US20220059236A1 - Method for modeling behavior and health changes - Google Patents

Method for modeling behavior and health changes Download PDF

Info

Publication number
US20220059236A1
US20220059236A1 US17/518,043 US202117518043A US2022059236A1 US 20220059236 A1 US20220059236 A1 US 20220059236A1 US 202117518043 A US202117518043 A US 202117518043A US 2022059236 A1 US2022059236 A1 US 2022059236A1
Authority
US
United States
Prior art keywords
patient
block
subgroup
treatment
patients
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Abandoned
Application number
US17/518,043
Inventor
Anmol Madan
Sai Thejasvee Moturu
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Orangedot Inc
Original Assignee
Orangedot Inc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Orangedot Inc filed Critical Orangedot Inc
Priority to US17/518,043 priority Critical patent/US20220059236A1/en
Assigned to Ginger.io, Inc. reassignment Ginger.io, Inc. ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: MADAN, Anmol, MOTURU, SAI THEJASVEE
Publication of US20220059236A1 publication Critical patent/US20220059236A1/en
Assigned to PACIFIC WESTERN BANK reassignment PACIFIC WESTERN BANK SECURITY INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: Ginger.io, Inc.
Assigned to OrangeDot, Inc. reassignment OrangeDot, Inc. MERGER AND CHANGE OF NAME (SEE DOCUMENT FOR DETAILS). Assignors: Ginger.io, Inc., OrangeDot, Inc., SPICER MERGER SUB, INC.
Assigned to Ginger.io, Inc. reassignment Ginger.io, Inc. RELEASE BY SECURED PARTY (SEE DOCUMENT FOR DETAILS). Assignors: PACIFIC WESTERN BANK
Assigned to OXFORD FINANCE LLC reassignment OXFORD FINANCE LLC SECURITY INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: Ginger.io, Inc.
Abandoned legal-status Critical Current

Links

Images

Classifications

    • GPHYSICS
    • G16INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16HHEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/50ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for simulation or modelling of medical disorders
    • GPHYSICS
    • G16INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16HHEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H40/00ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices
    • G16H40/60ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the operation of medical equipment or devices
    • GPHYSICS
    • G16INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16HHEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H40/00ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices
    • G16H40/60ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the operation of medical equipment or devices
    • G16H40/63ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the operation of medical equipment or devices for local operation
    • GPHYSICS
    • G16INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16HHEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/30ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
    • GPHYSICS
    • G16INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16ZINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS, NOT OTHERWISE PROVIDED FOR
    • G16Z99/00Subject matter not provided for in other main groups of this subclass
    • GPHYSICS
    • G16INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16HHEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H10/00ICT specially adapted for the handling or processing of patient-related medical or healthcare data
    • G16H10/20ICT specially adapted for the handling or processing of patient-related medical or healthcare data for electronic clinical trials or questionnaires
    • GPHYSICS
    • G16INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16HHEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H20/00ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance
    • G16H20/10ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance relating to drugs or medications, e.g. for ensuring correct administration to patients
    • GPHYSICS
    • G16INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16HHEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H40/00ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices
    • G16H40/60ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the operation of medical equipment or devices
    • G16H40/67ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the operation of medical equipment or devices for remote operation

Definitions

  • This invention relates generally to the field of patient health and more specifically to a new and useful method for modeling behavior and health changes in the field of patient health.
  • FIG. 1A is a flowchart representation of a first method of the invention
  • FIG. 1B is a flowchart representation of a variation of the first method
  • FIG. 1C is a flowchart representation of a variation of the first method
  • FIG. 1D is a flowchart representation of a variation of the first method
  • FIG. 2A is a flowchart representation of a second method of the invention.
  • FIG. 2B is a flowchart representation of a variation of the second method
  • FIG. 2C is a flowchart representation of a variation of the second method
  • FIG. 2D is a flowchart representation of a variation of the second method
  • FIG. 3 is a flowchart representation of variations in accordance with the first and second methods
  • FIG. 4 is a graphical representation of variations in accordance with the first and second methods.
  • FIG. 5 is a graphical representation of variations in accordance with the first and second methods.
  • a first method S 100 for supporting a patient associated with a health condition through a treatment regimen includes: accessing a first log of use of a native communication application executing on a mobile computing device by the patient within a first time period in Block S 110 ; receiving a first survey response corresponding to the first time period from the patient in Block S 112 ; estimating a first adherence to the treatment regimen by the patient within the first time period based on the first survey response in Block S 114 ; correlating the first log of use of the native communication application with the first adherence to the treatment regimen in Block S 116 ; accessing a second log of use of the native communication application by the patient within a second time period in Block S 120 ; receiving a second survey response from the patient within the second time period in Block S 122 ; estimating a second adherence of the patient within the second time period based on the second survey response in Block S 124 ; correlating the second log of use of the native communication application with the second adherence to the treatment regimen in Block
  • one variation of the first method S 100 includes: accessing a log of use of a native communication application executing on a mobile computing device by a patient in Block S 110 ; selecting a subgroup of a patient population based on the log of use of the native communication application and a communication behavior common to the subgroup in Block S 160 ; retrieving a regimen adherence model associated with the subgroup, the regimen adherence model defining a correlation between treatment regimen adherence and communication behavior for patients within the subgroup in Block S 162 ; predicting patient adherence to the treatment regimen based on the log of use of the native communication application and the regimen adherence model in Block S 144 ; and presenting a treatment-related notification based on the patient adherence through the mobile computing device in Block S 150 .
  • another variation of the first method includes: accessing a log of use of a native communication application executing on a mobile computing device by a patient within a period of time in Block S 110 ; selecting a subgroup of a patient population based on the log of use of the native communication application and a communication behavior common to the subgroup in Block S 160 ; retrieving a regimen adherence model associated with the subgroup, the regimen adherence model defining a correlation between treatment regimen adherence and communication behavior for patients within the subgroup in Block S 162 ; predicting adherence to the treatment regimen by the patient based on the log of use of the native communication application and the regimen adherence model in Block S 144 ; extracting a treatment response of the patient from a patient survey corresponding to the period of time in Block S 112 ; estimating an efficacy of the treatment regimen in treating a health condition of the patient according to a comparison between the treatment response and the adherence to the treatment regimen by the patient in Block S 172 ; transmitting a notification to a
  • yet another variation of the first method includes: accessing a log of use of a native communication application executing on a mobile computing device by the patient in Block S 110 ; selecting a subgroup of a patient population based on the log of use of the native communication application and a communication behavior common to the subgroup in Block S 160 ; retrieving a health risk model associated with the subgroup, the health risk model defining a correlation between risk of change in a medical symptom and communication behavior for patients within the subgroup in Block S 162 ; predicting a risk of change in a medical symptom for the patient based on the log of use of the native communication application and the health risk model in Block S 172 ; and transmitting a notification to a care provider associated with the patient in response to the risk of change in the medical symptom for the patient that exceeds a threshold risk in Block S 150 .
  • a second method S 200 includes: selecting a subgroup of patients from a patient population in Block S 220 , patients within the subgroup exhibiting similar behavioral characteristics and associated with a health condition; for a patient within the subgroup, estimating adherence of the patient to a prescribed treatment regimen during a period of time based on survey responses entered by the patient through a corresponding mobile computing device in Block S 230 ; for a patient within the subgroup, characterizing communication behavior of the patient based on use of a native communication application executing on a corresponding mobile computing device by the patient during the period of time in Block S 240 ; for a patient within the subgroup, correlating communication behavior of the patient with a health status of the patient in Block S 250 ; estimating an efficacy of the treatment regimen in treating the health condition for patients within the subgroup based on adherence to prescribed treatment regimens and health statuses of patients within the subgroup in Block S 260 ; and generating a treatment regimen report specific to the subgroup based on the effic
  • one variation of the method includes: selecting a population of patients prescribed a treatment regimen for a health condition in Block S 210 ; for a patient within the population, characterizing communication behavior of the patient based on use of a native communication application executing on a corresponding mobile computing device by the patient prior to initiation of a treatment regimen and during administration of the treatment regimen to the patient in Block S 240 ; selecting a subgroup of patients from the population of patients based on similar communication behaviors prior to initiation of the treatment regimen in Block S 220 ; extracting treatment responses from surveys completed by patients within the subgroup during administration of the treatment regimen in Block S 230 ; generating a treatment regimen model for the subgroup in Block S 280 , the treatment regimen model defining a correlation between communication behavior, treatment responses, and treatment regimen outcomes for patients within the subgroup; characterizing communication behavior of a subsequent patient based on use of a native communication application executing on a corresponding mobile computing device by the subsequent patient in Block S 242 ; and generating a predicted treatment regimen
  • another variation of the second method includes: selecting a subgroup of patients associated with a health condition from a population of patients, patients in the subgroup exhibiting similar behavioral characteristics in Block S 220 ; for patients within the subgroup, characterizing communication behavior of a patient based on use of a native communication application executing on a corresponding mobile computing device by the patient during the period of time in Block S 240 ; extracting characteristics of medical symptoms of patients within the subgroup from surveys submitted by patients within the subgroup in Block S 230 ; identifying a relationship between communication behaviors of patients within the subgroup, characteristics of medical symptoms of patients within the subgroup, and a treatment regimen administered to patients within the subgroup in Block S 232 ; generating a treatment efficacy model for the subgroup based on the relationship, the treatment efficacy model defining a correlation between a change in communication behavior and efficacy of the treatment regimen in alleviating medical symptoms of patients within the subgroup in Block S 280 .
  • yet another variation of the second method includes: selecting a subgroup of patients associated with a health condition from a population of patients, patients in the subgroup exhibiting similar behavioral characteristics in Block S 220 ; for patients within the subgroup, characterizing communication behavior of a patient based on use of a native communication application executing on a corresponding mobile computing device by the patient during the period of time in Block S 240 ; extracting characteristics of medical symptoms of patients within the subgroup from surveys submitted by patients within the subgroup in Block S 230 ; identifying a relationship between communication behaviors of patients within the subgroup and characteristics of medical symptoms of patients within the subgroup in Block S 232 ; generating a health risk model for the subgroup based on the relationship, the health risk model defining a correlation between a change in communication behavior and risk of change in a medical symptom in Block S 280 .
  • the first and second methods S 100 , S 200 function to collect communication data of a patient (a user or an at-risk individual) from a mobile computing device associated with the patient and to anticipate a health status of the patient based on the patient's communication data.
  • the methods can subsequently apply the anticipated health status of the patient to suggest an action to the patient and/or to inform a nurse, care provider, a family member, a pharmacist, a pharmacologist, an insurance producer, a hospital, or other health professional or network, etc. with anticipated health-related information of the patient.
  • the methods can additionally or alternatively implement an anticipated patient health status to drive automated or manual targeted intervention of for a patient via a phone call, email, health tip notification, insight, or other electronic communication.
  • the methods can also apply an anticipated health status of a patient to remind the patient to fulfill a treatment, to model the patient's progress through a treatment program or regimen, to predict an outcome of the patient's treatment program, to modify or customize a treatment program for the patient, to generate a model of treatment regimens and/or treatment regimen outcomes for a group of similar patients within a population of patients, to predict or monitor a change in patient symptoms or health status change, etc., such as based on volunteered survey data and/or volunteered or predicted treatment regimen adherence by the patient and/or other similar patients.
  • the methods collect data related to the patient's daily phone calls, text messages, and emails, including frequency, duration, length of time to respond to inbound communications, time-related communication patterns, and/or diversity of contacts through one or more native applications executing on a smartphone (or tablet) associated with the patient (e.g., a phone call application, a text messaging application, and an email application).
  • the methods can survey the patient for his health risk assessment, symptom score, and/or degree of current symptoms (e.g., how the patient is feeling) and then correlate the patient's health status, symptom score etc. to generate a health risk model.
  • the methods can generate a quantitative risk score corresponding to a predicted level of risk of change in a medical symptom for a patient based on a correlation between patient behavior and a corresponding health risk model.
  • the methods can subsequently implement the health risk model to predict a level of the patient's symptoms at a later time and/or if a notification pertaining to the patient's health should be generated based (solely or predominantly) on the patient's communication behavior through a computing device.
  • the first method S 100 can extrapolate a patient health status from patient communication data and survey responses and combine patient health status with patient communication behavior to generate a health risk model, and the first method can later anticipate a change in patient health risk by feeding subsequent patient communication data into the health risk model to determine subsequent patient health risk and generate a notification for the patient accordingly.
  • the methods collect data related to the patient's daily phone calls, text messages, emails, and/or other inbound and/or outbound communications from the patient's mobile computing device, including frequency, duration, length of time to respond to an inbound communication, time-related communication patterns, diversity of contacts, etc. through one or more native communication applications executing on a smartphone (or tablet) associated with the patient (e.g., a phone call application, a text messaging application, and an email application).
  • a smartphone or tablet
  • a phone call application e.g., a text messaging application, and an email application.
  • the methods can thus aggregate and manipulate any of the foregoing data to estimate patient adherence to a treatment regimen (e.g., whether the patient took his medications) and how the patient is ‘feeling’ (i.e., degree of symptoms) and subsequently correlate the patient's treatment adherence and feelings to the patient's communication behavior to generate a regimen adherence model.
  • the methods can subsequently predict how the patient is feeling and/or if the patient followed his treatment regimen based on the patient's communication behavior through his smartphone.
  • the first method S 100 can extrapolate patient treatment regimen adherence from a survey response, combine patient treatment regimen adherence with patient communication behavior to generate a regimen adherence model, and later anticipate a patient regimen adherence by feeding subsequent communication behavior into the regimen adherence model and then generate a notification for the patient accordingly.
  • the methods apply behavior data of the patient, such as prior to and/or during a treatment program or regimen, to select a regimen adherence model associated with a subgroup of patients exhibiting communication behaviors similar to those of a particular patient.
  • the methods can thus predict how the patient is feeling (i.e., presentation of the patient's symptoms) without additional patient info (e.g., additional survey results).
  • the methods can thus combine survey responses and patient communication behaviors to generate patient population subgroup models and later anticipate a patient health risk—and generate a notification accordingly—by feeding subsequent communication behavior of a patient into the model(s).
  • the first method S 100 applies behavior data of the patient, such as prior to and/or during a treatment program or regimen, to select a regimen adherence model associated with the subgroup exhibiting communication behaviors similar to those of the patient.
  • the first method S 100 can thus predict how the patient is feeling (i.e., presentation of the patient's symptoms) at a particular time and/or anticipate how or when the patient is fulfilling his treatment regimen, such as by consuming the prescribed medications at prescribed times and at prescribed dosages, without additional patient info (e.g., additional survey results).
  • this variation of the first method S 100 can thus extrapolate patient treatment regimen adherence from a survey response, combine patient treatment regimen adherence with patient communication behavior to select a patient population subgroup and a related regimen adherence model, and later anticipate a patient regimen adherence by feeding subsequent communication behavior into the regimen adherence model and then generate a notification for the patient accordingly.
  • the methods implement behavior data of a patient to select a health risk model associated with a particular subgroup exhibiting communication behaviors similar to those of the patient.
  • the methods can this predict how the patient is feeling (i.e., presentation of the patient's symptoms) at a particular time and/or anticipate how or when the a notifications indicating patient risk should be transmitted to a care provider, all without necessitating additional patient information (e.g. additional survey results).
  • the first method S 100 applies behavior data of the patient, such as prior to and/or during a treatment program or regimen, to select a regimen adherence model associated with the subgroup exhibiting communication behaviors similar to those of the patient.
  • the first method S 100 can thus predict how the patient is feeling (i.e., presentation of the patient's symptoms) at a particular time and/or anticipate how or when the patient is fulfilling his treatment regimen, such as by consuming the prescribed medications at prescribed times and at prescribed dosages, without additional patient info (e.g., additional survey results).
  • the second method S 200 collects communication data from multiple patients, groups patients according to identified communication behaviors, and anticipates progress through a treatment program and the effects of the treatment program on patients with a selected group. Rather than relying on data entered manually by patients and/or care providers, estimating patient activity or action from motion data (which can be imprecise, require substantial computing powers, and lack suitable precision for the health care space), etc., the second method S 200 can instead generate models linking treatments (e.g., pharmacotherapy regimens, physical therapy programs) to symptom presentation and final health condition outcome for patients based on how, when, and/or with whom, etc. patients communicate over time (e.g., on a daily or four-hour basis).
  • treatments e.g., pharmacotherapy regimens, physical therapy programs
  • the second method S 200 can thus select a subgroup of patients within a patient population based on communication behaviors of the patient population, estimate treatment regimen adherence within the subgroup based on survey responses volunteered by patients within the subgroup, estimate the health status of patients within the subgroup based on communication behaviors of the patients, and combine treatment adherence and communication behavior of patients within the subgroup to estimate the efficacy of the treatment regimen specifically for the subgroup of patients.
  • the second method S 200 can further apply such models to particular patients to anticipate how a patient will respond to a treatment regimen and/or to customize a treatment regimen for the patient based on similarities and/or differences between the patient's communication behaviors and communication behavior of other patients (e.g., previous or current patients with similar diagnosed health conditions and/or prescribed treatment regimens).
  • the first and second methods S 100 , S 200 can therefore derive meaningful health-related data for individual patients and/or groups of patients from patient communication behaviors extrapolated from phone call, text message, email, and/or other communication data collected passively through mobile computing devices associated with the patient(s).
  • Such derived health-related data can be implemented within a patient's mobile computing device to privately guide a patient through a treatment program by a care provider to anticipate risk of increasing medical symptoms or risk of a change in health status of the patient, by a doctor to monitor a patient's progress or to modify a prescription, by a nurse to identify a need to provide manual support to the patient, and/or by a pharmacologist to research drug therapies and drug responses within a population, etc.
  • Blocks of the first and second methods S 100 , S 200 can be implemented to collect patient data for an experiment, a research study, a commercial launch, a marketing study, a patient community study, such as on the recommendation of a healthcare provider (e.g., a doctor or a medical institution (e.g., a hospital center) to improve patient care.
  • a healthcare provider e.g., a doctor or a medical institution (e.g., a hospital center) to improve patient care.
  • the first and second methods S 100 , S 200 can interface with a patient-facing interface, a doctor-facing interface, a nurse-facing interface, and/or a pharmacologist-facing interface, etc. to deliver notifications and related data to the patient, doctor, nurse, pharmacologist, and/or other care provider directly or indirectly associated with the patient.
  • the first and second method S 100 , S 200 can be implemented on a computing device associated with a particular patient as a personal heath-tracking tool.
  • the first and second methods S 100 , S 200 interface with a native data collection application executing on a patient's mobile computing device (e.g., smartphone, tablet, personal data assistant (PDA), personal music player, vehicle, etc.) to retrieve patient communication data.
  • a patient's mobile computing device e.g., smartphone, tablet, personal data assistant (PDA), personal music player, vehicle, etc.
  • the native data collection application can be installed on the patient's mobile computing device, execute substantially continuously while the mobile computing device is in use and/or “ON,” record times, durations, and contact types (e.g., family, friend, coworker, business associate) of each inbound and outbound communication from the patient's mobile computing device.
  • the mobile computing device can then upload this data to a remote database, such as in real-time, every hour, at the end of each day, etc.
  • the first and/or second method implemented on a computer network (e.g., the “cloud”), can retrieve the patient's communication data from the remote database, analyze the patient's communication data to anticipate the patient's symptoms and/or therapy adherence, and generate a notification for the patient, and the patient's mobile computing device can download and subsequently display the notification for the patient, actions that can be handled automatically and in the background by the native data collection application or an alternative patient-facing native application executing on the mobile computing device.
  • the computer network can additionally or alternatively generate a patient regimen adherence model, a patient outcome model, a patient behavioral model, a health risk model etc. and transmit any one or more models to a doctor, a nurse, a pharmacologist, a therapist, etc.
  • Blocks of the first and second methods S 100 , S 200 can be implemented on one or more computer systems, such as a cloud-based computer system (e.g., Amazon EC3), a mainframe computer system, a grid-computer system, or any other suitable computer system.
  • Blocks of the first and second methods S 100 , S 200 can collect patient data from one or more devices over the Internet, such as communication data directly from a natively application executing on the patient's smartphone and/or motion data from a wearable sensor collected by the patient's smartphone and then uploaded to a remote database over an Internet connection.
  • Collection, manipulation, and transmission of patient data in Blocks of the first and second methods S 100 , S 200 can further adhere to health-related privacy laws, such as by privatizing or anonymizing patient data and transmitting encrypted or private notifications—such as pertaining to a patient's therapy adherence—to the patient and/or a doctor, a nurse, a pharmacologist, a researcher, etc. associated with the patient.
  • health-related privacy laws such as by privatizing or anonymizing patient data and transmitting encrypted or private notifications—such as pertaining to a patient's therapy adherence—to the patient and/or a doctor, a nurse, a pharmacologist, a researcher, etc. associated with the patient.
  • the native application can prompt the patient to create a profile or account.
  • the account can be stored locally on the patient's mobile computing device and/or remotely on the computer network and can contain a name, age, gender, location, occupation, list of health conditions, list of current health-related treatments and medications, medical history, a primary care physician (including name, office, contact information, etc.), health insurance information, prescription insurance plan, local pharmacy information, or other demographic and/or health-related information, any of which can be added by the patient, a family member, a doctor, a nurse, or other individual associated with the patient and/or retrieved from a medical, insurance, social network, or other related database.
  • the first and second methods S 100 , S 200 can also update the patient's account with additional patient information over time, such as presentation of symptoms, estimated patient treatment adherence, and/or predicted treatment regimen outcome for the patient.
  • Blocks of the methods can function to qualify and/or quantify a correlation between a health outcome or symptom risk and observed behavior changes for a patient based on data gathered passively from a patient's mobile computing device (e.g., mobile phone, smartphone and/or tablet, etc.) substantially without manual data entry from the patient, a doctor, a nurse, a pharmacist, etc.
  • Blocks of the methods can similarly quantify a correlation between a patient health status and an observed patient behavior gathered from the patient's mobile computing device and further combine passive behavior data collection with self-reported patient data (i.e., survey responses) to enhance the quality of a health-related model, extrapolated health-related notification triggers, patient inferences, patient symptom risk predictions, etc.
  • the first and second methods S 100 , S 200 can identify a relationship between a health outcome and a communication-related behavior change (e.g., a change in the frequency of phone calls, text messages, and emails over a preset period of time), such as changes in a patient's communication behavior from prior to administration of a treatment to during and/or after administration of the treatment for the patient.
  • a communication-related behavior change e.g., a change in the frequency of phone calls, text messages, and emails over a preset period of time
  • the first and second methods S 100 , S 200 can also identify a correlation between health outcome and patient activity data (i.e., accelerometer and gyroscope data from a corresponding mobile device or wearable sensor), local environmental data, patient location data, patient survey data, etc.
  • the first and second methods S 100 , S 200 can then implement an identified relationship between a (communication) behavior change and a change in a patient's health or quality of life over a period of time to anticipate a change in patient health risk, trigger or target an automated or manual intervention (e.g., an automated notification) to assist the patient through a treatment regimen, anticipate symptom relapse and estimate risk of hospital readmission for the patient, etc.
  • an automated or manual intervention e.g., an automated notification
  • the methods can similarly implement the foregoing identified relationship to identify categories, clusters, or subgroups of similar patients within a patient population, to modify or customize treatment regimens or patient health risk models for particular patient clusters, to prescribe a customized or subgroup-specific treatment regimen to a subsequent patient, to assist a care provider in patient triage, to support or justify adjustment to health care utilization, to compare health outcomes (e.g., admission/readmission rates and health care utilization) within and/or across a patient population, etc.
  • the methods can thus function to improve health outcomes for a particular patient and/or within a subgroup of a patient population and to improve health care utilization.
  • the first and second methods S 100 , S 200 can function to qualify and/or quantify a correlation between a health outcome related to a treatment, therapy, and/or medication, etc. and observed behavior changes based on data gathered passively from a patient's mobile computing device (e.g., mobile phone, smartphone and/or tablet, etc.) substantially without manual data entry from the patient, a doctor, a nurse, a pharmacist, etc.
  • a patient's mobile computing device e.g., mobile phone, smartphone and/or tablet, etc.
  • the first and second methods S 100 , S 200 can similarly quantify a correlation between a patient health status and an observed patient behavior based on patient data gathered from the patient's mobile computing device and combine passive behavior data collection with self-reported patient data (i.e., from surveys) to enhance the quality of a treatment regimen, inferences, and treatment predictions for a patient and/or a subgroup of patients.
  • the first and second methods S 100 , S 200 can identify a relationship between a health outcome and a communication-related behavior change (e.g., phone calls, text messages, emails), such as changes in a patient's communication behavior from prior to administration of a treatment to during and/or after the treatment for an individual patient.
  • a communication-related behavior change e.g., phone calls, text messages, emails
  • the methods can also identify a change in composite work-life balance for the patient, such as based on time spent by the patient in work-related phone calls or sending work-related emails or at a physical office or work location relative to a total communication time of the patient, a total patient waking time per day, or an amount of time spent at home.
  • the methods can thus correlate a change in patient composite work-life balance with a declining patient health status or increased patient health risk (e.g., for recurring or symptoms of increased severity) based on the health risk model.
  • the methods can compute a composite quality of life score for the patient, based on time spent in phone calls and at physical locations related to social events, familial and work phone calls, and work locations and transitions between such locations, and the methods can correlate a change in composite quality—of-life with a declining patient health status or increased patient health risk based on the health risk model.
  • the first and second methods S 100 , S 200 can additionally determine a correlation between health outcome and patient mobile phone usage (i.e. screen unlocks, mobile application use) derived from an operating system or the task manager executing on the patient's mobile computing device to determine periods of patient activity, hyper-activity (e.g. frequent unlocks late at night when the patient is unable to sleep), and inactivity.
  • the first and second methods S 100 , S 200 can then implement an identified relationship between a mobile usage behavior change and a change in a patient's health or quality of life over a period of time based on a corresponding health risk model.
  • the first and second methods S 100 , S 200 can also determine a correlation between health outcome and patient activity data (i.e., accelerometer and gyroscope data from the mobile computing device or external wearable sensor), local environmental data, patient location data, patient survey data, etc.
  • patient activity data i.e., accelerometer and gyroscope data from the mobile computing device or external wearable sensor
  • local environmental data i.e., local environmental data, patient location data, patient survey data, etc.
  • the first and second methods S 100 , S 200 can then implement an identified relationship between a (communication) behavior change and a change in a patient's health or quality of life over a period of time to anticipate a treatment efficacy and/or patient response to a treatment, estimate a patient's treatment adherence (i.e., compliance), determine patient satisfaction in a treatment, trigger or target automated or manual interventions (e.g., automated notifications) to assist a patient through a treatment, anticipate symptom relapse and estimate risk of hospital readmission.
  • a (communication) behavior change and a change in a patient's health or quality of life over a period of time to anticipate a treatment efficacy and/or patient response to a treatment, estimate a patient's treatment adherence (i.e., compliance), determine patient satisfaction in a treatment, trigger or target automated or manual interventions (e.g., automated notifications) to assist a patient through a treatment, anticipate symptom relapse and estimate risk of hospital readmission.
  • the first and second methods S 100 , S 200 can similarly implement such an identified relationship to identify categories, clusters, or subgroups of similar patients within a patient population, to modify or customize treatment regimens or notifications for particular patient clusters, to prescribe a customized or subgroup-specific treatment regimen to a subsequent patient, to assist a care provider in patient triage, to support adjustment health care utilization, to compare treatment outcomes (e.g., admission/readmission rates and health care utilization) within and/or across a patient population, etc. to improve treatment outcomes, therapy efficacy, and adherence to a treatment program for a particular patient and/or within a subgroup of a patient population, and health care utilization.
  • treatment outcomes e.g., admission/readmission rates and health care utilization
  • Blocks of first and second methods S 100 , S 200 can implement similar functionality and techniques to identify patient risk and assess treatment efficacy.
  • Block S 110 of the first method S 100 recites identifying a first log of use of a native communication application executing on a mobile computing device by the patient within a first time period.
  • Block S 120 and Block S 140 similarly recite identifying a second log and a third log of use of the native communication application by the patient within a second and a third time period, respectively.
  • Blocks S 110 , S 120 , and S 140 function to unobtrusively collect and/or retrieve communication-related data from a patient's mobile computing device, such as through integration within or by interfacing with a native data collection application to collect patient data, as described above.
  • the native data collection application can launch on the patient's mobile computing device as a background process that gathers patient data once the patient logs in to his account.
  • Block S 110 collects communication data and/or native communication application usage data, generated by the patient's mobile computing device, to identify how and how often (i.e., with what frequency) the patient interacts with and communicates with other individuals through phone calls, e-mail, instant messaging, an online social network, etc.
  • Block S 110 collects phone call-related data, including a number of sent and/or received calls, call duration, call start and/or end time, location of patient before, during, and/or after a call, and number and times of missed or ignored calls.
  • Block S 110 , etc. can also collect text messaging (e.g., SMS test messaging) data, including number of messages sent and/or received, message length, message entry speed, efficiency, and/or accuracy, time of sent and/or received messages, and location of the patient when receiving and/or sending a message.
  • text messaging e.g., SMS test messaging
  • textual messages sent through other communication venues can collect similar types of data on textual messages sent through other communication venues, such as public and/or private textual messages sent to contacts of the patient through an online social networking system, reviews of products, services, or businesses through an online ranking and/or review service, and/or any other text-based communication generated by the patient and communicated to another individual and/or computer network.
  • Block S 110 , etc. can further collect location data of the patient before, during, and/or after (or in the absence of) communication with another individual (e.g., a phone call) and/or computer network (e.g., a social networking message), such as by retrieving a GPS location from a GPS sensor within the patient's mobile computing device, estimating the location of the patient's mobile computing device through triangulation of local cellular towers, or identifying a geo-located local Wi-Fi hotspot, etc., during a phone call.
  • Block S 110 , etc. can apply this data to track patient behavior characteristics, such as patient mobility, periods of patient isolated, patient work-life (e.g., balance based on time spent at specific locations), etc.
  • Block S 110 , etc. can therefore track the patient's mobility during a communication.
  • Block S 110 , etc. can additionally or alternatively collect data pertaining to individuals in contact with the patient during the first period of time, etc., such as an individual's location during a phone call, phone number, contact duration and/or type with the patient, relationship to the patient or patient contact group (e.g., top contact, spouse, family member, friend, coworker, business associate, etc.) specified by the patient or learned from previous patient communications, etc.
  • Blocks S 110 , S 120 , etc. can further capture mobile usage data like screen unlocks and mobile application usage, such as by retrieving usage information from mobile operating system logs or task manager on the mobile computing device.
  • Blocks of the methods can therefore track variations and periods of activity and inactivity for a patient through data automatically collected on the patient's mobile computing device, such as to estimate extended periods when the patient was hyperactive on the device or not asleep.
  • Block S 110 also collects or retrieves patient physical activity- or physical action-related data (e.g., accelerometer and gyroscope data), local environmental data, patient nutrition or diet-related data, etc. such as recorded through sensors within the patient's mobile computing device or through a wearable or other peripheral device in communication with the patient's mobile computing device.
  • patient physical activity- or physical action-related data e.g., accelerometer and gyroscope data
  • local environmental data e.g., patient nutrition or diet-related data, etc.
  • patient nutrition or diet-related data e.g., etc., etc.
  • a wireless-enabled scale, blood pressure sensor, and a pulse-dosimeter sensor can transmit the patient's weight, blood pressure, and blood oxygen level to the patient's mobile computing device, and Block S 110 can add this data to the patient's account to further augment patient behavior data.
  • Block S 110 can subsequently aggregate phone, text message, email, social networking, and/or other patient communication data for a particular period of time into a qualitative and/or quantitative feature for the patient for the particular time period.
  • the feature can be specific to a day, a week, a month, a day period (e.g., morning, afternoon, evening, night), a time block during a day (e.g., one hour), a specific communication action (e.g., a single phone call, a set of communication actions of the same type (e.g., a set of phone calls within a two-hour period), all communications within a period of time, etc.).
  • Block S 110 can generate a weighted composite of the frequency, duration (i.e., length), timing (i.e., start and/or termination), and contact diversity of all outgoing voice (e.g., phone call) communications and a frequency, length, and timing and/or response time to (i.e., time to accept) incoming voice communications within the first period of time through a phone call application executing on the patient's mobile computing device.
  • Block S 110 can additionally or alternatively assess incoming and/or outgoing textual communications from a textual messaging application executing on the mobile computing device.
  • Block S 110 can also generate a quantitative assessment of a frequency of, a duration of, and a response time to both incoming and/or outgoing phone calls and textual communications to the mobile computing device during the first time period as a single qualitative and/or quantitative feature corresponding to the first period of time.
  • Block S 120 and/or S 140 can implement similar methods to generate a feature for the second period of time and third period of time, respectively.
  • Blocks S 112 , S 122 , etc. can further extract features based on voice communications, textual communications, mobile application activity usage, location data, etc., which can be based on variance, entropy, or other mathematical and probabilistic computations of basic data, such as a composite activity score, a composite socialization score, a work-life balance score, a quality-of-life score, etc.
  • Block S 110 implements machine learning, data mining, and statistical approaches to process patient communication data into relevant patient communication behavior features (e.g., data points).
  • Block S 110 can implement similar techniques to similarly process patient motion data, local environmental, and other automatically/passively collected data.
  • Block S 240 of the second method S 200 recites, for patients within the subgroup, characterizing communication behavior of a patient based on use of a native communication application executing on a corresponding mobile computing device by the patient during the period of time.
  • Block S 240 can thus implement functionality similar to Block S 110 , Block S 120 , etc. to extrapolate patient communication behavior from phone calls, text messages, social networking communications, emails, and/or other communications originating and/or terminating through one or more communication applications executing on one or more mobile computing devices associated with one or more patients.
  • Block 240 characterizes communication behavior of multiple patients based on use of native communication applications executing on corresponding mobile computing devices by the patients both prior to initiation of and during administration of an action item specified in an automated notification directed to the patient and/or an associated nurse, doctor, care provider, etc.
  • Block S 240 characterizes communication behavior of multiple patients based on use of native communication applications executing on corresponding mobile computing devices by the patients both prior to initiation of similar treatment regimens and during administration of the treatment regimens.
  • Block S 242 of the second method which recites characterizing communication behavior of a subsequent patient based on use of a native communication application executing on a corresponding mobile computing device by the subsequent patient, can also implement such functionality to characterize communication behavior of a new (i.e., subsequent) patient.
  • Block S 242 can pass the communication behavior characterization of the new patient to Block S 282 , and Block S 282 can determine a similarity between the new patient and a subgroup of patients based on communication behaviors of the two and, from this determined similarity and an effectiveness of the health-related notifications within the subgroup, generate a health-related notification for the new patient.
  • Block S 242 passes the communication behavior characterization of the new patient to Block S 282 , wherein Block 282 determines a similarly between the new patient and a subgroup of patients based on communication behaviors of the two and, from this determined similarity, predicts an effectiveness in a treatment regimen for the new patient based on the effectiveness of the treatment regimen within the subgroup.
  • Blocks S 110 , S 120 , S 140 , S 240 , S 242 , etc. can function in any other way to identify and/or characterize use of a native communication application by a patient within a period of time.
  • Block S 112 of the first method S 100 recites receiving a first survey response corresponding to the first time period from the patient.
  • Block S 122 similarly recites receiving a second survey response from the patient within the second time period.
  • Block S 112 , etc. functions to prompt the patient to self-report additional health-related data that can be implemented within the first and/or second method 5200 to qualify patient communication data, such as to teach a patient regimen adherence model relating patient communication to patient treatment adherence and/or patient symptom presentation.
  • Block S 112 transmits a survey to the patient's mobile computing device
  • the native data collection application executing on the mobile computing device opens the survey and prompts the patient to enter relevant data
  • Block S 112 receives patient responses to the survey from the mobile computing device once the survey is complete.
  • Block S 170 can function to extract a treatment response of the patient from a corresponding survey.
  • Block S 112 can generate the survey that includes prompts to enter a pain level (e.g., on a body-location-specific basis), presentation of symptoms (e.g., on a health-condition-specific basis), adherence to a treatment regimen (e.g., if and when the patient took a prescribed medication of some dosage), changes in presentation of symptoms after a treatment, how the patient “feels” generally, experience with a treatment and corresponding effect, a mood, a sleep quality or wakefulness, etc.
  • the survey can also prompt the patient to supply information related to a diagnosed disease or condition, such as major depressive disorder, diabetes, or chronic obstructive pulmonary disease.
  • Block S 112 can retrieve relevant patient health data from a medical record, history, or profile of the patient.
  • Block S 112 can import a patient medication record from an Electronic Medical Records (EMR) system hosted by a healthcare provider or health insurance company, such as via a supported Application Programmable Interface (API).
  • EMR Electronic Medical Records
  • API Application Programmable Interface
  • Block S 112 can prompt the patient to fill out a two-question survey every morning at 9:00 or after each meal, including whether he has taken prescribed medication(s) and his overall satisfaction with the treatment in mitigating symptoms.
  • presentation of surveys can be triggered by a determined patient behavior or symptom change.
  • Block S 112 can trigger presentation of a survey to the patient in response to a disparity in actual and anticipated patient communication behavior, such as significantly more or significantly less phone call time by the patient than expected based on past patient communication behavior.
  • Block S 112 can trigger delivery of a pain survey launched within the application either at a pre-determined time and/or once a period of time (e.g., 72 hours) transpires without a detected change in the patient's location (e.g., the patient has not left his house) and/or without a detected phone call initiated by the patient.
  • a period of time e.g., 72 hours
  • Surveys can include a single or combination of question-and-answer types, such as multiple choice questions each with a single answer option, multiple choice questions each with multiple answer options, textual or numerical manual entry, slider (e.g., for easy number or level selection from a range), icon animation selection (e.g., icons indicating different intensities or symptoms), etc., as shown in FIG. 1A .
  • a survey can include a Patient Health Questionnaire (PHQ9), a WHO Well-being Index (WHO5), or a one-question Pain Rating Scale question.
  • Block S 112 can additionally or alternatively generate a prompt targeted at a family member, care providers, health organization, etc. associated with the patient.
  • Block S 112 can prompt and collect survey responses from a nurse, including manually-entered patient symptom and treatment data.
  • Block S 112 can retrieve this information from the personal mobile computing device of a family member of the patient.
  • Block S 112 can similarly prompt a pharmacist or pharmacologist associated with the patient's treatment to enter patient prescription data directly into a survey.
  • Block S 112 can retrieve this data automatically from an associated pharmacy database.
  • Block S 112 can collect patient treatment data from a third-party system or device used by the patient or associated care provider, such as an Internet-enabled pillbox with embedded sensors.
  • Block S 112 can further function to compensate the participant for completing a survey, such as with cash, a gift card, a pharmacy discount, a health insurance discount, etc., such as in an application of the first and/or second method S 200 in which patient data is supplied to pharmaceutical researches to generate treatment regimens.
  • Block S 112 can be implemented within the native data collection application, on a computer network in communication with the mobile computing device (e.g., via an Internet connection), or in any other suitable way.
  • Block S 122 , etc. can implement similar techniques or functions, though Block S 112 , Block S 122 , etc. can function in any other way to collect survey responses corresponding to various time periods from the patient.
  • the first and second methods S 100 , S 200 can store data locally on the patent's mobile computing device and/or in a remote database on a computer network.
  • private health-related patient data can be stored temporarily on the patient's mobile computing device in a locked and encrypted file folder on integrated or removable memory.
  • the patient's data can be encrypted and uploaded to the remote database once a secure Internet connection is established.
  • patient data can be stored on any other local device or remote data in any other suitable way and transmitted between the two over any other connection via any other suitable communication and/or encryption protocol.
  • Block S 114 of the first method S 100 recites estimating a first adherence to the treatment regimen by the patient within the first time period based on the first survey response.
  • Block S 124 of the first method S 100 similarly recites estimating a second adherence of the patient within the second time period based on the second survey response.
  • Block S 230 of the second method S 200 recites, for patients within the subgroup, estimating adherence of a patient to a prescribed treatment regimen during a period of time based on survey responses entered by the patient through a corresponding mobile computing device.
  • Blocks S 114 , S 124 , S 230 , etc. function to extract the patient adherence to a treatment regimen over time from corresponding survey responses collected in Blocks S 112 , S 122 , etc.
  • a survey can include an explicit inquiry into if, when, and in what dosage the patient took a medication specified in a pharmacotherapy regimen and/or if the patient completed and the duration of a physical therapy session, and Block S 114 can identify the patient's response to the prompt, compare the patient's response to a treatment regimen assigned to the patient for the corresponding time period, and thus determine if and/or to what extent the patient fulfilled the treatment regimen for the corresponding time period. Block S 114 can subsequently pass the degree of patient adherence to the treatment regimen to Block S 116 .
  • Block S 124 and Block S 230 , etc. can implement similar functionality.
  • Block S 116 of the first method S 100 recites correlating the first log of use of the native communication application with the first adherence to the treatment regimen
  • Block S 126 similarly recites correlating the first log of use of the native communication application with the first adherence to the treatment regimen
  • Block S 126 recites correlating the second log of use of the native communication application with the second adherence to the treatment regimen.
  • Block S 116 , Block S 126 , etc. function to define a relationship between a degree of patient adherence to a treatment regimen and a quality and/or quantity of communication (i.e., communication behavior) of the patient for a corresponding period of time.
  • Block S 116 characterizes patient-provided treatment adherence data from the first survey and associates the treatment adherence characterization with the communication behavior characterization from Block S 110 .
  • the first method S 100 can thus anticipate the patient's treatment adherence at the later time to mimic the patient's treatment behavior during the first time period.
  • Block S 116 correlates a log of use of a first set of outgoing voice communications from a phone call application during the first period with conformity to the treatment regimen by the patient, and Block S 126 correlates a frequency of a second set of outgoing voice communications within the phone call application with neglect of the treatment regimen by the patient, as shown in FIG. 4 , wherein the log of use of the first set of outgoing voice communications is greater than the frequency of the second set of outgoing voice communications.
  • Block S 116 can additionally or alternatively correlate adherence to the treatment regimen and/or patient communication behavior to patient health status, such as presentation of patient symptoms.
  • Block S 116 extracts a patient symptom level from the first survey and presentation of patient symptoms (e.g., how well the patient is feeling) with patient communication behavior (e.g., how, when, and who the patient contacts during a period of time).
  • Block S 116 can determine that a period during which the patient sends a low volume of text messages to a large breadth of contact types and engages in long phone conversations with a limited number of contacts corresponds to minimal symptom presentation, and Block S 126 can determine that a period during which the patient sends a low volume of text messages to a low number of unique contacts and engages in short phone conversations with a single contact corresponds to a high degree of symptom presentation.
  • Block S 116 can also correlate minimal symptom presentation with a certain degree of treatment adherence during the first period (e.g., the patient followed the treatment adherence properly), and Block S 126 can further correlate higher degree of symptom presentation with a lesser degree of treatment adherence during the second period (e.g., the patient missed a treatment or took an improper dosage).
  • the first method S 100 can repeat elements of Blocks S 110 , S 112 , S 114 , S 116 , etc. to generate a set of communication behavior-treatment adherence (and health status) features for the patient over time, and these features can be fed into Block S 130 , Block S 160 , Block S 144 , Block S 150 , Block S 172 , and/or Block S 180 , etc. to enable further functionality.
  • Block S 250 of the second method which recites correlating communication behavior of a patient with health statuses of patients within the subgroup, can implement similar functionality to generate communication behavior-treatment adherence (and health status) features within the subgroup of patients.
  • Block S 250 can pair anonymized communication data from various patients exhibiting similar behaviors over time, such as before, during, and after a treatment regime, with survey responses from corresponding patients at corresponding times to output a collection of features specific to the subgroup.
  • Block S 250 can thus pass this collection of features to Block S 280 .
  • Block S 280 can thus implement the collection of features to identify trends or patterns between communication behavior, treatment adherence, and presentation of symptoms across the population and generate a corresponding treatment regimen model for the subgroup, as described below.
  • Block S 160 of the first method S 100 recites selecting a subgroup of a patient population based on the first log of use of the native communication application and a communication behavior common to the subgroup.
  • Block S 160 functions to identify one or more patients within a patient population exhibiting characteristics similar to those of the (current, new, or subsequent) patient, such that subsequent Blocks of the first method S 100 can apply preexisting communication behavior data, treatment adherence data, health status data, etc. and corresponding characterizations, patterns, models, etc. of this data to inform and/or improve manipulation of the patient's health-related data.
  • Block S 110 characterizes patient communication behavior prior to beginning a treatment regimen
  • Block S 160 selects a set of other patients within a patient population who exhibit(ed) similar communication behaviors prior to beginning similar treatment regimens.
  • Block S 160 can further select and/or filter the patient population based on a diagnosis, therapy prescription, age, gender, location, and/or other demographic, activity behaviors prior to beginning similar treatment programs, and/or other factors or variables similar to those of the patient.
  • Block S 160 can further select the set of other patients and/or update the set of other patients during the patient's treatment regimen based on similar in-treatment communication behaviors (and other behaviors or variables) between the patient and other patients in the patient population.
  • Block S 160 can therefore receive patient communication behavior characterizations of the patient and other patients within the patient population, such as from Block S 110 and Block S 120 , to identify communication behaviors common to the patient and to the subgroup.
  • Block S 160 can also select the subgroup as pertinent to the patient based on treatment adherence data and/or trends and/or symptom data and/or trends common to the patient and the subgroup, such as before and/or during a related treatment regimen.
  • Block S 160 can select the subgroup based on determined relationships between treatment adherence and communication data common to the patient and the other patients within the subgroup.
  • Block S 210 of the second method which recites identifying a population of patients diagnosed with a health condition
  • Block S 220 which recites selecting a subgroup of patients within the population exhibiting a similar behavioral characteristic
  • Block S 210 can implement similar functionality to identify a subset of patients within a patient population that share commonalities, such as one or more diagnoses, health conditions, symptoms, treatment regimens, communication behavior characteristics, treatment adherence characteristics, etc.
  • Block S 210 can identify a population of patients prescribed a treatment regimen for a particular diagnosed health condition, such as by identifying anonymized patient profiles tagged with the particular diagnosed health condition
  • Block S 220 can select a subgroup of patients from the population of patients based on similar communication behavior prior to initiation of the treatment regimen in Block S 220 .
  • Blocks S 230 , S 240 , and S 250 can therefore analyze and manipulate communication and/or survey data associated with selected patients within the subgroup (i.e., to estimate treatment regimen adherence, to characterize communication behavior, and to correlate communication behavior and health status of patients, respectively) before and/or after the subgroup is selected in Block S 220 .
  • Blocks S 160 , S 210 , and S 220 can function in any other way to select a subgroup of related patients from a population of patients.
  • Blocks of the first and second methods S 100 , S 200 can implement a Behavior Feature Engine (BFE) to handle patient data before, during, and/or following a treatment regimen and to then to manipulate patient data into an effective model.
  • BFE Behavior Feature Engine
  • the BFE includes a collection of statistics—a combination of unique variables or “features”—that vary amongst patients, conditions, and disease states. A subset of features defined in the BFE can thus be implemented to predict the health status of a patient, such as given communication behavior and/or treatment adherence of the patient.
  • the methods can collect raw patient data from the data collection application described above, such as through Block S 110 , S 112 , etc. and can then convert this raw data into statistical features to create the BFE, such as on a remote server for a subgroup of patients or locally on a mobile computing device for a particular patient.
  • the BFE can extract independent features from one of various types of data or modalities (e.g., phone calls, text messages, instant messages). For example, the BFE can aggregate patient communication data, including a total number of calls made, received, accepted, missed, etc. and a total number of text (e.g., SMS text) messages sent, received, ignored, etc. on a particular day or within another period of time, as shown in FIG. 1A .
  • data or modalities e.g., phone calls, text messages, instant messages.
  • the BFE can aggregate patient communication data, including a total number of calls made, received, accepted, missed, etc. and a total number of text (e.g., SMS text) messages sent, received, ignored, etc. on a particular day or within another period of time, as shown in FIG. 1A .
  • the BFE can estimate an interaction balance for the patient, such as by calculating a ratio of incoming communications (e.g., phone calls and text messages) to total communication within a period of time, as well as a patient responsiveness, such as by calculating a number of “missed” interactions within the time frame based on a number of calls not accepted or text messages ignored by the patient with the period of time.
  • incoming communications e.g., phone calls and text messages
  • the BFE can also calculate patient interaction diversity, including a total number of individuals with whom the patient interacts, such as through a voice communication application or text messaging application, within the period of time.
  • the BFE can further extract a patient mobility from patient communication location data, including an approximation of total patient movement during a particular phone call or during the total of a corresponding period of time, such as based on an estimated distance traveled by foot or bike (excluding a distance traveled by car, train, etc.) based on GPS location data of the patient's mobile computing device (e.g., smartphone).
  • the BFE can then generate a mobility radius factor—defining an approximate radius of an imaginary circle encompassing locations visited by a patient within a period of time—and tag or associate patient communication data with this mobility radius factor.
  • the BFE can also calculate other location data features, such as time spent at home, time spent at work, transition between home/work/social locations, isolation at a particular location for extended periods of time, unpredictability (i.e., entropy) in location information, etc.
  • the BFE can also implement mobile phone usage, such as screen unlocks and application usage, to calculate features that summarize variations and periods of patient activity and inactivity, identify extended periods during which the patient is hyperactive on the device, and/or determine sleep patterns.
  • the BFE can further handle other types of patient data.
  • the BFE can estimate an amount of physical activity undertaken by the patient within the period of time based on motion sensor (e.g., accelerometer) data recorded on the patient's mobile computing device or wearable device, as shown in FIG. 1B .
  • the BFE can also implement machine learning, such as feature extraction or pattern matching, to correlate motion data with a particular type of action, such as walking, playing tennis, or eating.
  • the BFE can then associate communication data (e.g., times and types) with the activity data, such as an amount of time the patient spent in phone calls during an activity, set of activities, or period of time including one or more particular activities.
  • the BFE can handle, manipulate, and/or aggregate any other type of patient data, such as any of the data collected as described above.
  • the BFE can segment patient behavior to enable identification of patient behavioral patterns.
  • the BFE can segment patient data according to any suitable period of time or time frame, such as by whole day, a preset number of hours, daytime and nighttime, weekdays and weekends, months, etc., and thus extrapolate an hourly, daily, weekly, or other time-dependent behavioral pattern.
  • the BFE can similarly segment patient behavior data by patient location, such as home, office, etc., and thus identify location-related patient behavioral patterns.
  • Feature extraction of patient data within the BFE can be driven by predefined metrics, such as distinctness, aggregation, entropy, or percentage change over time for a particular patient, and the BFE can group such features by time, patient, patient subgroup, etc.
  • Features can be stored as standalone features or, as behavior changes, features generated through comparison of two or more standalone features from different times (or across a group or in comparison to a baseline feature).
  • the BFE can then extract a time-dependent behavior pattern for a patient or a subgroup of patients from these features. For example, the BFE can identify and measure statistically significant changes in patient behavior throughout a treatment, including before, during, and/or after administration of the treatment (e.g., a pharmacological treatment, physical therapy, etc.).
  • the BFE can also generate comparative representations of base patient behavior features. For example, the BFE can compare a patient behavior feature value for a specific period of time against any one or more of a baseline (i.e., typical or normal) behavior for the patient as indicated by patient data gathered over time, historical patient behavior data (e.g., at a similar times of day or days of a week), an expected or healthy behavior as suggested by an expert, healthcare professional, literature, a past study, and/or an average or common behavior observed in patient population or patient subgroup exhibiting behaviors similar to the those of the patient.
  • the BFE can output such comparative representations in the form of differences, ratios, percentages, etc. and can handle the comparative representations as discrete features.
  • Blocks S 130 , S 280 , S 284 of the first and second methods S 100 , S 200 recite generating models based on communication and survey data of one or more patients.
  • Blocks of the first and second methods S 100 , S 200 can implement the BFE in conjunction with a Predictive Modeling Engine (PME) to generate patient-, patient subgroup-, and/or treatment-specific models.
  • PME Predictive Modeling Engine
  • the PME functions to identify and measure statistically significant changes in patient behavior through periods of varied symptom presentation or throughout a treatment, including before, during, and/or after administration of the treatment (e.g., a pharmacological treatment, physical therapy, etc.), to generate a (patient- or subgroup-level) predictive model accordingly, and to train the predictive model with additional patient data over time.
  • the PME can identify correlations between patient behavior(s), symptoms, treatment adherence, and other features extracted in the BFE to transform a large amount of patient data into a predictive model defining links between such observed patient data and associated efficacy, mitigation, abatement of patient symptoms and/or health condition via one or more treatments (i.e., therapies), patient wellness, quality of life, and/or general health of the patient.
  • the PME can therefore implement patient communication data to construct a patient behavior model from multiple patient-related features, such as including total call duration over a period of time (e.g., for a day, for a week), proportion of time spent on calls on weekdays and weekends (or days and nights), number of unique contacts in communication with the patient, and percentage change in patient communication with certain contacts over a baseline period of time.
  • patient-related features such as including total call duration over a period of time (e.g., for a day, for a week), proportion of time spent on calls on weekdays and weekends (or days and nights), number of unique contacts in communication with the patient, and percentage change in patient communication with certain contacts over a baseline period of time.
  • the PME can also associate such features—relative to the baseline—with patient treatment adherence, symptom presentation, and health condition progression to generate a holistic patient- or subgroup-specific treatment model.
  • the PME generates a model including a derived link between patient health status and patient behavior features extracted in the BFE by identifying and measuring statistically significant changes in behavior during days characterized by enhanced patient symptoms and days characterized by relatively normal patient symptoms for an individual patient or across a patient population subgroup.
  • the PME can identify a correlation between low patient activity and enhanced symptoms and generate a predictive model accordingly.
  • Such data can be applied to a particular patient, a patient subgroup, and/or a patient population to quantify how effectively a treatment enables a patient to be more active and return to a normal work-life balance in comparison to an alternative treatment.
  • the PME analyzes behavior data of a bipolar patient to correlate patient behaviors with patient symptoms.
  • the PME and BFE can cooperate to identify baseline patient communication behavior and associate this with a normal patient state, identify a period of low patient activity and isolated communication behavior and associate this with a depressive episode, and identify a period of patient hyper-activity, long work hours, and a high degree of socializing and associate this with a manic episode.
  • the PME can thus generate a patient-specific model relating patient behavior with normal, manic, and depressive states.
  • the PME can therefore manipulate changes in patient activity or behavior, work, work-life balance, productivity, stress and anxiety, health and wellness, treatment outcome (e.g., self-reported by the patient or entered by a caregiver, family-member, provider, employer, etc.), etc. to identify correlations between patient behavior and patient health and build predictive models based on behavior change, treatment adherence, symptom presentation, and/or treatment outcome patterns.
  • the PME can implement machine learning, data mining, and/or statistical approaches to generate one or more models specific to a patient, a patient subgroup, or a patient population.
  • the PME can apply statistical approaches to inform an understanding of underlying patient behavioral data through distributions, correlations, hypothesis testing, etc. by pairing original features extraction within the BFE with clinical insight and domain-specific intuition, such as from a doctor, nurse, pharmacist, etc.
  • the PME can apply feature selection approaches to determine a most predictive subset of features.
  • the PME can implement correlation-based feature selection (CFS), minimum redundancy maximum relevance (mRMR), Relief-F, symmetrical uncertainty, information gain, or other statistical methods and statistic fitting techniques to select a set of features output from the BFE.
  • CFS correlation-based feature selection
  • mRMR minimum redundancy maximum relevance
  • Relief-F symmetrical uncertainty
  • information gain or other statistical methods and statistic fitting techniques to select a set of features output from the BFE.
  • the PME can also implement support vector machines, ensemble learning (e.g., boosting, random forest), logistic regression, Bayesian learning (e.g., wherein an outcome variable is a predefined class or category), or other machine learning techniques to predict patient behavior, symptoms, treatment outcome, etc.
  • additional complexity can be added to this process.
  • the PME can implement cost-sensitive learning and sampling approaches when distribution of data between categories or classes is unequal, such as when a symptom of interest is depression, which is detectable within a patient on fewer than 10% of total days of a patient study.
  • the PME applies regression approaches and generalized linear models for model outputs that include continuous values, such as prediction of symptom severity.
  • the PME applies Support Vector Machine (SVM) classification approaches for model outputs that include discrete values, such as for patient health risk prediction.
  • the PME applies clustering techniques for model outputs that include a set of undefined clusters, such as a centroid-based, density-based, connectivity-based, and/or a distribution-based approach to define a cluster based on patient behaviors, as shown in FIG. 5 .
  • the PME can apply clustering techniques for a model output that specifies a group of patients with similar symptoms or behaviors together.
  • the PME can apply similar techniques to detect anomalies and/or outliers in aberrant cases in collected data, such as patients within a population who do not respond to a treatment or show adverse effects uncharacteristic of the greater patient population.
  • the PME can further train a linear model by fitting a line to collected data or features, and the PME can train a Bayesian network classification model and/or a Gaussian mixture classification model through Expectation-Maximization.
  • the PME can additionally or alternatively model subgroups of patients with a population of patients.
  • the PME identifies a subset (i.e., subgroup) of patients with a certain degree of similarity with one subset of patients in the population and a certain degree of dissimilarity with another subset of patients in the population, such as with respect to one or more behaviors. For example, for a particular therapy, the PME can select a first subgroup of patients who become more active during the therapy and a second subgroup—exclusive of the first subgroup—of patients who show little marked behavior change during the therapy. In this example, the PME can also assign different intervention triggers, reminder types, notification triggers, etc. to each of the first and second subgroups. The PME can therefore apply behavior data passively collected from a patient population to identify clusters of related patients within the population, as shown in FIG. 5 .
  • the PME can generate models that output predicted patient treatment adherence, patient symptom presentation, treatment efficacy, patient response to a treatment, etc.
  • the PME can generate models that additionally or alternatively incorporate or output one or more of a risk of admission or re-admission to an associated care provider, risk of a pending emergency room visit, risk of retrogression of clinical symptoms, etc.
  • a model generated through the PME can therefore support or enable early detection of higher-risk patients, prediction of patient hospitalization or symptom relapse, and timely patient intervention and messaging to improve or change patient behaviors.
  • the PME can further train or update a model over time, such as when new data from a specific patient or anonymized data from a patient subgroup becomes available. For example, the PME can verify a previous prediction outcome model for a specific patient in response to receiving feedback from the patients. Alternatively, the PME associates the patient with an alternative subgroup and selects an alternative model for the patient, accordingly, based on new patient feedback.
  • Block S 232 of the second method 5200 of the method recites identifying a relationship between communication behaviors of patients within the subgroup, characteristics of medical symptoms of patients within the subgroup, and a treatment regimen administered to patients within the subgroup.
  • Block S 280 of the second method S 200 further recites generating a health risk model for the subgroup based on the relationship, the health risk model defining a correlation between a change in communication behavior and risk of change in a medical symptom.
  • an individual patient and can be diagnosed or at-risk for a certain condition, and Blocks of the first method can therefore estimate symptom severity, health status, or health risk change] for the patient based on volunteered survey results and/or an applicable health risk model for a corresponding subgroup of patients.
  • the first method S 100 can implement the health risk model generated in the second method S 200 to predict a health risk of an individual patient.
  • Block S 162 of the first method can retrieve the health risk model associated with the subgroup (and defining a correlation between risk of change in a medical symptom and communication behavior for patients within the subgroup), and Block S 172 can predict a risk of change in a medical symptom for the patient based on the log of use of the native communication application and the health risk model.
  • Blocks of the methods can therefore extract patient symptom severity or health status over time from corresponding survey responses collected in corresponding Blocks of the methods (e.g., Blocks S 112 , S 122 , etc.).
  • a survey delivered to a patient can include an explicit inquiry into if and when a symptom occurred and a severity of symptom presentation, and Block S 114 can analyze the patient's response to the survey to determine if and/or to what extent the patient symptoms presented for the corresponding time period. Block S 114 can subsequently pass the degree of symptom severity to Block S 116 .
  • Block S 124 , Block S 230 , etc. can implement similar functionality to deliver and assess patient surveys.
  • Block S 116 (and Block S 126 , etc.) can subsequently define a relationship between a degree of symptom severity or health status and a quality and/or quantity of communication (i.e., communication behavior, location behavior, and/or phone usage behavior) of the patient for a corresponding period of time.
  • Block S 116 can then pass this relationship to Block S 172 to predict symptom severity or health status for the patient and/or for other patients in a corresponding subgroup.
  • Block S 116 characterizes patient-provided symptom severity and health status data from the first survey and associates this characterization with the communication behavior characterization from Block S 110 .
  • the first method S 100 can thus anticipate the patient's symptom severity and health status at the later time to mimic the patient's treatment behavior during a corresponding time period.
  • Block S 116 correlates a frequency of a first set of outgoing voice communications from a phone call application during the first period with high symptom severity for the patient
  • Block S 126 correlates a frequency of a second set of outgoing voice communications within the phone call application with low symptom severity for the patient, wherein the frequency of the first set of outgoing voice communications is greater than the frequency of the second set of outgoing voice communications.
  • Block S 116 can additionally or alternatively correlate patient symptom severity or health status and/or patient communication behavior to a change in healthcare utilization, such as readmission or increased hospital visits.
  • Block S 116 extracts a patient symptom severity from the first survey and presentation of patient symptoms (e.g., how well the patient is feeling) with patient communication behavior (e.g., how, when, and who the patient contacts during a period of time).
  • Block S 116 can determine that a period during which the patient sends a low volume of text messages to a large breadth of contact types and engages in long phone conversations with a limited number of contacts corresponds to low symptom severity
  • Block S 126 can determine that a period during which the patient sends a low volume of text messages to a low number of unique contacts and engages in short phone conversations with a single contact corresponds to a high symptom severity.
  • Block S 116 can also correlate low symptom severity with a certain amount or pattern of healthcare utilization during the first period (e.g., the patient did not have to visit the doctor for anything other than regular checkups), and Block S 126 can further correlate high symptom severity with increased healthcare utilization during the second period (e.g., the patient had to visit the doctor several times beyond the regular visits).
  • the first method S 100 can repeat elements of Blocks S 110 , S 112 , S 114 , S 116 , etc. to generate a set of communication behavior-symptom severity (and health status) features for the patient over time, and these features can be fed into Block S 130 , Block S 160 , Block S 162 , Block S 144 , Block S 150 , Block S 172 , and/or Block S 180 , etc. to enable further functionality.
  • Block S 232 of the second method S 200 which recites identifying a relationship between communication behaviors of patients within the subgroup, characteristics of medical symptoms of patients within the subgroup, and a treatment regimen administered to patients within the subgroup, can implement similar functionality to generate communication behavior-symptom severity (and health status) features within the subgroup of patients.
  • Block S 232 can pair anonymized communication data from various patients exhibiting similar behaviors over time, such as while presenting low or high symptom severities, with survey responses from corresponding patients at corresponding times to output a collection of features specific to the subgroup. Block S 232 can thus pass this collection of features to Block S 280
  • Block S 280 applies the collection of features to identify trends or patterns between communication behavior, symptom severity, and healthcare utilization across the population and generates a corresponding intervention plan for the subgroup, as described below.
  • Blocks of the first and second methods S 100 , S 200 can implement the PME to generate patient- and/or subgroup-specific models.
  • Block S which recites generating a health risk model including the first log (e.g., frequency) of use of the native communication application, the second log of use of the native communication application, the first symptom severity or health status, and the second symptom severity or health status, can implement the PME to generate a model linking patient communication behavior and health risk (e.g., symptom severity, risk of relapse, risk of hospital readmission, health status, etc.).
  • Block S 130 can implement the BFE to determine that low communication periods commonly follow periods in which the patient correlates with high symptom severity and that period characterized by relatively high levels of patient activity correspond to low symptom severity.
  • Block S 130 can then implement the PME to generate a corresponding model specific to the patient.
  • Block S 130 can subsequently pass this model to Block S 144 to estimate a subsequent patient health risk at a subsequent time based on patient behavior at the subsequent time.
  • Block S 144 can determine that the patient is at a high risk for increased symptom severity if the patient's behavior (e.g., communication behavior and/or activity level) substantially matches recorded patient behavior associated with that symptom severity (as defined by the health risk model).
  • the patient's behavior e.g., communication behavior and/or activity level
  • Block S 162 which can recite retrieving a health risk model associated with the subgroup, can select a preexisting health risk (or risk identification) model—generated by the PME and associated with a subgroup of patients—based on selection of the subgroup in Block S 160 .
  • Block S 144 can similarly function to pass subsequent patient behavior data into the health risk model to predict a subsequent level of patient risk for presenting symptom greater than a threshold severity.
  • Block S 172 can additionally or alternatively predict short-term and/or long-term patient risk for high symptom severity or change in health status or disease progression, such as based on a change in communication behavior of the patient over a certain time period.
  • Block S 172 can predict that the patient will exhibit a similar decrease in symptom severity.
  • Block S 130 and Block S 160 can function in any other way to generate and retrieve a predictive model defining a relationship between symptom severity or health status and patient communication behavior over time, respectively.
  • Block S 172 can further estimate an efficacy of the patient intervention (e.g., nurse outreach, doctor-patient communication, automated health tips delivered through the patient's mobile computing device) in improving the health condition of the patient according to a comparison between changes in symptom severity or health status of the patient.
  • Block S 172 can therefore implement elements of the BFE and/or the PME to predict the outcome of the intervention for the patient. For example, Block S 172 can identify a pattern of communication behavior of the patient, correlate the pattern of communication behavior with intervention outcomes, and then compare the interventions (e.g., based on symptom presentation recorded in various survey responses entered by the patient) to changes in healthcare utilization.
  • Block S 172 can estimate a low efficacy of the intervention in treating the patient's condition. Furthermore, in this example, Block S 172 can compare the patient to existing health condition data and/or a health risk model for a subgroup of patients to determine if worsening symptoms despite delivering interventions is common or expected for the patient's diagnosed health condition and/or for similar patients (i.e., patients within the subgroup). Block S 172 can thus predict the efficacy of the interventions accordingly.
  • Block S 260 of the second method 5200 similarly recites estimating an efficacy of the interventions in improving the health condition for patients within the subgroup based on of the delivery of interventions to patients within the subgroup.
  • Block S 260 implements techniques similar to Block S 172 to apply elements of the BFE and/or the PME to predict the outcome of the interventions for the subgroup of patients within a patient population.
  • Block S 260 can identify an increase in use of native communication applications by patients from prior to initiation of the interventions to the patients to during and after delivery of the interventions to the patients, correlating the increase in use of the native communication applications by the patients with improved health status of the patients, and correlating improved health status of the patients with efficacy of the interventions in improving the health condition of patients within the subgroups.
  • Block S 260 can apply survey response and communication behavior of patients within the subgroup—and a known health condition common to patients within the subgroup—to predict the efficacy of a nurse outreach program in providing support for mental health issues of patients within the subgroup. Additionally or alternatively, Block S 260 can implement such patient data to estimate the efficacy of a health tips program in improving health-related behaviors of diabetic patients within the subgroup.
  • block S 172 and Block S 260 output quantitative values of intervention efficacy for the patient and the patient subgroup, respectively.
  • Block S 172 can output a predicted intervention efficacy of 80%, indicating that four out of five similar patients show a reduction in symptom severity or an improvement in health status.
  • Block S 172 can additionally output a confidence interval indicating a statistical confidence in the intervention regimen, which can be dependent on a size of a patient subgroup associated with an intervention efficacy model used to estimate treatment efficacy for the patient or dependent on a predicted accuracy and/or availability of patient communication and survey response data, as shown in FIGS. 1A and 1C .
  • Block S 172 and Block S 260 output binary indicators of intervention efficacy. For example, if predicted intervention efficacy falls below a threshold, such as below a 70% predicted improvement rate, Block S 172 and/or Block S 260 can output a negative indicator of intervention efficacy. In response to a negative predicted intervention efficacy, Block S 172 can prompt Block S 180 to automatically trigger a notification to care providers suggesting that a different intervention plan might better serve the patient. If data regarding past intervention successes is available modification of the treatment regimen or prompt a care provider to modify the existing treatment regimen or to prescribe a new treatment regimen for the patient, as described below. Similarly, in response to a negative predicted treatment efficacy, Block S 260 can prompt Block S 284 to implement similar functionality to automatically modify or prompt manual modification of the treatment regimen for the subgroup of patients.
  • Block S 180 which recites generating an updated intervention regimen according to the efficacy of the intervention regimen.
  • Block S 180 functions to modify an intervention plan based on an intervention efficacy prediction output in Block S 172 .
  • Block S 180 can prompt a doctor, pharmacologist, etc. to prescribe an alternative intervention program.
  • Block S 284 of one variation of the second method S 200 similarly recites updating the intervention regimen for patients within the subgroup and currently prescribed the intervention regimen in accordance with the intervention efficacy predicted in Block S 260 .
  • Block S 284 the second method S 200 characterizes communication behavior of a patient following diagnosis of a medical condition and before administration of an intervention program, applies the characterized communication behavior to select a subgroup of patients with similar diagnoses and communication behaviors prior to administration of intervention programs, and selects an intervention program for the patient based on a predicted efficacy of the intervention for the patient, which is informed by an actual efficacy of the intervention within the selected subgroup (and/or a corresponding intervention efficacy model for the subgroup, as described above).
  • Block S 284 can function to pair a new patient to a subgroup of current and/or previous patients and to automatically prescribe an intervention regimen to the new patient accordingly.
  • Blocks of the first and second methods S 100 , S 200 can implement the PME to generate patient- and/or subgroup-specific models.
  • Block S 130 which recites generating a patient regimen adherence model including the first log of use of the native communication application, the second log of use of the native communication application, the first adherence, and the second adherence, can implement the PME to generate a model linking patient communication behavior and treatment adherence.
  • Block S 130 can implement the BFE to determine that low communication periods commonly follow periods in which the patient diverts from a prescribed treatment regimen and that period characterized by relatively high levels of patient activity correspond to suitable adherence to the treatment regimen.
  • Block S 130 can then implement the PME to generate a corresponding model specific to the patient.
  • Block S 130 can subsequently pass this model to Block S 144 to estimate a subsequent patient treatment adherence at a subsequent time based on patient behavior at the subsequent time.
  • Block S 144 can determine that the patient has effectively complied with his prescribed treatment program if the patient's behavior (e.g., communication behavior and/or activity level) substantially matches recorded patient behavior associated with treatment adherence compliance (as defined by the patient treatment adherence model), and Block S 144 can determine that patient has not suitably complied with his prescribed treatment program when the patient's behavior substantially matches recorded patient behavior associated with treatment adherence neglect.
  • the patient's behavior e.g., communication behavior and/or activity level
  • Block S 144 which recites estimating a third adherence within the third time period based on the patient regimen adherence model and the third log of use of the native communication application, functions to pass data from a subsequent third period of time into the adherence model generated in Block S 130 to predict patient adherence to the treatment regimen during the third period of time.
  • Block S 144 can additionally or alternatively estimate a patient response to a most-recent administration of the treatment program, such as based on a patient response component of the treatment adherence model.
  • Block S 162 which recites retrieving a regimen adherence model associated with the subgroup, can select a preexisting treatment adherence model—generated by the PME and associated with a subgroup of patients—based on selection of the subgroup in Block S 160 .
  • Block S 144 can similarly function to pass subsequent patient behavior data into the regimen adherence model to predict a subsequent level of patient adherence to a prescribed treatment.
  • Block S 172 can additionally or alternatively predict short-term and/or long-term patient response to the treatment, such as based on a change in communication behavior of the patient soon after beginning a treatment program.
  • Block S 172 can predict that the patient will exhibit a similar positive response to the treatment regimen. Block S 172 can similarly predict patient satisfaction in the treatment regimen based on data of other patients within the subgroup and the regimen adherence model.
  • Block S 130 and Block S 160 can function in any other way to generate and retrieve a predictive model defining a relationship between treatment regimen adherence and patient communication behavior over time, respectively.
  • Block S 280 recites generating a treatment regimen model for the subgroup, the treatment regimen model defining a correlation between communication behavior, treatment responses, and treatment regimen outcomes for patients within the subgroup.
  • Block S 280 can implement the PME to generate a model correlating patient behavior to patient treatment outcome for a subgroup of patients.
  • Block 242 can implement behavior data of a subsequent patient to match the subsequent patient to the subgroup, and Block S 282 can feed behavior data of the subsequent patient into the corresponding treatment regimen model to generate a predicted treatment regimen outcome for the subsequent patient following administration of the treatment regimen as shown in FIG. 2B .
  • Block S 282 can similarly predict a health condition relapse risk for the subsequent patient based on the treatment regimen model, such as for the treatment regimen model that includes a relapse risk output component.
  • Block S 280 implements the PME to generate a treatment regimen model for the subgroup based on adherence to the treatment regimen, communication behavior, and survey responses of patients within the subgroup throughout a period of time.
  • Block S 280 can implement the PME to generate the treatment regimen model for the subgroup by identifying patterns in communication behavior during administration of the treatment regimen and correlating patterns in communication behavior with treatment responses of patients within the subgroup.
  • Block S 282 can thus feed any one or more of adherence to the treatment regimen, communication behavior, and a survey response of a subsequent patient into the treatment regimen model to output a predicted effect of the associated treatment on the subsequent patient, such as for a particular period of time during administration of the treatment program to the subsequent patient.
  • Block S 280 can function in any other way to generate a predictive model defining a relationship between treatment outcome for a patient and patient communication behavior over time.
  • Block S 172 which recites estimating an efficacy of the treatment regimen in treating the health condition of the patient according to a comparison between the treatment response and the adherence to the treatment regimen by the patient, functions to implements elements of the BFE and/or the PME to predict the outcome of the treatment regimen for the patient.
  • Block S 172 can identify a pattern of communication behavior of the patient, correlate the pattern of communication behavior with a pattern of treatment adherence, and then compare the pattern of treatment adherence to a trend in patient treatment response (e.g., based on symptom presentation recorded in various survey responses entered by the patient). In this example, if the patient exhibits strong treatment adherence but worsening symptoms, Block S 172 can estimate a low efficacy of the treatment in treating the patient's condition.
  • Block S 172 can compare the patient to existing health condition data and/or a treatment outcome model for a subgroup of patients to determine if worsening symptoms despite strong treatment adherence is common or expected for the patient's diagnosed health condition and/or for similar patients (i.e., patients within the subgroup). Block S 172 can thus predict the efficacy of the treatment accordingly.
  • Block S 260 of the second method S 200 similarly recites estimating an efficacy of the treatment regimen in treating the health condition for patients within the subgroup based on adherence to prescribed treatment regimens and health statuses of patients within the subgroup.
  • Block S 260 implements techniques similar to Block S 172 to apply elements of the BFE and/or the PME to predict the outcome of the treatment regimen for the subgroup of patients within a patient population.
  • Block S 260 can identify an increase in use of native communication applications by patients from prior to initiation of the treatment regimen by the patients to during administration of the treatment regimen by the patients, correlating the increase in use of the native communication applications by the patients with improved health status (e.g., increased wellness, reduced symptoms) of the patients, and correlating improved health status of the patients with efficacy of the treatment in treating the health condition of patients within the subgroups.
  • Block S 260 can apply survey response and communication behavior of patients within the subgroup—and a known health condition common to patients within the subgroup—to predict an efficacy of a physical therapy regimen in treating physical handicaps for patients within the subgroup. Additionally or alternatively, Block S 260 can implement such patient data to estimate the efficacy of a pharmacotherapy regimen in treating mental disorders for patients within the subgroup.
  • Block S 172 and Block S 260 output quantitative values—from a scale or spectrum of treatment values—of treatment efficacy for the patient and the patient subgroup, respectively.
  • Block S 172 can output a predicted treatment efficacy of 80%, indicating that four out of five similar patients recover from a diagnosed condition following the corresponding treatment regimen.
  • Block S 172 can additionally output a confidence interval indicating a statistical confidence in the predicted treatment regimen, which can be dependent on a size of a patient subgroup associated with a treatment efficacy model used to estimate treatment efficacy for the patient or dependent on a predicted accuracy and/or availability of patient communication and survey response data, as shown in FIGS. 1A and 1C .
  • Block S 172 and Block S 260 output binary indicators of treatment efficacy. For example, if predicted treatment efficacy falls below a threshold efficacy, such as below a 70% predicted success rate, Block S 172 and/or Block S 260 can output a negative indicator of treatment efficacy. In response to a negative predicted treatment efficacy, Block S 172 can prompt Block S 180 to automatically trigger modification of the treatment regimen or prompt a care provider to modify the existing treatment regimen or to prescribe a new treatment regimen for the patient, as described below. Similarly, in response to a negative predicted treatment efficacy, Block S 260 can prompt Block S 284 to implement similar functionality to automatically modify or prompt manual modification of the treatment regimen for the subgroup of patients.
  • a threshold efficacy such as below a 70% predicted success rate
  • Block S 172 and/or Block S 260 can output a negative indicator of treatment efficacy.
  • Block S 172 can prompt Block S 180 to automatically trigger modification of the treatment regimen or prompt a care provider to modify the existing treatment regimen or to prescribe a new treatment
  • Block S 180 which recites generating an updated treatment regimen according to the efficacy of the treatment regimen.
  • Block S 180 functions to modify a treatment prescription based on a treatment efficacy prediction output in Block S 172 .
  • Block S 180 can respond to such improved symptoms by modifying the prescribed medication dosage specified in the treatment regimen for the patient, as shown in FIG. 1C .
  • Block S 180 can cancel the current treatment program and prescribe an alterative automatically or prompt a doctor, pharmacologist, etc. to prescribe an alternative treatment program.
  • Block S 284 of one variation of the second method S 200 similarly recites updating the treatment regimen for patients within the subgroup and currently prescribed the treatment regimen in accordance with the treatment efficacy predicted in Block S 260 .
  • the second method S 200 characterizes communication behavior of a patient following diagnosis of a medical condition and before administration of a treatment program, applies the characterized communication behavior to select a subgroup of patients with similar diagnoses and communication behaviors prior to administration of treatment programs, and selects a treatment program for the patient based on a predicted efficacy of the treatment for the patient, which is informed by an actual efficacy of the treatment within the selected subgroup (and/or a corresponding treatment efficacy model for the subgroup, as described above).
  • Block S 282 can prescribe a custom medication and custom dosage for the patient based on the predicted treatment regimen outcome for the patient, such as by passing different medications and/or dosages into the subgroup regimen outcome model and selecting a particular medication and/or dosage that yields a highest predicted treatment efficacy. Therefore, Block S 284 can function to pair a new patient to a subgroup of current and/or previous patients and to automatically prescribe a treatment regimen to the new patient accordingly.
  • Block S 150 of the first method S 100 recites presenting a treatment-related notification based on the third adherence through the mobile computing device.
  • Block S 150 functions to generate and present a notification to the patient through the patient's mobile computing device to provide guidance to the patient during the treatment program, such as in the form of a reminder to perform a particular task (shown in FIG. 1A ), suggest or enforce a behavior change, provide positive reinforcement for a behavior, provide behavioral insights, or share a patient symptom and/or behavior timeline.
  • Block S 150 can generate a notification including a reminder to take a prescribed medication and transmit the reminder to the patient's mobile computing device, wherein a native application executing on the mobile computing device displays the notification, such as in the form of a pop-up notification.
  • Block S 150 can transmit the notification to the mobile computing device for display at a prescribed treatment administration time, such as a 9 AM and 5 PM—dosage times prescribed by a doctor—on a day following estimated treatment program neglect by the patient.
  • Block S 150 can transmit the notification to the patient's mobile computing device substantially in real time in response to a patient action, inaction, behavior, or behavior change.
  • Block S 150 can generate a notification reciting “everything all right?” and transmit the notification to the patient in response to a detected abrupt change in a patient behavior pattern (e.g., the patient calls the same family member every morning at 9 AM, but 9:15 AM passes on a present day without an outgoing or incoming phone call).
  • a patient behavior pattern e.g., the patient calls the same family member every morning at 9 AM, but 9:15 AM passes on a present day without an outgoing or incoming phone call.
  • Block S 150 can also incorporate additional patient information in the notification.
  • Block S 150 can generate a notification that specifies a medication dosage for the patient, such as including how many pills of a certain type (or size, shape, color, etc.).
  • the notification can incorporate dosage information based on an updated treatment regimen, such as output in Block S 180 .
  • Block S 150 can additionally or alternatively generate the notification that includes past patient data indicating an expected outcome if the patient adheres to the treatment regimen and/or an expected outcome if the patient neglects the treatment regimen, such as based on a patient or subgroup treatment adherence model.
  • Block S 150 can generate the notification that includes a prompt requesting confirmation of patient implementation of the action item specified in the notification (e.g., the patient took his prescribed medication). For example, the notification can recite “It's time to take your meds” and include input regions reciting “Roger that!” and “Remind me later,” as shown in FIG. 1A . Block S 150 can subsequently pass a “Roger that!” selection back into the treatment adherence model of the patient as adherence to the treatment program, and Block S 150 can pass a “Remind me later” selection into the treatment adherence model as neglect or delay of the treatment program, as shown in FIG. 3 .
  • Block S 150 can additionally or alternatively interface with a care provider, such as a nurse or doctor (e.g., through a care provider interface executing on a corresponding computing device), to generate a custom notification.
  • a care provider such as a nurse or doctor
  • Block S 150 can transmit or share patient information, including health condition, treatment regimen, actual and estimated patient adherence, survey response, and/or actual and estimated treatment response data of the patient, with the care provider and guide the care provider in identifying a patient problem or need.
  • Block S 150 can then receive from the care provider a selection for a default notification from a pre-populated list of available and/or applicable default notifications.
  • Block S 150 can prompt the care provider to enter a custom textual and/or image-based (e.g., infographic) notification for the patient and then pass the notification back to the patient.
  • Block S 172 predicts a risk of change in a medical symptom for the patient based on a risk of increased symptom severity for symptoms associated with a particular disease associated with (e.g., diagnosed in) the selected subgroup, and Block S 150 transmits the notification to the care provider if to the risk of change in the medical symptom for the patient exceeds a threshold risk specific to the particular disease, such as a 70% probability that patient symptom severity will systematically increase in the (near) future for a patient diagnosed with diabetes and an 88% probability that patient symptom severity will systematically increase in the (near) future for a patient diagnosed with ADHD.
  • a threshold risk specific to the particular disease such as a 70% probability that patient symptom severity will systematically increase in the (near) future for a patient diagnosed with diabetes and an 88% probability that patient symptom severity will systematically increase in the
  • Block S 150 can additionally or alternatively generate a notification to prompt or reminder the patient to complete a survey, schedule a visit with a doctor or health care provider, to attend a scheduled health-related visit, etc.
  • Block S 150 can generate a notification including any other information and presented to the patient in any other suitable way or through any other suitable medium.
  • the first and second methods S 100 , S 200 can function to provide one or more services to a care provider associated with a patient and with a subgroup of patients, respectively.
  • the first and second methods S 100 , S 200 support a disease management program by enabling provision of lower cost treatments and/or therapies to patients in exchange for feedback, wherein the first and second methods S 100 , S 200 collect patient feedback as described above.
  • the first method S 100 can support tools and one or more mobile applications (i.e., native applications for mobile computing devices) to enable better patient-driven condition management and to reduce symptom flare-ups
  • the second method S 200 can aggregate and extract valuable information from patient data across a subgroup to support disease management programs by a care provider.
  • the second method S 200 can manage or provide analyzed data to a care provider to manage a population with a chronic condition, such as by collecting data about patients outside a clinical setting and generate reports and/or models to aid triage for higher-risk patients, and the first method S 100 —in conjunction within the second method S 200 —can support a communication venue or network to enable more effective outreach from a care provider to a patient.
  • the first method S 100 accesses a second log of use of a native communication application executing on a second mobile computing device by a second patient and predicts a second risk of change in a medical symptom for the second patient based on the second log of use of the native communication application and the health risk model, and Block S 150 transmits a scored patient triage list to a care provider (e.g., nurse, hospital, clinic, etc.) according to the risk of change in the medical symptom for the patient and the second risk of change in the medical symptom for the second patient.
  • a care provider e.g., nurse, hospital, clinic, etc.
  • Block S 150 transmits a notification to the care provider comprises prompting the care provider to schedule a visit with a particular patient.
  • the first and second methods S 100 , S 200 can therefore target patient interventions to improve patient health and reduce healthcare utilization (e.g., care costs, emergency room visits, pharmaceutical reliance, etc.).
  • the first and second methods S 100 , S 200 can generate and implement models to identify patients at higher-risk for declining health, increased healthcare utilization, and higher hospital readmissions. Once such patients are identified, the first and second methods S 100 , S 200 can prompt a care provider to intervene, thereby initiating targeted care interventions to improve patient health and wellbeing.
  • the first and second methods S 100 , S 200 can additionally or alternatively extrapolate insights from population-level data to guide care provider improvement in practices, thereby guiding care providers to better patient outcomes.
  • the second method S 200 can detect clusters of patients within a population who demonstrate similar behavioral and outcome patterns, identify two clusters of patients who show significantly different readmission rates, identify different treatment practices across the two clusters, and thus guide a care provider in designing a treatment practice based on the readmission rates and the treatment practices across with the two clusters of patients.
  • Block S 172 can communicate a predicted (e.g., estimated) efficacy of the treatment regimen for the patient to a healthcare provider associated with the patient. For example, if the predicted efficacy of the treatment falls below a threshold efficacy for the health condition for the patient, Block S 172 can transmit an alert or other notification to a doctor, nurse, or other care provider noted in the patient's medical file or recorded as managing treatment of the particular health condition for the patient. In one implementation, Block S 172 transmits an alert to a corresponding care provider through a private care provider account accessible through a provider-specific interface (e.g., an online healthcare dashboard), such as a care application logged-in to a care provider's private account and executing on a computing device associated with the care provider.
  • a provider-specific interface e.g., an online healthcare dashboard
  • the first method S 100 can also enable the care provider to access additional patient information through the care application and/or private care account, such as through an online dashboard.
  • Block S 172 can also guide the care provider in modifying an existing treatment program or selecting an alternative treatment program for the patient, as described above, such as by displaying a list of available treatment programs for the patient's health condition in a scored list based on similarities between the patient's behaviors and behavior within a subgroup of patients and treatment efficacies within the subgroup (e.g., via a treatment efficacy model).
  • Block S 172 can function in any other way to communicate a predicted efficacy of the treatment regimen for the patient to a corresponding care provider.
  • Block S 290 of the second method which recites transmitting the predicted treatment regimen outcome following administration of the treatment regimen by the subsequent patient to a care provider specified in a digital health profile associated with the subsequent patient, can implement similar techniques and/or methods to share health- and treatment-related information to a care provider related to the (subsequent) patient.
  • Block S 290 can transmit a notification including a prompt to respond to a current health status of the subsequent patient to a corresponding care provider.
  • Block S 290 can function in any other way to communicate a predicted treatment outcome for a patient to a care provider.
  • Block S 270 of the second method which recites generating a treatment regimen report specific to the subgroup based on the efficacy of the treatment regimen, can implement similar techniques or functionality to communicate a subgroup-wide efficacy report for a particular treatment regimen to a care provider.
  • Block S 270 functions to formulate a report for the efficacy of the treatment regimen with the population based on the estimated efficacy of the treatment regimen output in Block S 260 .
  • Block S 270 can generate the report that includes charts relating patterns in patent symptoms to treatment patterns in treatment adherence, changes in patterns in patient behaviors and survey responses, etc. for within the subgroup over time.
  • Block S 270 generates a treatment report specific to a particular patient within the subgroup, including a predicted risk of health condition decline for the particular patient during administration of the treatment regimen, as shown in FIG. 2A .
  • Block S 270 can thus prompt a care provider to modify a treatment program for a particular patient within a subgroup, such as based on predicted or extrapolated patient risk.
  • Block S 270 can also guide a care provider in modifying or changing a treatment regimen for the subgroup, such as by identifying and displaying (unexpected) patterns of treatment negligence or patient ‘customization’ associated with improved patient symptoms and/or patient behaviors within the subgroup.
  • the second method S 200 can implement behavior data of patients within the subgroup to compare the wellbeing of the patients and identify those most in need accordingly.
  • Block S 270 can thus generate the treatment regimen report that includes a ranking of patients within the subgroup based on need and symptoms.
  • a nurse practitioner managing dozens of patients can review the treatment report for the subgroup to more effectively triage and determine the patients that are (likely) in need of support.
  • Block S 270 and the second method S 200 in general—can therefore enable a nurse or other care provider to make real-time decisions and intervene and/or reach out to a particular patient based on real patient data.
  • Block S 270 can generate a report that includes any other suitable patient information and can function in any other way to make the report available to one or more care providers, such as related to one or more patients and/or a prescribed treatment regimen for the subgroup.
  • the first and second methods S 100 , S 200 can implement similar functionality and techniques to identify patient risk and assess treatment efficacy.
  • the systems and methods of the embodiments can be embodied and/or implemented at least in part as a machine configured to receive a computer-readable medium storing computer-readable instructions.
  • the instructions can be executed by computer-executable components integrated with the application, applet, host, server, network, website, communication service, communication interface, hardware/firmware/software elements of a patient computer or mobile device, or any suitable combination thereof.
  • Other systems and methods of the embodiments can be embodied and/or implemented at least in part as a machine configured to receive a computer-readable medium storing computer-readable instructions.
  • the instructions can be executed by computer-executable components integrated by computer-executable components integrated with apparatuses and networks of the type described above.
  • the computer-readable medium can be stored on any suitable computer readable media such as RAMs, ROMs, flash memory, EEPROMs, optical devices (CD or DVD), hard drives, floppy drives, or any suitable device.
  • the computer-executable component can be a processor, though any suitable dedicated hardware device can (alternatively or additionally) execute the instructions.

Abstract

One method for supporting a patient through a treatment regimen includes: accessing a log of use of a native communication application executing on a mobile computing device by a patient; selecting a subgroup of a patient population based on the log of use of the native communication application and a communication behavior common to the subgroup; retrieving a regimen adherence model associated with the subgroup, the regimen adherence model defining a correlation between treatment regimen adherence and communication behavior for patients within the subgroup; predicting patient adherence to the treatment regimen based on the log of use of the native communication application and the regimen adherence model; and presenting a treatment-related notification based on the patient adherence through the mobile computing device.

Description

    CROSS-REFERENCE TO RELATED APPLICATIONS
  • This application is a continuation of U.S. patent application Ser. No. 16/057,199, filed 7 Aug. 2018, which is a continuation of U.S. patent application Ser. No. 15/807,323, filed 8 Nov. 2017, which is a continuation of U.S. patent application Ser. No. 13/969,349, filed 16 Aug. 2013, which claims the benefit of U.S. Provisional Application No. 61/683,867, filed 16 Aug. 2012, and U.S. Provisional Application No. 61/683,869, filed 16 Aug. 2012, which are all incorporated in their entireties herein by this reference.
  • TECHNICAL FIELD
  • This invention relates generally to the field of patient health and more specifically to a new and useful method for modeling behavior and health changes in the field of patient health.
  • BRIEF DESCRIPTION OF THE FIGURES
  • FIG. 1A is a flowchart representation of a first method of the invention;
  • FIG. 1B is a flowchart representation of a variation of the first method;
  • FIG. 1C is a flowchart representation of a variation of the first method;
  • FIG. 1D is a flowchart representation of a variation of the first method;
  • FIG. 2A is a flowchart representation of a second method of the invention;
  • FIG. 2B is a flowchart representation of a variation of the second method;
  • FIG. 2C is a flowchart representation of a variation of the second method;
  • FIG. 2D is a flowchart representation of a variation of the second method;
  • FIG. 3 is a flowchart representation of variations in accordance with the first and second methods;
  • FIG. 4 is a graphical representation of variations in accordance with the first and second methods; and
  • FIG. 5 is a graphical representation of variations in accordance with the first and second methods.
  • DESCRIPTION OF THE EMBODIMENTS
  • The following description of the embodiments of the invention is not intended to limit the invention to these embodiments, but rather to enable any person skilled in the art to make and use this invention.
  • 1. Methods
  • As shown in FIG. 1A, a first method S100 for supporting a patient associated with a health condition through a treatment regimen includes: accessing a first log of use of a native communication application executing on a mobile computing device by the patient within a first time period in Block S110; receiving a first survey response corresponding to the first time period from the patient in Block S112; estimating a first adherence to the treatment regimen by the patient within the first time period based on the first survey response in Block S114; correlating the first log of use of the native communication application with the first adherence to the treatment regimen in Block S116; accessing a second log of use of the native communication application by the patient within a second time period in Block S120; receiving a second survey response from the patient within the second time period in Block S122; estimating a second adherence of the patient within the second time period based on the second survey response in Block S124; correlating the second log of use of the native communication application with the second adherence to the treatment regimen in Block S126; generating a patient regimen adherence model including the first log of use of the native communication application, the second log of use of the native communication application, the first adherence, and the second adherence in Block S130; accessing a third log of use of the native communication application by the patient within a third time period in Block S140; estimating a third adherence to the treatment regimen within the third time period based on the patient regimen adherence model and the third log of use of the native communication application in Block S144; and presenting a treatment-related notification based on the third adherence through the mobile computing device in Block S150.
  • As shown in FIG. 1B, one variation of the first method S100 includes: accessing a log of use of a native communication application executing on a mobile computing device by a patient in Block S110; selecting a subgroup of a patient population based on the log of use of the native communication application and a communication behavior common to the subgroup in Block S160; retrieving a regimen adherence model associated with the subgroup, the regimen adherence model defining a correlation between treatment regimen adherence and communication behavior for patients within the subgroup in Block S162; predicting patient adherence to the treatment regimen based on the log of use of the native communication application and the regimen adherence model in Block S144; and presenting a treatment-related notification based on the patient adherence through the mobile computing device in Block S150.
  • As shown in FIG. 1C, another variation of the first method includes: accessing a log of use of a native communication application executing on a mobile computing device by a patient within a period of time in Block S110; selecting a subgroup of a patient population based on the log of use of the native communication application and a communication behavior common to the subgroup in Block S160; retrieving a regimen adherence model associated with the subgroup, the regimen adherence model defining a correlation between treatment regimen adherence and communication behavior for patients within the subgroup in Block S162; predicting adherence to the treatment regimen by the patient based on the log of use of the native communication application and the regimen adherence model in Block S144; extracting a treatment response of the patient from a patient survey corresponding to the period of time in Block S112; estimating an efficacy of the treatment regimen in treating a health condition of the patient according to a comparison between the treatment response and the adherence to the treatment regimen by the patient in Block S172; transmitting a notification to a care provider associated with the patient in response to the efficacy of the treatment regimen that falls below a threshold efficacy in Block S150.
  • As shown in FIG. 1D, yet another variation of the first method includes: accessing a log of use of a native communication application executing on a mobile computing device by the patient in Block S110; selecting a subgroup of a patient population based on the log of use of the native communication application and a communication behavior common to the subgroup in Block S160; retrieving a health risk model associated with the subgroup, the health risk model defining a correlation between risk of change in a medical symptom and communication behavior for patients within the subgroup in Block S162; predicting a risk of change in a medical symptom for the patient based on the log of use of the native communication application and the health risk model in Block S172; and transmitting a notification to a care provider associated with the patient in response to the risk of change in the medical symptom for the patient that exceeds a threshold risk in Block S150.
  • As shown in FIG. 2A, a second method S200 includes: selecting a subgroup of patients from a patient population in Block S220, patients within the subgroup exhibiting similar behavioral characteristics and associated with a health condition; for a patient within the subgroup, estimating adherence of the patient to a prescribed treatment regimen during a period of time based on survey responses entered by the patient through a corresponding mobile computing device in Block S230; for a patient within the subgroup, characterizing communication behavior of the patient based on use of a native communication application executing on a corresponding mobile computing device by the patient during the period of time in Block S240; for a patient within the subgroup, correlating communication behavior of the patient with a health status of the patient in Block S250; estimating an efficacy of the treatment regimen in treating the health condition for patients within the subgroup based on adherence to prescribed treatment regimens and health statuses of patients within the subgroup in Block S260; and generating a treatment regimen report specific to the subgroup based on the efficacy of the treatment regimen in Block S270.
  • As shown in FIG. 2B, one variation of the method includes: selecting a population of patients prescribed a treatment regimen for a health condition in Block S210; for a patient within the population, characterizing communication behavior of the patient based on use of a native communication application executing on a corresponding mobile computing device by the patient prior to initiation of a treatment regimen and during administration of the treatment regimen to the patient in Block S240; selecting a subgroup of patients from the population of patients based on similar communication behaviors prior to initiation of the treatment regimen in Block S220; extracting treatment responses from surveys completed by patients within the subgroup during administration of the treatment regimen in Block S230; generating a treatment regimen model for the subgroup in Block S280, the treatment regimen model defining a correlation between communication behavior, treatment responses, and treatment regimen outcomes for patients within the subgroup; characterizing communication behavior of a subsequent patient based on use of a native communication application executing on a corresponding mobile computing device by the subsequent patient in Block S242; and generating a predicted treatment regimen outcome for the subsequent patient based on a similarity between communication behavior of the subsequent patient and communication behavior common within the subgroup in Block S282.
  • As shown in FIG. 2C, another variation of the second method includes: selecting a subgroup of patients associated with a health condition from a population of patients, patients in the subgroup exhibiting similar behavioral characteristics in Block S220; for patients within the subgroup, characterizing communication behavior of a patient based on use of a native communication application executing on a corresponding mobile computing device by the patient during the period of time in Block S240; extracting characteristics of medical symptoms of patients within the subgroup from surveys submitted by patients within the subgroup in Block S230; identifying a relationship between communication behaviors of patients within the subgroup, characteristics of medical symptoms of patients within the subgroup, and a treatment regimen administered to patients within the subgroup in Block S232; generating a treatment efficacy model for the subgroup based on the relationship, the treatment efficacy model defining a correlation between a change in communication behavior and efficacy of the treatment regimen in alleviating medical symptoms of patients within the subgroup in Block S280.
  • As shown in FIG. 2D, yet another variation of the second method includes: selecting a subgroup of patients associated with a health condition from a population of patients, patients in the subgroup exhibiting similar behavioral characteristics in Block S220; for patients within the subgroup, characterizing communication behavior of a patient based on use of a native communication application executing on a corresponding mobile computing device by the patient during the period of time in Block S240; extracting characteristics of medical symptoms of patients within the subgroup from surveys submitted by patients within the subgroup in Block S230; identifying a relationship between communication behaviors of patients within the subgroup and characteristics of medical symptoms of patients within the subgroup in Block S232; generating a health risk model for the subgroup based on the relationship, the health risk model defining a correlation between a change in communication behavior and risk of change in a medical symptom in Block S280.
  • 2. Applications of the Methods
  • Generally, the first and second methods S100, S200 function to collect communication data of a patient (a user or an at-risk individual) from a mobile computing device associated with the patient and to anticipate a health status of the patient based on the patient's communication data. The methods can subsequently apply the anticipated health status of the patient to suggest an action to the patient and/or to inform a nurse, care provider, a family member, a pharmacist, a pharmacologist, an insurance producer, a hospital, or other health professional or network, etc. with anticipated health-related information of the patient. The methods can additionally or alternatively implement an anticipated patient health status to drive automated or manual targeted intervention of for a patient via a phone call, email, health tip notification, insight, or other electronic communication. The methods can also apply an anticipated health status of a patient to remind the patient to fulfill a treatment, to model the patient's progress through a treatment program or regimen, to predict an outcome of the patient's treatment program, to modify or customize a treatment program for the patient, to generate a model of treatment regimens and/or treatment regimen outcomes for a group of similar patients within a population of patients, to predict or monitor a change in patient symptoms or health status change, etc., such as based on volunteered survey data and/or volunteered or predicted treatment regimen adherence by the patient and/or other similar patients.
  • In one example, over time, the methods collect data related to the patient's daily phone calls, text messages, and emails, including frequency, duration, length of time to respond to inbound communications, time-related communication patterns, and/or diversity of contacts through one or more native applications executing on a smartphone (or tablet) associated with the patient (e.g., a phone call application, a text messaging application, and an email application). Initially, the methods can survey the patient for his health risk assessment, symptom score, and/or degree of current symptoms (e.g., how the patient is feeling) and then correlate the patient's health status, symptom score etc. to generate a health risk model. For example, the methods can generate a quantitative risk score corresponding to a predicted level of risk of change in a medical symptom for a patient based on a correlation between patient behavior and a corresponding health risk model.
  • The methods can subsequently implement the health risk model to predict a level of the patient's symptoms at a later time and/or if a notification pertaining to the patient's health should be generated based (solely or predominantly) on the patient's communication behavior through a computing device. As in this example, the first method S100 can extrapolate a patient health status from patient communication data and survey responses and combine patient health status with patient communication behavior to generate a health risk model, and the first method can later anticipate a change in patient health risk by feeding subsequent patient communication data into the health risk model to determine subsequent patient health risk and generate a notification for the patient accordingly.
  • In another example, over time, the methods collect data related to the patient's daily phone calls, text messages, emails, and/or other inbound and/or outbound communications from the patient's mobile computing device, including frequency, duration, length of time to respond to an inbound communication, time-related communication patterns, diversity of contacts, etc. through one or more native communication applications executing on a smartphone (or tablet) associated with the patient (e.g., a phone call application, a text messaging application, and an email application). The methods can thus aggregate and manipulate any of the foregoing data to estimate patient adherence to a treatment regimen (e.g., whether the patient took his medications) and how the patient is ‘feeling’ (i.e., degree of symptoms) and subsequently correlate the patient's treatment adherence and feelings to the patient's communication behavior to generate a regimen adherence model. The methods can subsequently predict how the patient is feeling and/or if the patient followed his treatment regimen based on the patient's communication behavior through his smartphone. As in this example, the first method S100 can extrapolate patient treatment regimen adherence from a survey response, combine patient treatment regimen adherence with patient communication behavior to generate a regimen adherence model, and later anticipate a patient regimen adherence by feeding subsequent communication behavior into the regimen adherence model and then generate a notification for the patient accordingly.
  • In a similar example, the methods apply behavior data of the patient, such as prior to and/or during a treatment program or regimen, to select a regimen adherence model associated with a subgroup of patients exhibiting communication behaviors similar to those of a particular patient. By feeding patient communication behavior data into a health risk model, the methods can thus predict how the patient is feeling (i.e., presentation of the patient's symptoms) without additional patient info (e.g., additional survey results). As in this example, the methods can thus combine survey responses and patient communication behaviors to generate patient population subgroup models and later anticipate a patient health risk—and generate a notification accordingly—by feeding subsequent communication behavior of a patient into the model(s).
  • In yet another example, the first method S100 applies behavior data of the patient, such as prior to and/or during a treatment program or regimen, to select a regimen adherence model associated with the subgroup exhibiting communication behaviors similar to those of the patient. By feeding subsequent patient communication behavior into the regimen adherence model, the first method S100 can thus predict how the patient is feeling (i.e., presentation of the patient's symptoms) at a particular time and/or anticipate how or when the patient is fulfilling his treatment regimen, such as by consuming the prescribed medications at prescribed times and at prescribed dosages, without additional patient info (e.g., additional survey results). As in this example, this variation of the first method S100 can thus extrapolate patient treatment regimen adherence from a survey response, combine patient treatment regimen adherence with patient communication behavior to select a patient population subgroup and a related regimen adherence model, and later anticipate a patient regimen adherence by feeding subsequent communication behavior into the regimen adherence model and then generate a notification for the patient accordingly.
  • In still another example, the methods implement behavior data of a patient to select a health risk model associated with a particular subgroup exhibiting communication behaviors similar to those of the patient. By feeding subsequent patient communication behavior data into the health risk model, the methods can this predict how the patient is feeling (i.e., presentation of the patient's symptoms) at a particular time and/or anticipate how or when the a notifications indicating patient risk should be transmitted to a care provider, all without necessitating additional patient information (e.g. additional survey results).
  • In a further example, the first method S100 applies behavior data of the patient, such as prior to and/or during a treatment program or regimen, to select a regimen adherence model associated with the subgroup exhibiting communication behaviors similar to those of the patient. By feeding subsequent patient communication behavior into the regimen adherence model, the first method S100 can thus predict how the patient is feeling (i.e., presentation of the patient's symptoms) at a particular time and/or anticipate how or when the patient is fulfilling his treatment regimen, such as by consuming the prescribed medications at prescribed times and at prescribed dosages, without additional patient info (e.g., additional survey results).
  • In another example, the second method S200 collects communication data from multiple patients, groups patients according to identified communication behaviors, and anticipates progress through a treatment program and the effects of the treatment program on patients with a selected group. Rather than relying on data entered manually by patients and/or care providers, estimating patient activity or action from motion data (which can be imprecise, require substantial computing powers, and lack suitable precision for the health care space), etc., the second method S200 can instead generate models linking treatments (e.g., pharmacotherapy regimens, physical therapy programs) to symptom presentation and final health condition outcome for patients based on how, when, and/or with whom, etc. patients communicate over time (e.g., on a daily or four-hour basis). In this example, the second method S200 can thus select a subgroup of patients within a patient population based on communication behaviors of the patient population, estimate treatment regimen adherence within the subgroup based on survey responses volunteered by patients within the subgroup, estimate the health status of patients within the subgroup based on communication behaviors of the patients, and combine treatment adherence and communication behavior of patients within the subgroup to estimate the efficacy of the treatment regimen specifically for the subgroup of patients.
  • In the foregoing example, the second method S200 can further apply such models to particular patients to anticipate how a patient will respond to a treatment regimen and/or to customize a treatment regimen for the patient based on similarities and/or differences between the patient's communication behaviors and communication behavior of other patients (e.g., previous or current patients with similar diagnosed health conditions and/or prescribed treatment regimens).
  • The first and second methods S100, S200 can therefore derive meaningful health-related data for individual patients and/or groups of patients from patient communication behaviors extrapolated from phone call, text message, email, and/or other communication data collected passively through mobile computing devices associated with the patient(s). Such derived health-related data can be implemented within a patient's mobile computing device to privately guide a patient through a treatment program by a care provider to anticipate risk of increasing medical symptoms or risk of a change in health status of the patient, by a doctor to monitor a patient's progress or to modify a prescription, by a nurse to identify a need to provide manual support to the patient, and/or by a pharmacologist to research drug therapies and drug responses within a population, etc. For example, Blocks of the first and second methods S100, S200 can be implemented to collect patient data for an experiment, a research study, a commercial launch, a marketing study, a patient community study, such as on the recommendation of a healthcare provider (e.g., a doctor or a medical institution (e.g., a hospital center) to improve patient care. In this example, the first and second methods S100, S200 can interface with a patient-facing interface, a doctor-facing interface, a nurse-facing interface, and/or a pharmacologist-facing interface, etc. to deliver notifications and related data to the patient, doctor, nurse, pharmacologist, and/or other care provider directly or indirectly associated with the patient. In another example, the first and second method S100, S200 can be implemented on a computing device associated with a particular patient as a personal heath-tracking tool.
  • In one implementation, the first and second methods S100, S200 interface with a native data collection application executing on a patient's mobile computing device (e.g., smartphone, tablet, personal data assistant (PDA), personal music player, vehicle, etc.) to retrieve patient communication data. For example, the native data collection application can be installed on the patient's mobile computing device, execute substantially continuously while the mobile computing device is in use and/or “ON,” record times, durations, and contact types (e.g., family, friend, coworker, business associate) of each inbound and outbound communication from the patient's mobile computing device. The mobile computing device can then upload this data to a remote database, such as in real-time, every hour, at the end of each day, etc. over an Internet connection, and the first and/or second method, implemented on a computer network (e.g., the “cloud”), can retrieve the patient's communication data from the remote database, analyze the patient's communication data to anticipate the patient's symptoms and/or therapy adherence, and generate a notification for the patient, and the patient's mobile computing device can download and subsequently display the notification for the patient, actions that can be handled automatically and in the background by the native data collection application or an alternative patient-facing native application executing on the mobile computing device. The computer network can additionally or alternatively generate a patient regimen adherence model, a patient outcome model, a patient behavioral model, a health risk model etc. and transmit any one or more models to a doctor, a nurse, a pharmacologist, a therapist, etc.
  • Therefore, Blocks of the first and second methods S100, S200 can be implemented on one or more computer systems, such as a cloud-based computer system (e.g., Amazon EC3), a mainframe computer system, a grid-computer system, or any other suitable computer system. Blocks of the first and second methods S100, S200 can collect patient data from one or more devices over the Internet, such as communication data directly from a natively application executing on the patient's smartphone and/or motion data from a wearable sensor collected by the patient's smartphone and then uploaded to a remote database over an Internet connection. Collection, manipulation, and transmission of patient data in Blocks of the first and second methods S100, S200 can further adhere to health-related privacy laws, such as by privatizing or anonymizing patient data and transmitting encrypted or private notifications—such as pertaining to a patient's therapy adherence—to the patient and/or a doctor, a nurse, a pharmacologist, a researcher, etc. associated with the patient.
  • As in the foregoing implementation, when a patient installs and/or authorizes collection and transmission of personal communication data through such a native data collection application, the native application can prompt the patient to create a profile or account. The account can be stored locally on the patient's mobile computing device and/or remotely on the computer network and can contain a name, age, gender, location, occupation, list of health conditions, list of current health-related treatments and medications, medical history, a primary care physician (including name, office, contact information, etc.), health insurance information, prescription insurance plan, local pharmacy information, or other demographic and/or health-related information, any of which can be added by the patient, a family member, a doctor, a nurse, or other individual associated with the patient and/or retrieved from a medical, insurance, social network, or other related database. The first and second methods S100, S200 can also update the patient's account with additional patient information over time, such as presentation of symptoms, estimated patient treatment adherence, and/or predicted treatment regimen outcome for the patient.
  • Blocks of the methods can function to qualify and/or quantify a correlation between a health outcome or symptom risk and observed behavior changes for a patient based on data gathered passively from a patient's mobile computing device (e.g., mobile phone, smartphone and/or tablet, etc.) substantially without manual data entry from the patient, a doctor, a nurse, a pharmacist, etc. Blocks of the methods can similarly quantify a correlation between a patient health status and an observed patient behavior gathered from the patient's mobile computing device and further combine passive behavior data collection with self-reported patient data (i.e., survey responses) to enhance the quality of a health-related model, extrapolated health-related notification triggers, patient inferences, patient symptom risk predictions, etc. for an individual patient and/or for a subgroup of patients. In particular, the first and second methods S100, S200 can identify a relationship between a health outcome and a communication-related behavior change (e.g., a change in the frequency of phone calls, text messages, and emails over a preset period of time), such as changes in a patient's communication behavior from prior to administration of a treatment to during and/or after administration of the treatment for the patient.
  • The first and second methods S100, S200 can also identify a correlation between health outcome and patient activity data (i.e., accelerometer and gyroscope data from a corresponding mobile device or wearable sensor), local environmental data, patient location data, patient survey data, etc. The first and second methods S100, S200 can then implement an identified relationship between a (communication) behavior change and a change in a patient's health or quality of life over a period of time to anticipate a change in patient health risk, trigger or target an automated or manual intervention (e.g., an automated notification) to assist the patient through a treatment regimen, anticipate symptom relapse and estimate risk of hospital readmission for the patient, etc. The methods can similarly implement the foregoing identified relationship to identify categories, clusters, or subgroups of similar patients within a patient population, to modify or customize treatment regimens or patient health risk models for particular patient clusters, to prescribe a customized or subgroup-specific treatment regimen to a subsequent patient, to assist a care provider in patient triage, to support or justify adjustment to health care utilization, to compare health outcomes (e.g., admission/readmission rates and health care utilization) within and/or across a patient population, etc. The methods can thus function to improve health outcomes for a particular patient and/or within a subgroup of a patient population and to improve health care utilization.
  • Generally, the first and second methods S100, S200 can function to qualify and/or quantify a correlation between a health outcome related to a treatment, therapy, and/or medication, etc. and observed behavior changes based on data gathered passively from a patient's mobile computing device (e.g., mobile phone, smartphone and/or tablet, etc.) substantially without manual data entry from the patient, a doctor, a nurse, a pharmacist, etc. The first and second methods S100, S200 can similarly quantify a correlation between a patient health status and an observed patient behavior based on patient data gathered from the patient's mobile computing device and combine passive behavior data collection with self-reported patient data (i.e., from surveys) to enhance the quality of a treatment regimen, inferences, and treatment predictions for a patient and/or a subgroup of patients. In particular, the first and second methods S100, S200 can identify a relationship between a health outcome and a communication-related behavior change (e.g., phone calls, text messages, emails), such as changes in a patient's communication behavior from prior to administration of a treatment to during and/or after the treatment for an individual patient. The methods can also identify a change in composite work-life balance for the patient, such as based on time spent by the patient in work-related phone calls or sending work-related emails or at a physical office or work location relative to a total communication time of the patient, a total patient waking time per day, or an amount of time spent at home. The methods can thus correlate a change in patient composite work-life balance with a declining patient health status or increased patient health risk (e.g., for recurring or symptoms of increased severity) based on the health risk model. Similarly, the methods can compute a composite quality of life score for the patient, based on time spent in phone calls and at physical locations related to social events, familial and work phone calls, and work locations and transitions between such locations, and the methods can correlate a change in composite quality—of-life with a declining patient health status or increased patient health risk based on the health risk model.
  • The first and second methods S100, S200 can additionally determine a correlation between health outcome and patient mobile phone usage (i.e. screen unlocks, mobile application use) derived from an operating system or the task manager executing on the patient's mobile computing device to determine periods of patient activity, hyper-activity (e.g. frequent unlocks late at night when the patient is unable to sleep), and inactivity. The first and second methods S100, S200 can then implement an identified relationship between a mobile usage behavior change and a change in a patient's health or quality of life over a period of time based on a corresponding health risk model.
  • The first and second methods S100, S200 can also determine a correlation between health outcome and patient activity data (i.e., accelerometer and gyroscope data from the mobile computing device or external wearable sensor), local environmental data, patient location data, patient survey data, etc. The first and second methods S100, S200 can then implement an identified relationship between a (communication) behavior change and a change in a patient's health or quality of life over a period of time to anticipate a treatment efficacy and/or patient response to a treatment, estimate a patient's treatment adherence (i.e., compliance), determine patient satisfaction in a treatment, trigger or target automated or manual interventions (e.g., automated notifications) to assist a patient through a treatment, anticipate symptom relapse and estimate risk of hospital readmission. The first and second methods S100, S200 can similarly implement such an identified relationship to identify categories, clusters, or subgroups of similar patients within a patient population, to modify or customize treatment regimens or notifications for particular patient clusters, to prescribe a customized or subgroup-specific treatment regimen to a subsequent patient, to assist a care provider in patient triage, to support adjustment health care utilization, to compare treatment outcomes (e.g., admission/readmission rates and health care utilization) within and/or across a patient population, etc. to improve treatment outcomes, therapy efficacy, and adherence to a treatment program for a particular patient and/or within a subgroup of a patient population, and health care utilization.
  • Though the following description describes the inventions particularly in the context of monitoring and/or estimating treatment adherence, Blocks of first and second methods S100, S200 can implement similar functionality and techniques to identify patient risk and assess treatment efficacy.
  • 3. Behavior Data
  • Block S110 of the first method S100 recites identifying a first log of use of a native communication application executing on a mobile computing device by the patient within a first time period. Block S120 and Block S140 similarly recite identifying a second log and a third log of use of the native communication application by the patient within a second and a third time period, respectively. Generally, Blocks S110, S120, and S140 function to unobtrusively collect and/or retrieve communication-related data from a patient's mobile computing device, such as through integration within or by interfacing with a native data collection application to collect patient data, as described above. For example, the native data collection application can launch on the patient's mobile computing device as a background process that gathers patient data once the patient logs in to his account. In particular, Block S110, etc. collects communication data and/or native communication application usage data, generated by the patient's mobile computing device, to identify how and how often (i.e., with what frequency) the patient interacts with and communicates with other individuals through phone calls, e-mail, instant messaging, an online social network, etc.
  • In one implementation, Block S110, etc. collects phone call-related data, including a number of sent and/or received calls, call duration, call start and/or end time, location of patient before, during, and/or after a call, and number and times of missed or ignored calls. Block S110, etc. can also collect text messaging (e.g., SMS test messaging) data, including number of messages sent and/or received, message length, message entry speed, efficiency, and/or accuracy, time of sent and/or received messages, and location of the patient when receiving and/or sending a message. Block S110, etc. can collect similar types of data on textual messages sent through other communication venues, such as public and/or private textual messages sent to contacts of the patient through an online social networking system, reviews of products, services, or businesses through an online ranking and/or review service, and/or any other text-based communication generated by the patient and communicated to another individual and/or computer network.
  • Block S110, etc. can further collect location data of the patient before, during, and/or after (or in the absence of) communication with another individual (e.g., a phone call) and/or computer network (e.g., a social networking message), such as by retrieving a GPS location from a GPS sensor within the patient's mobile computing device, estimating the location of the patient's mobile computing device through triangulation of local cellular towers, or identifying a geo-located local Wi-Fi hotspot, etc., during a phone call. Block S110, etc. can apply this data to track patient behavior characteristics, such as patient mobility, periods of patient isolated, patient work-life (e.g., balance based on time spent at specific locations), etc. Block S110, etc. can also collect patient location data before, during, and/or after communication with another individual, such as via a phone call and/or over a computer network (e.g., with a social networking message), and merge patient location with patient communication (or other) data. Block S110, etc. can therefore track the patient's mobility during a communication. Block S110, etc. can additionally or alternatively collect data pertaining to individuals in contact with the patient during the first period of time, etc., such as an individual's location during a phone call, phone number, contact duration and/or type with the patient, relationship to the patient or patient contact group (e.g., top contact, spouse, family member, friend, coworker, business associate, etc.) specified by the patient or learned from previous patient communications, etc.
  • Blocks S110, S120, etc., can further capture mobile usage data like screen unlocks and mobile application usage, such as by retrieving usage information from mobile operating system logs or task manager on the mobile computing device. Blocks of the methods can therefore track variations and periods of activity and inactivity for a patient through data automatically collected on the patient's mobile computing device, such as to estimate extended periods when the patient was hyperactive on the device or not asleep.
  • In one implementation, Block S110, etc. also collects or retrieves patient physical activity- or physical action-related data (e.g., accelerometer and gyroscope data), local environmental data, patient nutrition or diet-related data, etc. such as recorded through sensors within the patient's mobile computing device or through a wearable or other peripheral device in communication with the patient's mobile computing device. For example, a wireless-enabled scale, blood pressure sensor, and a pulse-dosimeter sensor can transmit the patient's weight, blood pressure, and blood oxygen level to the patient's mobile computing device, and Block S110 can add this data to the patient's account to further augment patient behavior data.
  • Block S110 can subsequently aggregate phone, text message, email, social networking, and/or other patient communication data for a particular period of time into a qualitative and/or quantitative feature for the patient for the particular time period. The feature can be specific to a day, a week, a month, a day period (e.g., morning, afternoon, evening, night), a time block during a day (e.g., one hour), a specific communication action (e.g., a single phone call, a set of communication actions of the same type (e.g., a set of phone calls within a two-hour period), all communications within a period of time, etc.). For example, Block S110 can generate a weighted composite of the frequency, duration (i.e., length), timing (i.e., start and/or termination), and contact diversity of all outgoing voice (e.g., phone call) communications and a frequency, length, and timing and/or response time to (i.e., time to accept) incoming voice communications within the first period of time through a phone call application executing on the patient's mobile computing device. Block S110 can additionally or alternatively assess incoming and/or outgoing textual communications from a textual messaging application executing on the mobile computing device. Block S110 can also generate a quantitative assessment of a frequency of, a duration of, and a response time to both incoming and/or outgoing phone calls and textual communications to the mobile computing device during the first time period as a single qualitative and/or quantitative feature corresponding to the first period of time. Block S120 and/or S140 can implement similar methods to generate a feature for the second period of time and third period of time, respectively.
  • Blocks S112, S122, etc. can further extract features based on voice communications, textual communications, mobile application activity usage, location data, etc., which can be based on variance, entropy, or other mathematical and probabilistic computations of basic data, such as a composite activity score, a composite socialization score, a work-life balance score, a quality-of-life score, etc.
  • In one example, Block S110 implements machine learning, data mining, and statistical approaches to process patient communication data into relevant patient communication behavior features (e.g., data points). Block S110 can implement similar techniques to similarly process patient motion data, local environmental, and other automatically/passively collected data.
  • Block S240 of the second method S200 recites, for patients within the subgroup, characterizing communication behavior of a patient based on use of a native communication application executing on a corresponding mobile computing device by the patient during the period of time. Block S240 can thus implement functionality similar to Block S110, Block S120, etc. to extrapolate patient communication behavior from phone calls, text messages, social networking communications, emails, and/or other communications originating and/or terminating through one or more communication applications executing on one or more mobile computing devices associated with one or more patients. In one example, Block 240 characterizes communication behavior of multiple patients based on use of native communication applications executing on corresponding mobile computing devices by the patients both prior to initiation of and during administration of an action item specified in an automated notification directed to the patient and/or an associated nurse, doctor, care provider, etc. In another example, Block S240 characterizes communication behavior of multiple patients based on use of native communication applications executing on corresponding mobile computing devices by the patients both prior to initiation of similar treatment regimens and during administration of the treatment regimens.
  • Block S242 of the second method, which recites characterizing communication behavior of a subsequent patient based on use of a native communication application executing on a corresponding mobile computing device by the subsequent patient, can also implement such functionality to characterize communication behavior of a new (i.e., subsequent) patient. For example, Block S242 can pass the communication behavior characterization of the new patient to Block S282, and Block S282 can determine a similarity between the new patient and a subgroup of patients based on communication behaviors of the two and, from this determined similarity and an effectiveness of the health-related notifications within the subgroup, generate a health-related notification for the new patient. In another example, Block S242 passes the communication behavior characterization of the new patient to Block S282, wherein Block 282 determines a similarly between the new patient and a subgroup of patients based on communication behaviors of the two and, from this determined similarity, predicts an effectiveness in a treatment regimen for the new patient based on the effectiveness of the treatment regimen within the subgroup.
  • However, Blocks S110, S120, S140, S240, S242, etc. can function in any other way to identify and/or characterize use of a native communication application by a patient within a period of time.
  • 4. Surveys
  • Block S112 of the first method S100 recites receiving a first survey response corresponding to the first time period from the patient. Block S122 similarly recites receiving a second survey response from the patient within the second time period. Generally, Block S112, etc. functions to prompt the patient to self-report additional health-related data that can be implemented within the first and/or second method 5200 to qualify patient communication data, such as to teach a patient regimen adherence model relating patient communication to patient treatment adherence and/or patient symptom presentation. For example, Block S112 transmits a survey to the patient's mobile computing device, the native data collection application executing on the mobile computing device opens the survey and prompts the patient to enter relevant data, and Block S112 receives patient responses to the survey from the mobile computing device once the survey is complete. In this example, Block S170 can function to extract a treatment response of the patient from a corresponding survey.
  • Block S112 can generate the survey that includes prompts to enter a pain level (e.g., on a body-location-specific basis), presentation of symptoms (e.g., on a health-condition-specific basis), adherence to a treatment regimen (e.g., if and when the patient took a prescribed medication of some dosage), changes in presentation of symptoms after a treatment, how the patient “feels” generally, experience with a treatment and corresponding effect, a mood, a sleep quality or wakefulness, etc. The survey can also prompt the patient to supply information related to a diagnosed disease or condition, such as major depressive disorder, diabetes, or chronic obstructive pulmonary disease. Alternatively, Block S112 can retrieve relevant patient health data from a medical record, history, or profile of the patient. In one example, Block S112 can import a patient medication record from an Electronic Medical Records (EMR) system hosted by a healthcare provider or health insurance company, such as via a supported Application Programmable Interface (API).
  • Surveys can be presented to the patient on the mobile computing device at preset or patient-selected launch times and/or frequency. For example, Block S112 can prompt the patient to fill out a two-question survey every morning at 9:00 or after each meal, including whether he has taken prescribed medication(s) and his overall satisfaction with the treatment in mitigating symptoms. Alternatively, presentation of surveys can be triggered by a determined patient behavior or symptom change. In one example, Block S112 can trigger presentation of a survey to the patient in response to a disparity in actual and anticipated patient communication behavior, such as significantly more or significantly less phone call time by the patient than expected based on past patient communication behavior. In another example, Block S112 can trigger delivery of a pain survey launched within the application either at a pre-determined time and/or once a period of time (e.g., 72 hours) transpires without a detected change in the patient's location (e.g., the patient has not left his house) and/or without a detected phone call initiated by the patient.
  • Surveys can include a single or combination of question-and-answer types, such as multiple choice questions each with a single answer option, multiple choice questions each with multiple answer options, textual or numerical manual entry, slider (e.g., for easy number or level selection from a range), icon animation selection (e.g., icons indicating different intensities or symptoms), etc., as shown in FIG. 1A. For example, a survey can include a Patient Health Questionnaire (PHQ9), a WHO Well-being Index (WHO5), or a one-question Pain Rating Scale question.
  • Block S112 can additionally or alternatively generate a prompt targeted at a family member, care providers, health organization, etc. associated with the patient. For example, Block S112 can prompt and collect survey responses from a nurse, including manually-entered patient symptom and treatment data. Alternatively, Block S112 can retrieve this information from the personal mobile computing device of a family member of the patient. Block S112 can similarly prompt a pharmacist or pharmacologist associated with the patient's treatment to enter patient prescription data directly into a survey. Alternatively, Block S112 can retrieve this data automatically from an associated pharmacy database. Yet alternatively, Block S112 can collect patient treatment data from a third-party system or device used by the patient or associated care provider, such as an Internet-enabled pillbox with embedded sensors.
  • Block S112 can further function to compensate the participant for completing a survey, such as with cash, a gift card, a pharmacy discount, a health insurance discount, etc., such as in an application of the first and/or second method S200 in which patient data is supplied to pharmaceutical researches to generate treatment regimens.
  • Block S112 can be implemented within the native data collection application, on a computer network in communication with the mobile computing device (e.g., via an Internet connection), or in any other suitable way. Block S122, etc. can implement similar techniques or functions, though Block S112, Block S122, etc. can function in any other way to collect survey responses corresponding to various time periods from the patient.
  • 5. Data Storage
  • The first and second methods S100, S200 can store data locally on the patent's mobile computing device and/or in a remote database on a computer network. For example, private health-related patient data can be stored temporarily on the patient's mobile computing device in a locked and encrypted file folder on integrated or removable memory. In this example, the patient's data can be encrypted and uploaded to the remote database once a secure Internet connection is established. However, patient data can be stored on any other local device or remote data in any other suitable way and transmitted between the two over any other connection via any other suitable communication and/or encryption protocol.
  • 6. Treatment Adherence
  • Block S114 of the first method S100 recites estimating a first adherence to the treatment regimen by the patient within the first time period based on the first survey response. Block S124 of the first method S100 similarly recites estimating a second adherence of the patient within the second time period based on the second survey response. Furthermore, Block S230 of the second method S200 recites, for patients within the subgroup, estimating adherence of a patient to a prescribed treatment regimen during a period of time based on survey responses entered by the patient through a corresponding mobile computing device. Generally, Blocks S114, S124, S230, etc. function to extract the patient adherence to a treatment regimen over time from corresponding survey responses collected in Blocks S112, S122, etc. For example, a survey can include an explicit inquiry into if, when, and in what dosage the patient took a medication specified in a pharmacotherapy regimen and/or if the patient completed and the duration of a physical therapy session, and Block S114 can identify the patient's response to the prompt, compare the patient's response to a treatment regimen assigned to the patient for the corresponding time period, and thus determine if and/or to what extent the patient fulfilled the treatment regimen for the corresponding time period. Block S114 can subsequently pass the degree of patient adherence to the treatment regimen to Block S116. Block S124 and Block S230, etc. can implement similar functionality.
  • Block S116 of the first method S100 recites correlating the first log of use of the native communication application with the first adherence to the treatment regimen, and Block S126 similarly recites correlating the first log of use of the native communication application with the first adherence to the treatment regimen, and Block S126 recites correlating the second log of use of the native communication application with the second adherence to the treatment regimen. Generally, Block S116, Block S126, etc. function to define a relationship between a degree of patient adherence to a treatment regimen and a quality and/or quantity of communication (i.e., communication behavior) of the patient for a corresponding period of time.
  • In one implementation, Block S116 characterizes patient-provided treatment adherence data from the first survey and associates the treatment adherence characterization with the communication behavior characterization from Block S110. At a later time when the patient's communication behavior mimics the communication behavior during the first period, the first method S100 can thus anticipate the patient's treatment adherence at the later time to mimic the patient's treatment behavior during the first time period. In one example, Block S116 correlates a log of use of a first set of outgoing voice communications from a phone call application during the first period with conformity to the treatment regimen by the patient, and Block S126 correlates a frequency of a second set of outgoing voice communications within the phone call application with neglect of the treatment regimen by the patient, as shown in FIG. 4, wherein the log of use of the first set of outgoing voice communications is greater than the frequency of the second set of outgoing voice communications.
  • Block S116 can additionally or alternatively correlate adherence to the treatment regimen and/or patient communication behavior to patient health status, such as presentation of patient symptoms. In one example, Block S116 extracts a patient symptom level from the first survey and presentation of patient symptoms (e.g., how well the patient is feeling) with patient communication behavior (e.g., how, when, and who the patient contacts during a period of time). In this example, Block S116 can determine that a period during which the patient sends a low volume of text messages to a large breadth of contact types and engages in long phone conversations with a limited number of contacts corresponds to minimal symptom presentation, and Block S126 can determine that a period during which the patient sends a low volume of text messages to a low number of unique contacts and engages in short phone conversations with a single contact corresponds to a high degree of symptom presentation. In this example, Block S116 can also correlate minimal symptom presentation with a certain degree of treatment adherence during the first period (e.g., the patient followed the treatment adherence properly), and Block S126 can further correlate higher degree of symptom presentation with a lesser degree of treatment adherence during the second period (e.g., the patient missed a treatment or took an improper dosage).
  • The first method S100 can repeat elements of Blocks S110, S112, S114, S116, etc. to generate a set of communication behavior-treatment adherence (and health status) features for the patient over time, and these features can be fed into Block S130, Block S160, Block S144, Block S150, Block S172, and/or Block S180, etc. to enable further functionality.
  • Block S250 of the second method, which recites correlating communication behavior of a patient with health statuses of patients within the subgroup, can implement similar functionality to generate communication behavior-treatment adherence (and health status) features within the subgroup of patients. For example, Block S250 can pair anonymized communication data from various patients exhibiting similar behaviors over time, such as before, during, and after a treatment regime, with survey responses from corresponding patients at corresponding times to output a collection of features specific to the subgroup. Block S250 can thus pass this collection of features to Block S280. Block S280 can thus implement the collection of features to identify trends or patterns between communication behavior, treatment adherence, and presentation of symptoms across the population and generate a corresponding treatment regimen model for the subgroup, as described below.
  • 7. Subgroups
  • Block S160 of the first method S100 recites selecting a subgroup of a patient population based on the first log of use of the native communication application and a communication behavior common to the subgroup. Generally, Block S160 functions to identify one or more patients within a patient population exhibiting characteristics similar to those of the (current, new, or subsequent) patient, such that subsequent Blocks of the first method S100 can apply preexisting communication behavior data, treatment adherence data, health status data, etc. and corresponding characterizations, patterns, models, etc. of this data to inform and/or improve manipulation of the patient's health-related data.
  • In one implementation, Block S110 characterizes patient communication behavior prior to beginning a treatment regimen, and Block S160 selects a set of other patients within a patient population who exhibit(ed) similar communication behaviors prior to beginning similar treatment regimens. Block S160 can further select and/or filter the patient population based on a diagnosis, therapy prescription, age, gender, location, and/or other demographic, activity behaviors prior to beginning similar treatment programs, and/or other factors or variables similar to those of the patient. Block S160 can further select the set of other patients and/or update the set of other patients during the patient's treatment regimen based on similar in-treatment communication behaviors (and other behaviors or variables) between the patient and other patients in the patient population.
  • Block S160 can therefore receive patient communication behavior characterizations of the patient and other patients within the patient population, such as from Block S110 and Block S120, to identify communication behaviors common to the patient and to the subgroup. Block S160 can also select the subgroup as pertinent to the patient based on treatment adherence data and/or trends and/or symptom data and/or trends common to the patient and the subgroup, such as before and/or during a related treatment regimen. Similarly, Block S160 can select the subgroup based on determined relationships between treatment adherence and communication data common to the patient and the other patients within the subgroup.
  • Block S210 of the second method, which recites identifying a population of patients diagnosed with a health condition, and Block S220, which recites selecting a subgroup of patients within the population exhibiting a similar behavioral characteristic, can implement similar functionality to identify a subset of patients within a patient population that share commonalities, such as one or more diagnoses, health conditions, symptoms, treatment regimens, communication behavior characteristics, treatment adherence characteristics, etc. For example, Block S210 can identify a population of patients prescribed a treatment regimen for a particular diagnosed health condition, such as by identifying anonymized patient profiles tagged with the particular diagnosed health condition, and Block S220 can select a subgroup of patients from the population of patients based on similar communication behavior prior to initiation of the treatment regimen in Block S220. Blocks S230, S240, and S250 can therefore analyze and manipulate communication and/or survey data associated with selected patients within the subgroup (i.e., to estimate treatment regimen adherence, to characterize communication behavior, and to correlate communication behavior and health status of patients, respectively) before and/or after the subgroup is selected in Block S220. However, Blocks S160, S210, and S220 can function in any other way to select a subgroup of related patients from a population of patients.
  • 8. Behavior Feature Engine
  • Any one or more of the foregoing Blocks of the first and second methods S100, S200 can implement a Behavior Feature Engine (BFE) to handle patient data before, during, and/or following a treatment regimen and to then to manipulate patient data into an effective model.
  • The BFE includes a collection of statistics—a combination of unique variables or “features”—that vary amongst patients, conditions, and disease states. A subset of features defined in the BFE can thus be implemented to predict the health status of a patient, such as given communication behavior and/or treatment adherence of the patient. The methods can collect raw patient data from the data collection application described above, such as through Block S110, S112, etc. and can then convert this raw data into statistical features to create the BFE, such as on a remote server for a subgroup of patients or locally on a mobile computing device for a particular patient.
  • For a particular patient, the BFE can extract independent features from one of various types of data or modalities (e.g., phone calls, text messages, instant messages). For example, the BFE can aggregate patient communication data, including a total number of calls made, received, accepted, missed, etc. and a total number of text (e.g., SMS text) messages sent, received, ignored, etc. on a particular day or within another period of time, as shown in FIG. 1A. From this data, the BFE can estimate an interaction balance for the patient, such as by calculating a ratio of incoming communications (e.g., phone calls and text messages) to total communication within a period of time, as well as a patient responsiveness, such as by calculating a number of “missed” interactions within the time frame based on a number of calls not accepted or text messages ignored by the patient with the period of time.
  • The BFE can also calculate patient interaction diversity, including a total number of individuals with whom the patient interacts, such as through a voice communication application or text messaging application, within the period of time. The BFE can further extract a patient mobility from patient communication location data, including an approximation of total patient movement during a particular phone call or during the total of a corresponding period of time, such as based on an estimated distance traveled by foot or bike (excluding a distance traveled by car, train, etc.) based on GPS location data of the patient's mobile computing device (e.g., smartphone). The BFE can then generate a mobility radius factor—defining an approximate radius of an imaginary circle encompassing locations visited by a patient within a period of time—and tag or associate patient communication data with this mobility radius factor.
  • The BFE can also calculate other location data features, such as time spent at home, time spent at work, transition between home/work/social locations, isolation at a particular location for extended periods of time, unpredictability (i.e., entropy) in location information, etc. The BFE can also implement mobile phone usage, such as screen unlocks and application usage, to calculate features that summarize variations and periods of patient activity and inactivity, identify extended periods during which the patient is hyperactive on the device, and/or determine sleep patterns.
  • The BFE can further handle other types of patient data. For example, the BFE can estimate an amount of physical activity undertaken by the patient within the period of time based on motion sensor (e.g., accelerometer) data recorded on the patient's mobile computing device or wearable device, as shown in FIG. 1B. In this example, the BFE can also implement machine learning, such as feature extraction or pattern matching, to correlate motion data with a particular type of action, such as walking, playing tennis, or eating. The BFE can then associate communication data (e.g., times and types) with the activity data, such as an amount of time the patient spent in phone calls during an activity, set of activities, or period of time including one or more particular activities. However, the BFE can handle, manipulate, and/or aggregate any other type of patient data, such as any of the data collected as described above.
  • The BFE can segment patient behavior to enable identification of patient behavioral patterns. For example, the BFE can segment patient data according to any suitable period of time or time frame, such as by whole day, a preset number of hours, daytime and nighttime, weekdays and weekends, months, etc., and thus extrapolate an hourly, daily, weekly, or other time-dependent behavioral pattern. The BFE can similarly segment patient behavior data by patient location, such as home, office, etc., and thus identify location-related patient behavioral patterns.
  • Feature extraction of patient data within the BFE can be driven by predefined metrics, such as distinctness, aggregation, entropy, or percentage change over time for a particular patient, and the BFE can group such features by time, patient, patient subgroup, etc. Features can be stored as standalone features or, as behavior changes, features generated through comparison of two or more standalone features from different times (or across a group or in comparison to a baseline feature). The BFE can then extract a time-dependent behavior pattern for a patient or a subgroup of patients from these features. For example, the BFE can identify and measure statistically significant changes in patient behavior throughout a treatment, including before, during, and/or after administration of the treatment (e.g., a pharmacological treatment, physical therapy, etc.).
  • The BFE can also generate comparative representations of base patient behavior features. For example, the BFE can compare a patient behavior feature value for a specific period of time against any one or more of a baseline (i.e., typical or normal) behavior for the patient as indicated by patient data gathered over time, historical patient behavior data (e.g., at a similar times of day or days of a week), an expected or healthy behavior as suggested by an expert, healthcare professional, literature, a past study, and/or an average or common behavior observed in patient population or patient subgroup exhibiting behaviors similar to the those of the patient. The BFE can output such comparative representations in the form of differences, ratios, percentages, etc. and can handle the comparative representations as discrete features.
  • 9. Predictive Modeling Engine
  • Blocks S130, S280, S284 of the first and second methods S100, S200 recite generating models based on communication and survey data of one or more patients. Generally, Blocks of the first and second methods S100, S200 can implement the BFE in conjunction with a Predictive Modeling Engine (PME) to generate patient-, patient subgroup-, and/or treatment-specific models.
  • Generally, the PME functions to identify and measure statistically significant changes in patient behavior through periods of varied symptom presentation or throughout a treatment, including before, during, and/or after administration of the treatment (e.g., a pharmacological treatment, physical therapy, etc.), to generate a (patient- or subgroup-level) predictive model accordingly, and to train the predictive model with additional patient data over time. The PME can identify correlations between patient behavior(s), symptoms, treatment adherence, and other features extracted in the BFE to transform a large amount of patient data into a predictive model defining links between such observed patient data and associated efficacy, mitigation, abatement of patient symptoms and/or health condition via one or more treatments (i.e., therapies), patient wellness, quality of life, and/or general health of the patient.
  • The PME can therefore implement patient communication data to construct a patient behavior model from multiple patient-related features, such as including total call duration over a period of time (e.g., for a day, for a week), proportion of time spent on calls on weekdays and weekends (or days and nights), number of unique contacts in communication with the patient, and percentage change in patient communication with certain contacts over a baseline period of time. The PME can also associate such features—relative to the baseline—with patient treatment adherence, symptom presentation, and health condition progression to generate a holistic patient- or subgroup-specific treatment model.
  • In one example, the PME generates a model including a derived link between patient health status and patient behavior features extracted in the BFE by identifying and measuring statistically significant changes in behavior during days characterized by enhanced patient symptoms and days characterized by relatively normal patient symptoms for an individual patient or across a patient population subgroup. In this example, the PME can identify a correlation between low patient activity and enhanced symptoms and generate a predictive model accordingly. Such data can be applied to a particular patient, a patient subgroup, and/or a patient population to quantify how effectively a treatment enables a patient to be more active and return to a normal work-life balance in comparison to an alternative treatment. In another example, the PME analyzes behavior data of a bipolar patient to correlate patient behaviors with patient symptoms. In this example, the PME and BFE can cooperate to identify baseline patient communication behavior and associate this with a normal patient state, identify a period of low patient activity and isolated communication behavior and associate this with a depressive episode, and identify a period of patient hyper-activity, long work hours, and a high degree of socializing and associate this with a manic episode. The PME can thus generate a patient-specific model relating patient behavior with normal, manic, and depressive states. The PME can therefore manipulate changes in patient activity or behavior, work, work-life balance, productivity, stress and anxiety, health and wellness, treatment outcome (e.g., self-reported by the patient or entered by a caregiver, family-member, provider, employer, etc.), etc. to identify correlations between patient behavior and patient health and build predictive models based on behavior change, treatment adherence, symptom presentation, and/or treatment outcome patterns.
  • As described above, the PME can implement machine learning, data mining, and/or statistical approaches to generate one or more models specific to a patient, a patient subgroup, or a patient population. In particular, the PME can apply statistical approaches to inform an understanding of underlying patient behavioral data through distributions, correlations, hypothesis testing, etc. by pairing original features extraction within the BFE with clinical insight and domain-specific intuition, such as from a doctor, nurse, pharmacist, etc. The PME can apply feature selection approaches to determine a most predictive subset of features. For example, the PME can implement correlation-based feature selection (CFS), minimum redundancy maximum relevance (mRMR), Relief-F, symmetrical uncertainty, information gain, or other statistical methods and statistic fitting techniques to select a set of features output from the BFE. The PME can also implement support vector machines, ensemble learning (e.g., boosting, random forest), logistic regression, Bayesian learning (e.g., wherein an outcome variable is a predefined class or category), or other machine learning techniques to predict patient behavior, symptoms, treatment outcome, etc. In certain cases, additional complexity can be added to this process. For example, the PME can implement cost-sensitive learning and sampling approaches when distribution of data between categories or classes is unequal, such as when a symptom of interest is depression, which is detectable within a patient on fewer than 10% of total days of a patient study.
  • In one implementation, the PME applies regression approaches and generalized linear models for model outputs that include continuous values, such as prediction of symptom severity. In another implementation, the PME applies Support Vector Machine (SVM) classification approaches for model outputs that include discrete values, such as for patient health risk prediction. In another implementation, the PME applies clustering techniques for model outputs that include a set of undefined clusters, such as a centroid-based, density-based, connectivity-based, and/or a distribution-based approach to define a cluster based on patient behaviors, as shown in FIG. 5. For example, the PME can apply clustering techniques for a model output that specifies a group of patients with similar symptoms or behaviors together. The PME can apply similar techniques to detect anomalies and/or outliers in aberrant cases in collected data, such as patients within a population who do not respond to a treatment or show adverse effects uncharacteristic of the greater patient population. The PME can further train a linear model by fitting a line to collected data or features, and the PME can train a Bayesian network classification model and/or a Gaussian mixture classification model through Expectation-Maximization.
  • The PME can additionally or alternatively model subgroups of patients with a population of patients. In one implementation, the PME identifies a subset (i.e., subgroup) of patients with a certain degree of similarity with one subset of patients in the population and a certain degree of dissimilarity with another subset of patients in the population, such as with respect to one or more behaviors. For example, for a particular therapy, the PME can select a first subgroup of patients who become more active during the therapy and a second subgroup—exclusive of the first subgroup—of patients who show little marked behavior change during the therapy. In this example, the PME can also assign different intervention triggers, reminder types, notification triggers, etc. to each of the first and second subgroups. The PME can therefore apply behavior data passively collected from a patient population to identify clusters of related patients within the population, as shown in FIG. 5.
  • As described above, the PME can generate models that output predicted patient treatment adherence, patient symptom presentation, treatment efficacy, patient response to a treatment, etc. However, the PME can generate models that additionally or alternatively incorporate or output one or more of a risk of admission or re-admission to an associated care provider, risk of a pending emergency room visit, risk of retrogression of clinical symptoms, etc. A model generated through the PME can therefore support or enable early detection of higher-risk patients, prediction of patient hospitalization or symptom relapse, and timely patient intervention and messaging to improve or change patient behaviors.
  • The PME can further train or update a model over time, such as when new data from a specific patient or anonymized data from a patient subgroup becomes available. For example, the PME can verify a previous prediction outcome model for a specific patient in response to receiving feedback from the patients. Alternatively, the PME associates the patient with an alternative subgroup and selects an alternative model for the patient, accordingly, based on new patient feedback.
  • 10. Health Risk Identification
  • Block S232 of the second method 5200 of the method recites identifying a relationship between communication behaviors of patients within the subgroup, characteristics of medical symptoms of patients within the subgroup, and a treatment regimen administered to patients within the subgroup. Block S280 of the second method S200 further recites generating a health risk model for the subgroup based on the relationship, the health risk model defining a correlation between a change in communication behavior and risk of change in a medical symptom.
  • In this implementation of the first method, an individual patient and can be diagnosed or at-risk for a certain condition, and Blocks of the first method can therefore estimate symptom severity, health status, or health risk change] for the patient based on volunteered survey results and/or an applicable health risk model for a corresponding subgroup of patients. Generally, the first method S100 can implement the health risk model generated in the second method S200 to predict a health risk of an individual patient. In particular, Block S162 of the first method can retrieve the health risk model associated with the subgroup (and defining a correlation between risk of change in a medical symptom and communication behavior for patients within the subgroup), and Block S172 can predict a risk of change in a medical symptom for the patient based on the log of use of the native communication application and the health risk model.
  • Blocks of the methods can therefore extract patient symptom severity or health status over time from corresponding survey responses collected in corresponding Blocks of the methods (e.g., Blocks S112, S122, etc.). For example, a survey delivered to a patient can include an explicit inquiry into if and when a symptom occurred and a severity of symptom presentation, and Block S114 can analyze the patient's response to the survey to determine if and/or to what extent the patient symptoms presented for the corresponding time period. Block S114 can subsequently pass the degree of symptom severity to Block S116. Block S124, Block S230, etc. can implement similar functionality to deliver and assess patient surveys.
  • Block S116 (and Block S126, etc.) can subsequently define a relationship between a degree of symptom severity or health status and a quality and/or quantity of communication (i.e., communication behavior, location behavior, and/or phone usage behavior) of the patient for a corresponding period of time. Block S116 can then pass this relationship to Block S172 to predict symptom severity or health status for the patient and/or for other patients in a corresponding subgroup.
  • In one implementation, Block S116 characterizes patient-provided symptom severity and health status data from the first survey and associates this characterization with the communication behavior characterization from Block S110. At a later time when the patient's communication behavior mimics the communication behavior during the first period, the first method S100 can thus anticipate the patient's symptom severity and health status at the later time to mimic the patient's treatment behavior during a corresponding time period. In one example, Block S116 correlates a frequency of a first set of outgoing voice communications from a phone call application during the first period with high symptom severity for the patient, and Block S126 correlates a frequency of a second set of outgoing voice communications within the phone call application with low symptom severity for the patient, wherein the frequency of the first set of outgoing voice communications is greater than the frequency of the second set of outgoing voice communications.
  • Block S116 can additionally or alternatively correlate patient symptom severity or health status and/or patient communication behavior to a change in healthcare utilization, such as readmission or increased hospital visits. In one example, Block S116 extracts a patient symptom severity from the first survey and presentation of patient symptoms (e.g., how well the patient is feeling) with patient communication behavior (e.g., how, when, and who the patient contacts during a period of time). In this example, Block S116 can determine that a period during which the patient sends a low volume of text messages to a large breadth of contact types and engages in long phone conversations with a limited number of contacts corresponds to low symptom severity, and Block S126 can determine that a period during which the patient sends a low volume of text messages to a low number of unique contacts and engages in short phone conversations with a single contact corresponds to a high symptom severity. In this example, Block S116 can also correlate low symptom severity with a certain amount or pattern of healthcare utilization during the first period (e.g., the patient did not have to visit the doctor for anything other than regular checkups), and Block S126 can further correlate high symptom severity with increased healthcare utilization during the second period (e.g., the patient had to visit the doctor several times beyond the regular visits).
  • The first method S100 can repeat elements of Blocks S110, S112, S114, S116, etc. to generate a set of communication behavior-symptom severity (and health status) features for the patient over time, and these features can be fed into Block S130, Block S160, Block S162, Block S144, Block S150, Block S172, and/or Block S180, etc. to enable further functionality.
  • Block S232 of the second method S200, which recites identifying a relationship between communication behaviors of patients within the subgroup, characteristics of medical symptoms of patients within the subgroup, and a treatment regimen administered to patients within the subgroup, can implement similar functionality to generate communication behavior-symptom severity (and health status) features within the subgroup of patients. For example, Block S232 can pair anonymized communication data from various patients exhibiting similar behaviors over time, such as while presenting low or high symptom severities, with survey responses from corresponding patients at corresponding times to output a collection of features specific to the subgroup. Block S232 can thus pass this collection of features to Block S280
  • In one implementation, Block S280 applies the collection of features to identify trends or patterns between communication behavior, symptom severity, and healthcare utilization across the population and generates a corresponding intervention plan for the subgroup, as described below.
  • 11. Health Risk Model
  • As described above, Blocks of the first and second methods S100, S200 can implement the PME to generate patient- and/or subgroup-specific models.
  • In particular, Block S, which recites generating a health risk model including the first log (e.g., frequency) of use of the native communication application, the second log of use of the native communication application, the first symptom severity or health status, and the second symptom severity or health status, can implement the PME to generate a model linking patient communication behavior and health risk (e.g., symptom severity, risk of relapse, risk of hospital readmission, health status, etc.). For example, Block S130 can implement the BFE to determine that low communication periods commonly follow periods in which the patient correlates with high symptom severity and that period characterized by relatively high levels of patient activity correspond to low symptom severity. In this example, Block S130 can then implement the PME to generate a corresponding model specific to the patient. Block S130 can subsequently pass this model to Block S144 to estimate a subsequent patient health risk at a subsequent time based on patient behavior at the subsequent time. In particular, in this example, Block S144 can determine that the patient is at a high risk for increased symptom severity if the patient's behavior (e.g., communication behavior and/or activity level) substantially matches recorded patient behavior associated with that symptom severity (as defined by the health risk model).
  • Block S162, which can recite retrieving a health risk model associated with the subgroup, can select a preexisting health risk (or risk identification) model—generated by the PME and associated with a subgroup of patients—based on selection of the subgroup in Block S160. Block S144 can similarly function to pass subsequent patient behavior data into the health risk model to predict a subsequent level of patient risk for presenting symptom greater than a threshold severity. In this variation of the first method, Block S172 can additionally or alternatively predict short-term and/or long-term patient risk for high symptom severity or change in health status or disease progression, such as based on a change in communication behavior of the patient over a certain time period. For example, if the patient's total communications increase during a certain time period and this increase in communications is common for patients within the subgroup who show a decrease in symptom severity—as defined in the health risk model—Block S172 can predict that the patient will exhibit a similar decrease in symptom severity.
  • However, Block S130 and Block S160 can function in any other way to generate and retrieve a predictive model defining a relationship between symptom severity or health status and patient communication behavior over time, respectively.
  • Block S172 can further estimate an efficacy of the patient intervention (e.g., nurse outreach, doctor-patient communication, automated health tips delivered through the patient's mobile computing device) in improving the health condition of the patient according to a comparison between changes in symptom severity or health status of the patient. Block S172 can therefore implement elements of the BFE and/or the PME to predict the outcome of the intervention for the patient. For example, Block S172 can identify a pattern of communication behavior of the patient, correlate the pattern of communication behavior with intervention outcomes, and then compare the interventions (e.g., based on symptom presentation recorded in various survey responses entered by the patient) to changes in healthcare utilization. In this example, if the patient is delivered relevant interventions but shows worsening or unchanged symptoms, Block S172 can estimate a low efficacy of the intervention in treating the patient's condition. Furthermore, in this example, Block S172 can compare the patient to existing health condition data and/or a health risk model for a subgroup of patients to determine if worsening symptoms despite delivering interventions is common or expected for the patient's diagnosed health condition and/or for similar patients (i.e., patients within the subgroup). Block S172 can thus predict the efficacy of the interventions accordingly.
  • Block S260 of the second method 5200 similarly recites estimating an efficacy of the interventions in improving the health condition for patients within the subgroup based on of the delivery of interventions to patients within the subgroup. Generally, Block S260 implements techniques similar to Block S172 to apply elements of the BFE and/or the PME to predict the outcome of the interventions for the subgroup of patients within a patient population. For example, for patients within the selected subgroup, Block S260 can identify an increase in use of native communication applications by patients from prior to initiation of the interventions to the patients to during and after delivery of the interventions to the patients, correlating the increase in use of the native communication applications by the patients with improved health status of the patients, and correlating improved health status of the patients with efficacy of the interventions in improving the health condition of patients within the subgroups. In this example, Block S260 can apply survey response and communication behavior of patients within the subgroup—and a known health condition common to patients within the subgroup—to predict the efficacy of a nurse outreach program in providing support for mental health issues of patients within the subgroup. Additionally or alternatively, Block S260 can implement such patient data to estimate the efficacy of a health tips program in improving health-related behaviors of diabetic patients within the subgroup.
  • In one implementation, block S172 and Block S260 output quantitative values of intervention efficacy for the patient and the patient subgroup, respectively. For example, Block S172 can output a predicted intervention efficacy of 80%, indicating that four out of five similar patients show a reduction in symptom severity or an improvement in health status. As in this example, Block S172 can additionally output a confidence interval indicating a statistical confidence in the intervention regimen, which can be dependent on a size of a patient subgroup associated with an intervention efficacy model used to estimate treatment efficacy for the patient or dependent on a predicted accuracy and/or availability of patient communication and survey response data, as shown in FIGS. 1A and 1C.
  • Alternatively, Block S172 and Block S260 output binary indicators of intervention efficacy. For example, if predicted intervention efficacy falls below a threshold, such as below a 70% predicted improvement rate, Block S172 and/or Block S260 can output a negative indicator of intervention efficacy. In response to a negative predicted intervention efficacy, Block S172 can prompt Block S180 to automatically trigger a notification to care providers suggesting that a different intervention plan might better serve the patient. If data regarding past intervention successes is available modification of the treatment regimen or prompt a care provider to modify the existing treatment regimen or to prescribe a new treatment regimen for the patient, as described below. Similarly, in response to a negative predicted treatment efficacy, Block S260 can prompt Block S284 to implement similar functionality to automatically modify or prompt manual modification of the treatment regimen for the subgroup of patients.
  • As shown in FIG. 1C, one variation of the first method S100 includes Block S180, which recites generating an updated intervention regimen according to the efficacy of the intervention regimen. Generally, Block S180 functions to modify an intervention plan based on an intervention efficacy prediction output in Block S172. In one example, if the patient's symptoms worsen over time when improvement is expected despite delivering relevant interventions, Block S180 can prompt a doctor, pharmacologist, etc. to prescribe an alternative intervention program.
  • As shown in FIG. 2B, Block S284 of one variation of the second method S200 similarly recites updating the intervention regimen for patients within the subgroup and currently prescribed the intervention regimen in accordance with the intervention efficacy predicted in Block S260.
  • In one implementation of Block S284, the second method S200 characterizes communication behavior of a patient following diagnosis of a medical condition and before administration of an intervention program, applies the characterized communication behavior to select a subgroup of patients with similar diagnoses and communication behaviors prior to administration of intervention programs, and selects an intervention program for the patient based on a predicted efficacy of the intervention for the patient, which is informed by an actual efficacy of the intervention within the selected subgroup (and/or a corresponding intervention efficacy model for the subgroup, as described above). Block S284 can function to pair a new patient to a subgroup of current and/or previous patients and to automatically prescribe an intervention regimen to the new patient accordingly.
  • 12. Adherence Model
  • As described above, Blocks of the first and second methods S100, S200 can implement the PME to generate patient- and/or subgroup-specific models.
  • In particular, Block S130, which recites generating a patient regimen adherence model including the first log of use of the native communication application, the second log of use of the native communication application, the first adherence, and the second adherence, can implement the PME to generate a model linking patient communication behavior and treatment adherence. For example, Block S130 can implement the BFE to determine that low communication periods commonly follow periods in which the patient diverts from a prescribed treatment regimen and that period characterized by relatively high levels of patient activity correspond to suitable adherence to the treatment regimen. In this example, Block S130 can then implement the PME to generate a corresponding model specific to the patient. Block S130 can subsequently pass this model to Block S144 to estimate a subsequent patient treatment adherence at a subsequent time based on patient behavior at the subsequent time. In particular, in this example, Block S144 can determine that the patient has effectively complied with his prescribed treatment program if the patient's behavior (e.g., communication behavior and/or activity level) substantially matches recorded patient behavior associated with treatment adherence compliance (as defined by the patient treatment adherence model), and Block S144 can determine that patient has not suitably complied with his prescribed treatment program when the patient's behavior substantially matches recorded patient behavior associated with treatment adherence neglect.
  • In this foregoing implementation, Block S144, which recites estimating a third adherence within the third time period based on the patient regimen adherence model and the third log of use of the native communication application, functions to pass data from a subsequent third period of time into the adherence model generated in Block S130 to predict patient adherence to the treatment regimen during the third period of time. Block S144 can additionally or alternatively estimate a patient response to a most-recent administration of the treatment program, such as based on a patient response component of the treatment adherence model.
  • Alternatively, Block S162, which recites retrieving a regimen adherence model associated with the subgroup, can select a preexisting treatment adherence model—generated by the PME and associated with a subgroup of patients—based on selection of the subgroup in Block S160. Block S144 can similarly function to pass subsequent patient behavior data into the regimen adherence model to predict a subsequent level of patient adherence to a prescribed treatment. In this variation of the first method, Block S172 can additionally or alternatively predict short-term and/or long-term patient response to the treatment, such as based on a change in communication behavior of the patient soon after beginning a treatment program. For example, if the patient's total communications increase soon after beginning a treatment regimen for a particular condition and this increase in communications is common for patients within the subgroup who positively respond to the treatment regimen—as defined in the regimen adherence model—Block S172 can predict that the patient will exhibit a similar positive response to the treatment regimen. Block S172 can similarly predict patient satisfaction in the treatment regimen based on data of other patients within the subgroup and the regimen adherence model.
  • However, Block S130 and Block S160 can function in any other way to generate and retrieve a predictive model defining a relationship between treatment regimen adherence and patient communication behavior over time, respectively.
  • 13. Treatment Regimen Model
  • As shown in FIG. 2B, one variation of the second method S200 includes Block S280, which recites generating a treatment regimen model for the subgroup, the treatment regimen model defining a correlation between communication behavior, treatment responses, and treatment regimen outcomes for patients within the subgroup. Generally, Block S280 can implement the PME to generate a model correlating patient behavior to patient treatment outcome for a subgroup of patients. In one example, Block 242 can implement behavior data of a subsequent patient to match the subsequent patient to the subgroup, and Block S282 can feed behavior data of the subsequent patient into the corresponding treatment regimen model to generate a predicted treatment regimen outcome for the subsequent patient following administration of the treatment regimen as shown in FIG. 2B. Block S282 can similarly predict a health condition relapse risk for the subsequent patient based on the treatment regimen model, such as for the treatment regimen model that includes a relapse risk output component.
  • In one implementation, Block S280 implements the PME to generate a treatment regimen model for the subgroup based on adherence to the treatment regimen, communication behavior, and survey responses of patients within the subgroup throughout a period of time. For example, Block S280 can implement the PME to generate the treatment regimen model for the subgroup by identifying patterns in communication behavior during administration of the treatment regimen and correlating patterns in communication behavior with treatment responses of patients within the subgroup. In this implementation, Block S282 can thus feed any one or more of adherence to the treatment regimen, communication behavior, and a survey response of a subsequent patient into the treatment regimen model to output a predicted effect of the associated treatment on the subsequent patient, such as for a particular period of time during administration of the treatment program to the subsequent patient.
  • However, Block S280 can function in any other way to generate a predictive model defining a relationship between treatment outcome for a patient and patient communication behavior over time.
  • 14. Treatment Efficacy Model
  • Block S172, which recites estimating an efficacy of the treatment regimen in treating the health condition of the patient according to a comparison between the treatment response and the adherence to the treatment regimen by the patient, functions to implements elements of the BFE and/or the PME to predict the outcome of the treatment regimen for the patient. For example, Block S172 can identify a pattern of communication behavior of the patient, correlate the pattern of communication behavior with a pattern of treatment adherence, and then compare the pattern of treatment adherence to a trend in patient treatment response (e.g., based on symptom presentation recorded in various survey responses entered by the patient). In this example, if the patient exhibits strong treatment adherence but worsening symptoms, Block S172 can estimate a low efficacy of the treatment in treating the patient's condition. Furthermore, in this example, Block S172 can compare the patient to existing health condition data and/or a treatment outcome model for a subgroup of patients to determine if worsening symptoms despite strong treatment adherence is common or expected for the patient's diagnosed health condition and/or for similar patients (i.e., patients within the subgroup). Block S172 can thus predict the efficacy of the treatment accordingly.
  • Block S260 of the second method S200 similarly recites estimating an efficacy of the treatment regimen in treating the health condition for patients within the subgroup based on adherence to prescribed treatment regimens and health statuses of patients within the subgroup. Generally, Block S260 implements techniques similar to Block S172 to apply elements of the BFE and/or the PME to predict the outcome of the treatment regimen for the subgroup of patients within a patient population. For example, for patients within the selected subgroup, Block S260 can identify an increase in use of native communication applications by patients from prior to initiation of the treatment regimen by the patients to during administration of the treatment regimen by the patients, correlating the increase in use of the native communication applications by the patients with improved health status (e.g., increased wellness, reduced symptoms) of the patients, and correlating improved health status of the patients with efficacy of the treatment in treating the health condition of patients within the subgroups. In this example, Block S260 can apply survey response and communication behavior of patients within the subgroup—and a known health condition common to patients within the subgroup—to predict an efficacy of a physical therapy regimen in treating physical handicaps for patients within the subgroup. Additionally or alternatively, Block S260 can implement such patient data to estimate the efficacy of a pharmacotherapy regimen in treating mental disorders for patients within the subgroup.
  • In one implementation, Block S172 and Block S260 output quantitative values—from a scale or spectrum of treatment values—of treatment efficacy for the patient and the patient subgroup, respectively. For example, Block S172 can output a predicted treatment efficacy of 80%, indicating that four out of five similar patients recover from a diagnosed condition following the corresponding treatment regimen. As in this example, Block S172 can additionally output a confidence interval indicating a statistical confidence in the predicted treatment regimen, which can be dependent on a size of a patient subgroup associated with a treatment efficacy model used to estimate treatment efficacy for the patient or dependent on a predicted accuracy and/or availability of patient communication and survey response data, as shown in FIGS. 1A and 1C.
  • Alternatively, Block S172 and Block S260 output binary indicators of treatment efficacy. For example, if predicted treatment efficacy falls below a threshold efficacy, such as below a 70% predicted success rate, Block S172 and/or Block S260 can output a negative indicator of treatment efficacy. In response to a negative predicted treatment efficacy, Block S172 can prompt Block S180 to automatically trigger modification of the treatment regimen or prompt a care provider to modify the existing treatment regimen or to prescribe a new treatment regimen for the patient, as described below. Similarly, in response to a negative predicted treatment efficacy, Block S260 can prompt Block S284 to implement similar functionality to automatically modify or prompt manual modification of the treatment regimen for the subgroup of patients.
  • As shown in FIG. 1C, one variation of the first method S100 includes Block S180, which recites generating an updated treatment regimen according to the efficacy of the treatment regimen. Generally, Block S180 functions to modify a treatment prescription based on a treatment efficacy prediction output in Block S172. In one example, if the patient exhibits improved symptoms when a medication dosage is skipped or inadvertently reduced by the patient (or care giver, etc.), Block S180 can respond to such improved symptoms by modifying the prescribed medication dosage specified in the treatment regimen for the patient, as shown in FIG. 1C. In another example, if the patient's symptoms worsen over time when improvement is expected despite strong patient adherence to the treatment program, Block S180 can cancel the current treatment program and prescribe an alterative automatically or prompt a doctor, pharmacologist, etc. to prescribe an alternative treatment program.
  • As shown in FIG. 2B, Block S284 of one variation of the second method S200 similarly recites updating the treatment regimen for patients within the subgroup and currently prescribed the treatment regimen in accordance with the treatment efficacy predicted in Block S260.
  • In one implementation of Block S284, the second method S200 characterizes communication behavior of a patient following diagnosis of a medical condition and before administration of a treatment program, applies the characterized communication behavior to select a subgroup of patients with similar diagnoses and communication behaviors prior to administration of treatment programs, and selects a treatment program for the patient based on a predicted efficacy of the treatment for the patient, which is informed by an actual efficacy of the treatment within the selected subgroup (and/or a corresponding treatment efficacy model for the subgroup, as described above). For example, Block S282 can prescribe a custom medication and custom dosage for the patient based on the predicted treatment regimen outcome for the patient, such as by passing different medications and/or dosages into the subgroup regimen outcome model and selecting a particular medication and/or dosage that yields a highest predicted treatment efficacy. Therefore, Block S284 can function to pair a new patient to a subgroup of current and/or previous patients and to automatically prescribe a treatment regimen to the new patient accordingly.
  • 15. Notifications
  • Block S150 of the first method S100 recites presenting a treatment-related notification based on the third adherence through the mobile computing device. Generally, Block S150 functions to generate and present a notification to the patient through the patient's mobile computing device to provide guidance to the patient during the treatment program, such as in the form of a reminder to perform a particular task (shown in FIG. 1A), suggest or enforce a behavior change, provide positive reinforcement for a behavior, provide behavioral insights, or share a patient symptom and/or behavior timeline. For example, once Block S144 identifies a period of low treatment adherence or treatment neglect, Block S150 can generate a notification including a reminder to take a prescribed medication and transmit the reminder to the patient's mobile computing device, wherein a native application executing on the mobile computing device displays the notification, such as in the form of a pop-up notification. Block S150 can transmit the notification to the mobile computing device for display at a prescribed treatment administration time, such as a 9 AM and 5 PM—dosage times prescribed by a doctor—on a day following estimated treatment program neglect by the patient. Alternatively, Block S150 can transmit the notification to the patient's mobile computing device substantially in real time in response to a patient action, inaction, behavior, or behavior change. For example, Block S150 can generate a notification reciting “everything all right?” and transmit the notification to the patient in response to a detected abrupt change in a patient behavior pattern (e.g., the patient calls the same family member every morning at 9 AM, but 9:15 AM passes on a present day without an outgoing or incoming phone call).
  • Block S150 can also incorporate additional patient information in the notification. For example, Block S150 can generate a notification that specifies a medication dosage for the patient, such as including how many pills of a certain type (or size, shape, color, etc.). In this example, the notification can incorporate dosage information based on an updated treatment regimen, such as output in Block S180. Block S150 can additionally or alternatively generate the notification that includes past patient data indicating an expected outcome if the patient adheres to the treatment regimen and/or an expected outcome if the patient neglects the treatment regimen, such as based on a patient or subgroup treatment adherence model.
  • Block S150 can generate the notification that includes a prompt requesting confirmation of patient implementation of the action item specified in the notification (e.g., the patient took his prescribed medication). For example, the notification can recite “It's time to take your meds” and include input regions reciting “Roger that!” and “Remind me later,” as shown in FIG. 1A. Block S150 can subsequently pass a “Roger that!” selection back into the treatment adherence model of the patient as adherence to the treatment program, and Block S150 can pass a “Remind me later” selection into the treatment adherence model as neglect or delay of the treatment program, as shown in FIG. 3.
  • Block S150 can additionally or alternatively interface with a care provider, such as a nurse or doctor (e.g., through a care provider interface executing on a corresponding computing device), to generate a custom notification. For example, Block S150 can transmit or share patient information, including health condition, treatment regimen, actual and estimated patient adherence, survey response, and/or actual and estimated treatment response data of the patient, with the care provider and guide the care provider in identifying a patient problem or need. Block S150 can then receive from the care provider a selection for a default notification from a pre-populated list of available and/or applicable default notifications. Alternatively, Block S150 can prompt the care provider to enter a custom textual and/or image-based (e.g., infographic) notification for the patient and then pass the notification back to the patient. In one example, Block S172 predicts a risk of change in a medical symptom for the patient based on a risk of increased symptom severity for symptoms associated with a particular disease associated with (e.g., diagnosed in) the selected subgroup, and Block S150 transmits the notification to the care provider if to the risk of change in the medical symptom for the patient exceeds a threshold risk specific to the particular disease, such as a 70% probability that patient symptom severity will systematically increase in the (near) future for a patient diagnosed with diabetes and an 88% probability that patient symptom severity will systematically increase in the (near) future for a patient diagnosed with ADHD.
  • In addition to generating a reminder for the patient, such as to adhere to a treatment program, Block S150 can additionally or alternatively generate a notification to prompt or reminder the patient to complete a survey, schedule a visit with a doctor or health care provider, to attend a scheduled health-related visit, etc. However, Block S150 can generate a notification including any other information and presented to the patient in any other suitable way or through any other suitable medium.
  • 16. Services
  • The first and second methods S100, S200 can function to provide one or more services to a care provider associated with a patient and with a subgroup of patients, respectively. In one implementation, the first and second methods S100, S200 support a disease management program by enabling provision of lower cost treatments and/or therapies to patients in exchange for feedback, wherein the first and second methods S100, S200 collect patient feedback as described above. In this implementation, the first method S100 can support tools and one or more mobile applications (i.e., native applications for mobile computing devices) to enable better patient-driven condition management and to reduce symptom flare-ups, and the second method S200 can aggregate and extract valuable information from patient data across a subgroup to support disease management programs by a care provider. For example, the second method S200 can manage or provide analyzed data to a care provider to manage a population with a chronic condition, such as by collecting data about patients outside a clinical setting and generate reports and/or models to aid triage for higher-risk patients, and the first method S100—in conjunction within the second method S200—can support a communication venue or network to enable more effective outreach from a care provider to a patient.
  • In one example, the first method S100 accesses a second log of use of a native communication application executing on a second mobile computing device by a second patient and predicts a second risk of change in a medical symptom for the second patient based on the second log of use of the native communication application and the health risk model, and Block S150 transmits a scored patient triage list to a care provider (e.g., nurse, hospital, clinic, etc.) according to the risk of change in the medical symptom for the patient and the second risk of change in the medical symptom for the second patient. In another example, Block S150 transmits a notification to the care provider comprises prompting the care provider to schedule a visit with a particular patient.
  • The first and second methods S100, S200 can therefore target patient interventions to improve patient health and reduce healthcare utilization (e.g., care costs, emergency room visits, pharmaceutical reliance, etc.). For example, the first and second methods S100, S200 can generate and implement models to identify patients at higher-risk for declining health, increased healthcare utilization, and higher hospital readmissions. Once such patients are identified, the first and second methods S100, S200 can prompt a care provider to intervene, thereby initiating targeted care interventions to improve patient health and wellbeing.
  • The first and second methods S100, S200 can additionally or alternatively extrapolate insights from population-level data to guide care provider improvement in practices, thereby guiding care providers to better patient outcomes. For example, the second method S200 can detect clusters of patients within a population who demonstrate similar behavioral and outcome patterns, identify two clusters of patients who show significantly different readmission rates, identify different treatment practices across the two clusters, and thus guide a care provider in designing a treatment practice based on the readmission rates and the treatment practices across with the two clusters of patients.
  • In particular, various Blocks of the first and second methods S100, S200 interface with care providers in various ways.
  • As described above, Block S172 can communicate a predicted (e.g., estimated) efficacy of the treatment regimen for the patient to a healthcare provider associated with the patient. For example, if the predicted efficacy of the treatment falls below a threshold efficacy for the health condition for the patient, Block S172 can transmit an alert or other notification to a doctor, nurse, or other care provider noted in the patient's medical file or recorded as managing treatment of the particular health condition for the patient. In one implementation, Block S172 transmits an alert to a corresponding care provider through a private care provider account accessible through a provider-specific interface (e.g., an online healthcare dashboard), such as a care application logged-in to a care provider's private account and executing on a computing device associated with the care provider. The first method S100 can also enable the care provider to access additional patient information through the care application and/or private care account, such as through an online dashboard. Block S172 can also guide the care provider in modifying an existing treatment program or selecting an alternative treatment program for the patient, as described above, such as by displaying a list of available treatment programs for the patient's health condition in a scored list based on similarities between the patient's behaviors and behavior within a subgroup of patients and treatment efficacies within the subgroup (e.g., via a treatment efficacy model). However, Block S172 can function in any other way to communicate a predicted efficacy of the treatment regimen for the patient to a corresponding care provider.
  • Block S290 of the second method, which recites transmitting the predicted treatment regimen outcome following administration of the treatment regimen by the subsequent patient to a care provider specified in a digital health profile associated with the subsequent patient, can implement similar techniques and/or methods to share health- and treatment-related information to a care provider related to the (subsequent) patient. For example, Block S290 can transmit a notification including a prompt to respond to a current health status of the subsequent patient to a corresponding care provider. However, Block S290 can function in any other way to communicate a predicted treatment outcome for a patient to a care provider.
  • Block S270 of the second method, which recites generating a treatment regimen report specific to the subgroup based on the efficacy of the treatment regimen, can implement similar techniques or functionality to communicate a subgroup-wide efficacy report for a particular treatment regimen to a care provider. Generally, Block S270 functions to formulate a report for the efficacy of the treatment regimen with the population based on the estimated efficacy of the treatment regimen output in Block S260. For example, Block S270 can generate the report that includes charts relating patterns in patent symptoms to treatment patterns in treatment adherence, changes in patterns in patient behaviors and survey responses, etc. for within the subgroup over time. In another example, Block S270 generates a treatment report specific to a particular patient within the subgroup, including a predicted risk of health condition decline for the particular patient during administration of the treatment regimen, as shown in FIG. 2A. In this example, Block S270 can thus prompt a care provider to modify a treatment program for a particular patient within a subgroup, such as based on predicted or extrapolated patient risk.
  • As in Block S172, Block S270 can also guide a care provider in modifying or changing a treatment regimen for the subgroup, such as by identifying and displaying (unexpected) patterns of treatment negligence or patient ‘customization’ associated with improved patient symptoms and/or patient behaviors within the subgroup. For example, the second method S200 can implement behavior data of patients within the subgroup to compare the wellbeing of the patients and identify those most in need accordingly. Block S270 can thus generate the treatment regimen report that includes a ranking of patients within the subgroup based on need and symptoms. Thus, rather than randomly select patients within a subgroup to contact during the treatment regimen, a nurse practitioner managing dozens of patients can review the treatment report for the subgroup to more effectively triage and determine the patients that are (likely) in need of support. Block S270—and the second method S200 in general—can therefore enable a nurse or other care provider to make real-time decisions and intervene and/or reach out to a particular patient based on real patient data.
  • However, Block S270 can generate a report that includes any other suitable patient information and can function in any other way to make the report available to one or more care providers, such as related to one or more patients and/or a prescribed treatment regimen for the subgroup.
  • As noted above, though the foregoing description describes the inventions particularly in the context of monitoring and/or estimating treatment adherence, the first and second methods S100, S200 can implement similar functionality and techniques to identify patient risk and assess treatment efficacy.
  • The systems and methods of the embodiments can be embodied and/or implemented at least in part as a machine configured to receive a computer-readable medium storing computer-readable instructions. The instructions can be executed by computer-executable components integrated with the application, applet, host, server, network, website, communication service, communication interface, hardware/firmware/software elements of a patient computer or mobile device, or any suitable combination thereof. Other systems and methods of the embodiments can be embodied and/or implemented at least in part as a machine configured to receive a computer-readable medium storing computer-readable instructions. The instructions can be executed by computer-executable components integrated by computer-executable components integrated with apparatuses and networks of the type described above. The computer-readable medium can be stored on any suitable computer readable media such as RAMs, ROMs, flash memory, EEPROMs, optical devices (CD or DVD), hard drives, floppy drives, or any suitable device. The computer-executable component can be a processor, though any suitable dedicated hardware device can (alternatively or additionally) execute the instructions.
  • As a person skilled in the art will recognize from the previous detailed description and from the figures and claims, modifications and changes can be made to the embodiments of the invention without departing from the scope of this invention as defined in the following claims.

Claims (20)

We claim:
1. A method for improving mental health risk determination for a user, the method comprising:
receiving a dataset from a user device associated with the user;
deriving a set of features associated with the user based at least in part on the dataset;
selecting a mental health risk model for the user based on the set of features;
evaluating the mental health risk model to determine a set of mental health risk parameters for the user;
determining a mental health risk state for the user based on the set of mental health risk parameters; and
in response to and based on the set of mental health risk parameters, initiating provision of an intervention at the user device, the intervention operable to improve the mental health risk state of the user.
2. The method of claim 1, wherein the mental health risk model is determined based on aggregated data from a subgroup of multiple users, wherein the subgroup of multiple users is associated with the set of features.
3. The method of claim 2, wherein the mental health risk model is selected from a set of multiple health risk models.
4. The method of claim 2, wherein the mental health risk model is selected in response to assigning the user to the subgroup of multiple users based on the set of features.
5. The method of claim 1, wherein the set of features is further determined based on demographic information associated with the user.
6. The method of claim 1, wherein the dataset comprises passive information, wherein the passive information is collected passively from the user device.
7. The method of claim 6, wherein the passive information is collected passively from a communication application executing on the user device.
8. The method of claim 6, wherein the dataset further comprises active information, wherein the active information is entered by the user at the user device.
9. The method of claim 8, wherein the active information is collected from a set of surveys administered to the user at the user device.
10. The method of claim 1, wherein the set of mental health risk parameters for the user is further determined based on the dataset.
11. The method of claim 1, further comprising, receiving a second dataset from the user device in response to the intervention, and updating the mental health risk model based on the second dataset.
12. The method of Claim ii, wherein the second dataset comprises a survey dataset from a set of surveys administered to the user.
13. The method of claim 1, further comprising:
receiving a second dataset from a second user device associated with a second user;
deriving a second set of features associated with the second user based at least in part on the second dataset;
detecting an overlap between the second set of features and the first set of features; and
selecting the mental health risk model for the second user based on the overlap.
14. The method of claim 1, wherein initiating provision of the intervention comprises facilitating a communication between the user and a care provider.
15. The method of claim 1, wherein the mental health risk model defines a correlation between the dataset and the mental health risk state of the user.
16. The method of claim 1, wherein the dataset comprises at least one of: a communication dataset, a location dataset, and a motion dataset.
17. The method of claim 1, wherein the mental health risk model is a trained machine learning model, wherein the mental health risk model is trained based on data from an aggregated set of users, the aggregated set of users having a shared subset of the set of features of the user.
18. The method of claim 1, wherein the set of features is derived with a factor analysis approach.
19. The method of claim 1, wherein the intervention comprises an infographic notification displayed at the user device, wherein initiating provision of the intervention comprises automatically initiating provision of the infographic notification.
20. The method of claim 1, wherein determining the mental health risk state for the user comprises comparing the set of mental health risk parameters with a set of thresholds.
US17/518,043 2012-08-16 2021-11-03 Method for modeling behavior and health changes Abandoned US20220059236A1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US17/518,043 US20220059236A1 (en) 2012-08-16 2021-11-03 Method for modeling behavior and health changes

Applications Claiming Priority (6)

Application Number Priority Date Filing Date Title
US201261683869P 2012-08-16 2012-08-16
US201261683867P 2012-08-16 2012-08-16
US13/969,349 US9836581B2 (en) 2012-08-16 2013-08-16 Method for modeling behavior and health changes
US15/807,323 US10068673B2 (en) 2012-08-16 2017-11-08 Method for modeling behavior and health changes
US16/057,199 US11195626B2 (en) 2012-08-16 2018-08-07 Method for modeling behavior and health changes
US17/518,043 US20220059236A1 (en) 2012-08-16 2021-11-03 Method for modeling behavior and health changes

Related Parent Applications (1)

Application Number Title Priority Date Filing Date
US16/057,199 Continuation US11195626B2 (en) 2012-08-16 2018-08-07 Method for modeling behavior and health changes

Publications (1)

Publication Number Publication Date
US20220059236A1 true US20220059236A1 (en) 2022-02-24

Family

ID=50100691

Family Applications (7)

Application Number Title Priority Date Filing Date
US13/969,363 Abandoned US20140052465A1 (en) 2012-08-16 2013-08-16 Method for modeling behavior and health changes
US13/969,349 Active 2033-11-08 US9836581B2 (en) 2012-08-16 2013-08-16 Method for modeling behavior and health changes
US13/969,339 Abandoned US20140052474A1 (en) 2012-08-16 2013-08-16 Method for modeling behavior and health changes
US15/798,005 Active US10068672B2 (en) 2012-08-16 2017-10-30 Method for modeling behavior and health changes
US15/807,323 Active US10068673B2 (en) 2012-08-16 2017-11-08 Method for modeling behavior and health changes
US16/057,199 Active 2035-03-21 US11195626B2 (en) 2012-08-16 2018-08-07 Method for modeling behavior and health changes
US17/518,043 Abandoned US20220059236A1 (en) 2012-08-16 2021-11-03 Method for modeling behavior and health changes

Family Applications Before (6)

Application Number Title Priority Date Filing Date
US13/969,363 Abandoned US20140052465A1 (en) 2012-08-16 2013-08-16 Method for modeling behavior and health changes
US13/969,349 Active 2033-11-08 US9836581B2 (en) 2012-08-16 2013-08-16 Method for modeling behavior and health changes
US13/969,339 Abandoned US20140052474A1 (en) 2012-08-16 2013-08-16 Method for modeling behavior and health changes
US15/798,005 Active US10068672B2 (en) 2012-08-16 2017-10-30 Method for modeling behavior and health changes
US15/807,323 Active US10068673B2 (en) 2012-08-16 2017-11-08 Method for modeling behavior and health changes
US16/057,199 Active 2035-03-21 US11195626B2 (en) 2012-08-16 2018-08-07 Method for modeling behavior and health changes

Country Status (5)

Country Link
US (7) US20140052465A1 (en)
EP (1) EP2884888A4 (en)
JP (1) JP6185066B2 (en)
CN (2) CN104584017B (en)
WO (1) WO2014028888A2 (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20230199112A1 (en) * 2021-12-22 2023-06-22 Optum, Inc. Vulnerable callee monitoring system

Families Citing this family (175)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2754385A1 (en) 2010-09-08 2014-07-16 Covidien LP Catheter with imaging assembly
WO2017091730A2 (en) 2015-11-24 2017-06-01 Dacadoo Ag Automated health data acquisition, processing and communication system and method
US10741285B2 (en) 2012-08-16 2020-08-11 Ginger.io, Inc. Method and system for providing automated conversations
US10276260B2 (en) 2012-08-16 2019-04-30 Ginger.io, Inc. Method for providing therapy to an individual
WO2014028888A2 (en) 2012-08-16 2014-02-20 Ginger.io, Inc. Method for modeling behavior and health changes
US10068060B2 (en) * 2012-08-16 2018-09-04 Ginger.io, Inc. Method for modeling behavior and psychotic disorders
US10068670B2 (en) * 2012-08-16 2018-09-04 Ginger.io, Inc. Method for modeling behavior and depression state
US10740438B2 (en) 2012-08-16 2020-08-11 Ginger.io, Inc. Method and system for characterizing and/or treating poor sleep behavior
JP2015534701A (en) 2012-08-28 2015-12-03 デロス リビング エルエルシーDelos Living Llc Systems, methods, and articles for promoting wellness associated with living environments
USD717340S1 (en) * 2012-09-07 2014-11-11 Covidien Lp Display screen with enteral feeding icon
US9517184B2 (en) 2012-09-07 2016-12-13 Covidien Lp Feeding tube with insufflation device and related methods therefor
USD735343S1 (en) 2012-09-07 2015-07-28 Covidien Lp Console
US9198835B2 (en) 2012-09-07 2015-12-01 Covidien Lp Catheter with imaging assembly with placement aid and related methods therefor
US20140100858A1 (en) * 2012-10-04 2014-04-10 Robert Bosch Gmbh System and Method for Identification of Risk Indicators Based on Delays in Answering Survey Questions
CN103778171A (en) * 2012-10-24 2014-05-07 I-昂索夫特有限公司 System for health media recommendation based on wellbeing-index
US10395766B2 (en) * 2013-06-28 2019-08-27 Hitachi, Ltd. Diagnostic process analysis system
WO2015003247A1 (en) * 2013-07-10 2015-01-15 1Datapoint Diagnostics Inc. Systems and methods relating to subject monitoring, assessment and treatment
US11631484B1 (en) * 2013-10-21 2023-04-18 Aic Innovations Group, Inc. Method and apparatus for predicting, encouraging, and intervening to improve patient medication adherence
US9420970B2 (en) 2013-10-22 2016-08-23 Mindstrong, LLC Method and system for assessment of cognitive function based on mobile device usage
US9474481B2 (en) 2013-10-22 2016-10-25 Mindstrong, LLC Method and system for assessment of cognitive function based on electronic device usage
US11315694B2 (en) * 2013-12-02 2022-04-26 Ceresti Health, Inc. System and method for reducing acute incident risk
US10049185B2 (en) * 2014-01-28 2018-08-14 3M Innovative Properties Company Perfoming analytics on protected health information
US20150242593A1 (en) * 2014-02-21 2015-08-27 MAP Health Management, LLC System and method for generating survey questions
EP3111411A4 (en) 2014-02-28 2017-08-09 Delos Living, LLC Systems, methods and articles for enhancing wellness associated with habitable environments
US10321870B2 (en) 2014-05-01 2019-06-18 Ramot At Tel-Aviv University Ltd. Method and system for behavioral monitoring
US9977875B2 (en) 2014-05-16 2018-05-22 Corcept Therapeutics, Inc. Systems and methods of managing treatment of a chronic condition by symptom tracking
US10073951B2 (en) * 2014-06-23 2018-09-11 GoodRx, Inc. Demographically filterable interface for conveying information about a medication
US10614073B2 (en) * 2014-06-26 2020-04-07 Financialsharp, Inc. System and method for using data incident based modeling and prediction
WO2016003974A1 (en) * 2014-06-30 2016-01-07 Hc1.Com, Inc. System and method for multi-tenant healthcare relationship management and wellness automation
JP6316689B2 (en) * 2014-07-15 2018-04-25 株式会社 国際疾病管理研究所 Information display apparatus and method, and computer program
EP3175420A4 (en) * 2014-08-01 2018-05-16 ResMed Limited Patient management system
US10153058B2 (en) 2014-09-11 2018-12-11 The Regents Of The University Of Michigan Machine learning for hepatitis C
US20160117465A1 (en) * 2014-10-22 2016-04-28 Xerox Corporation Method and apparatus for calculating an overall health quality index and providing a health upside optimizing recommendation
US10123748B2 (en) * 2014-10-30 2018-11-13 International Business Machines Corporation Active patient risk prediction
KR102323211B1 (en) * 2014-11-04 2021-11-08 삼성전자주식회사 Mobile health care device and operating method thereof
US20210121082A1 (en) * 2014-11-11 2021-04-29 Well Universal Pty Ltd Method and a processor for determining health of an individual
GB201420436D0 (en) * 2014-11-18 2014-12-31 Ixico Ltd Device
EP3038023A1 (en) * 2014-12-23 2016-06-29 Telefonica Digital España, S.L.U. A method, a system and computer program products for assessing the behavioral performance of a user
US20160196394A1 (en) * 2015-01-07 2016-07-07 Amino, Inc. Entity cohort discovery and entity profiling
CN107251031A (en) 2015-01-13 2017-10-13 戴尔斯生活有限责任公司 System, method and product for monitoring and strengthening health
US11461848B1 (en) 2015-01-14 2022-10-04 Alchemy Logic Systems, Inc. Methods of obtaining high accuracy impairment ratings and to assist data integrity in the impairment rating process
US10489554B2 (en) 2015-02-09 2019-11-26 Forge Laboratories, Llc Computer assisted patient navigation and information systems and methods
US10354051B2 (en) 2015-02-09 2019-07-16 Forge Laboratories, Llc Computer assisted patient navigation and information systems and methods
US20160262681A1 (en) * 2015-03-13 2016-09-15 At&T Intellectual Property I, L.P. Detecting depression via mobile device data
EP3292534A4 (en) * 2015-05-04 2018-12-26 Kontigo Care AB Method and device for estimating a risk of relapse of addictive behaviour
WO2016181490A1 (en) * 2015-05-12 2016-11-17 株式会社日立製作所 Analysis system and analysis method
US10560135B1 (en) 2015-06-05 2020-02-11 Life365, Inc. Health, wellness and activity monitor
US11329683B1 (en) 2015-06-05 2022-05-10 Life365, Inc. Device configured for functional diagnosis and updates
US10185513B1 (en) 2015-06-05 2019-01-22 Life365, Inc. Device configured for dynamic software change
US9974492B1 (en) 2015-06-05 2018-05-22 Life365, Inc. Health monitoring and communications device
CN108289107A (en) * 2015-06-06 2018-07-17 深圳市前海安测信息技术有限公司 On-line off-line health control method, data terminal and server Internet-based
US10755812B1 (en) * 2015-07-31 2020-08-25 Iqvia Inc. System and method for timely notification of treatment
US20170061091A1 (en) * 2015-08-26 2017-03-02 Uptake Technologies, Inc. Indication of Outreach Options for Healthcare Facility to Facilitate Patient Actions
EP3350806A4 (en) * 2015-09-14 2019-08-07 Cogito Corporation Systems and methods for identifying human emotions and/or mental health states based on analyses of audio inputs and/or behavioral data collected from computing devices
CA3001350A1 (en) * 2015-10-06 2017-04-13 Raymond A. Berardinelli Smartwatch device and method
CN105232063B (en) * 2015-10-22 2017-03-22 广东小天才科技有限公司 Detection method for mental health of user and intelligent terminal
CN109155158A (en) * 2015-11-05 2019-01-04 360膝盖系统股份有限公司 Manage the patient of knee surgery
WO2017079047A1 (en) * 2015-11-06 2017-05-11 3M Innovative Properties Company Identification of low-efficacy patient population
US20170213006A1 (en) * 2016-01-27 2017-07-27 The Live Network Inc. Intelligent mobile homework adherence and feedback application for telehealth
AU2017218522A1 (en) 2016-02-08 2018-08-23 OutcomeMD, Inc. Determining a wellness, improvement, or effectiveness score
US11387000B2 (en) * 2016-02-08 2022-07-12 OutcomeMD, Inc. Systems and methods for determining and providing a display of a plurality of wellness scores for patients with regard to a medical condition and/or a medical treatment
US10420514B2 (en) 2016-02-25 2019-09-24 Samsung Electronics Co., Ltd. Detection of chronotropic incompetence
US11164596B2 (en) 2016-02-25 2021-11-02 Samsung Electronics Co., Ltd. Sensor assisted evaluation of health and rehabilitation
US20170249437A1 (en) * 2016-02-25 2017-08-31 Samsung Electronics Co., Ltd. Sensor assisted depression detection
US10362998B2 (en) 2016-02-25 2019-07-30 Samsung Electronics Co., Ltd. Sensor-based detection of changes in health and ventilation threshold
US20170286618A1 (en) * 2016-04-01 2017-10-05 Medisked, LLC. Support for linked individual care plans across multiple care providers
US20190148002A1 (en) * 2016-04-16 2019-05-16 Nancy NUNES System for affecting behavior of a subject
US10586614B1 (en) * 2016-04-22 2020-03-10 Iqvia Inc. System and method for timely multi-channel notification of treatment
EP3452932A1 (en) * 2016-05-04 2019-03-13 Koninklijke Philips N.V. Estimation and use of clinician assessment of patient acuity
KR102236614B1 (en) * 2016-05-11 2021-04-09 주식회사 칼라세븐 Smart communication device based low level light therapy service offer system and method
WO2017196052A1 (en) * 2016-05-11 2017-11-16 주식회사 칼라세븐 System and method for providing low-output light therapy service based on smart communication device
US11853973B1 (en) * 2016-07-26 2023-12-26 Alchemy Logic Systems, Inc. Method of and system for executing an impairment repair process
JP2018015327A (en) * 2016-07-28 2018-02-01 株式会社リコー Diagnostic apparatus, diagnostic method, and program
US20180032126A1 (en) * 2016-08-01 2018-02-01 Yadong Liu Method and system for measuring emotional state
WO2018045152A1 (en) * 2016-09-02 2018-03-08 Lumme Inc Systems and methods for health monitoring
CN106407676A (en) * 2016-09-12 2017-02-15 广东省农业科学院动物科学研究所 Monte Carlo simulation based drug effect calculating method for antibacterial drug
CN106355033A (en) * 2016-09-27 2017-01-25 无锡金世纪国民体质与健康研究有限公司 Life risk assessment system
CN106530375B (en) * 2016-09-28 2019-03-29 山东师范大学 A kind of crowd's animation producing method of personalization emotional appeal
JP6864460B2 (en) * 2016-11-02 2021-04-28 花王株式会社 Disease prediction device
US11854700B1 (en) 2016-12-06 2023-12-26 Alchemy Logic Systems, Inc. Method of and system for determining a highly accurate and objective maximum medical improvement status and dating assignment
US11188620B1 (en) * 2016-12-16 2021-11-30 Iqvia Inc. System and method to improve dynamic multi-channel information synthesis
CA3053245A1 (en) * 2017-02-09 2018-08-16 Cognoa, Inc. Platform and system for digital personalized medicine
CN110494919B (en) * 2017-02-15 2023-11-24 赫美特里克斯 Method for managing healthcare services by using a therapy management system
US10674476B1 (en) * 2017-03-10 2020-06-02 Wells Fargo Bank, N.A. Contextual aware electronic alert system
CN107016457A (en) * 2017-03-14 2017-08-04 上海市精神卫生中心 One kind realizes community's hazardous act pre-warning system and method
EP3382584A1 (en) * 2017-03-30 2018-10-03 Fujitsu Limited A system and a method to predict patient behaviour
US11253195B2 (en) 2017-04-26 2022-02-22 Morebetter, Ltd. Method and system for capturing patient feedback for a medical treatment
US10862927B2 (en) 2017-05-15 2020-12-08 Forcepoint, LLC Dividing events into sessions during adaptive trust profile operations
US10447718B2 (en) 2017-05-15 2019-10-15 Forcepoint Llc User profile definition and management
US10129269B1 (en) 2017-05-15 2018-11-13 Forcepoint, LLC Managing blockchain access to user profile information
US10999297B2 (en) 2017-05-15 2021-05-04 Forcepoint, LLC Using expected behavior of an entity when prepopulating an adaptive trust profile
US10999296B2 (en) 2017-05-15 2021-05-04 Forcepoint, LLC Generating adaptive trust profiles using information derived from similarly situated organizations
US10917423B2 (en) 2017-05-15 2021-02-09 Forcepoint, LLC Intelligently differentiating between different types of states and attributes when using an adaptive trust profile
US9882918B1 (en) 2017-05-15 2018-01-30 Forcepoint, LLC User behavior profile in a blockchain
US10943019B2 (en) 2017-05-15 2021-03-09 Forcepoint, LLC Adaptive trust profile endpoint
CN107478917B (en) * 2017-07-17 2019-10-29 国网江西省电力公司电力科学研究院 The determination method and device of a kind of area's degree of unbalancedness
US11668481B2 (en) 2017-08-30 2023-06-06 Delos Living Llc Systems, methods and articles for assessing and/or improving health and well-being
US11687800B2 (en) 2017-08-30 2023-06-27 P Tech, Llc Artificial intelligence and/or virtual reality for activity optimization/personalization
FI20175793A1 (en) * 2017-09-06 2019-03-07 Klinfys Oy Arrangement and method for prediction of data related to health conditions
US10497250B1 (en) 2017-09-27 2019-12-03 State Farm Mutual Automobile Insurance Company Real property monitoring systems and methods for detecting damage and other conditions
US10916344B2 (en) 2017-11-22 2021-02-09 Accenture Global Solutions Limited Utilizing a machine learning model to identify activities and deviations from the activities by an individual
EP3721435A4 (en) * 2017-12-04 2021-08-25 Kontigo Care AB Method and device for estimating a risk of relapse of addictive behavior
US11551815B2 (en) * 2017-12-12 2023-01-10 Medical Informatics Corp. Risk monitoring scores
US20190189288A1 (en) * 2017-12-18 2019-06-20 Koninklijke Philips N.V. Providing subject-specific information
US11495332B2 (en) 2017-12-28 2022-11-08 International Business Machines Corporation Automated prediction and answering of medical professional questions directed to patient based on EMR
US11031109B2 (en) * 2017-12-28 2021-06-08 International Business Machines Corporation Contextual EMR based dashboard graphical user interface elements
US11203314B2 (en) 2018-01-05 2021-12-21 Anagog Ltd. Automatic accident detection
CN108122613B (en) * 2018-01-15 2022-04-01 北京颐圣智能科技有限公司 Health prediction method and device based on health prediction model
CN111630606A (en) * 2018-01-16 2020-09-04 皇家飞利浦有限公司 Detecting recurrence of a medical condition
CN108511056A (en) * 2018-02-09 2018-09-07 上海长江科技发展有限公司 Therapeutic scheme based on patients with cerebral apoplexy similarity analysis recommends method and system
US20190252063A1 (en) * 2018-02-14 2019-08-15 International Business Machines Corporation Monitoring system for care provider
WO2019158496A1 (en) * 2018-02-19 2019-08-22 Koninklijke Philips N.V. System and method for providing model-based population insight generation
US20190290128A1 (en) * 2018-03-20 2019-09-26 Aic Innovations Group, Inc. Apparatus and method for user evaluation
JP2019164634A (en) 2018-03-20 2019-09-26 カシオ計算機株式会社 Wearable device, health management support method, and health management support program
US10937531B1 (en) 2018-04-09 2021-03-02 Iqvia Inc. System and method for timely notification of treatments to healthcare providers and patient
CN108735295B (en) * 2018-05-14 2024-01-26 平安科技(深圳)有限公司 Blood analysis method and terminal equipment based on regression tree model
EP3582225A1 (en) 2018-06-14 2019-12-18 Koninklijke Philips N.V. Monitoring a subject
WO2019246239A1 (en) 2018-06-19 2019-12-26 Ellipsis Health, Inc. Systems and methods for mental health assessment
US20190385711A1 (en) 2018-06-19 2019-12-19 Ellipsis Health, Inc. Systems and methods for mental health assessment
WO2020002284A1 (en) * 2018-06-29 2020-01-02 Koninklijke Philips N.V. Determining adherence to a medication dosage scheme
EP3624125A1 (en) * 2018-09-13 2020-03-18 Koninklijke Philips N.V. Determining adherence to a medication dosage scheme
US11195618B2 (en) * 2018-07-12 2021-12-07 International Business Machines Corporation Multi-level machine learning to detect a social media user's possible health issue
WO2020030597A1 (en) * 2018-08-06 2020-02-13 Grünenthal GmbH Method for carrying out structured rehabilitation
EP3850458A4 (en) 2018-09-14 2022-06-08 Delos Living, LLC Systems and methods for air remediation
US20210358628A1 (en) * 2018-10-09 2021-11-18 Valotec Digital companion for healthcare
US11625687B1 (en) 2018-10-16 2023-04-11 Alchemy Logic Systems Inc. Method of and system for parity repair for functional limitation determination and injury profile reports in worker's compensation cases
WO2020102175A1 (en) * 2018-11-12 2020-05-22 F. Hoffman-La Roche Ag Medical treatment metric modelling based on machine learning
FR3089331B1 (en) * 2018-11-29 2020-12-25 Veyron Jacques Henri Data processing system and method for determining the risk of an individual's emergency visit
US11309082B2 (en) 2018-12-10 2022-04-19 Groop Internet Platform, Inc. System and method for monitoring engagement
US11783948B2 (en) * 2018-12-17 2023-10-10 International Business Machines Corporation Cognitive evaluation determined from social interactions
US11694786B1 (en) * 2019-01-04 2023-07-04 Yasmine Van Wilt Recommendation methods, systems and devices
CN109658772B (en) * 2019-02-11 2021-01-26 三峡大学 Operation training and checking method based on virtual reality
JP2022521020A (en) 2019-02-22 2022-04-04 ペア セラピューティクス インコーポレイテッド Systems and methods for the treatment of symptoms associated with migraine
WO2020176503A1 (en) 2019-02-26 2020-09-03 Delos Living Llc Method and apparatus for lighting in an office environment
MX2021011493A (en) 2019-03-22 2022-01-04 Cognoa Inc Personalized digital therapy methods and devices.
WO2020198183A1 (en) 2019-03-25 2020-10-01 Delos Living Llc Systems and methods for acoustic monitoring
US11514339B2 (en) * 2019-04-24 2022-11-29 Optum, Inc. Machine-learning based recommendation engine providing transparency in generation of recommendations
US10853496B2 (en) 2019-04-26 2020-12-01 Forcepoint, LLC Adaptive trust profile behavioral fingerprint
TW202119301A (en) * 2019-06-17 2021-05-16 日商大塚製藥股份有限公司 A dynamically distorted time warping distance measure between continuous bounded discrete-time series
US11894129B1 (en) 2019-07-03 2024-02-06 State Farm Mutual Automobile Insurance Company Senior living care coordination platforms
US11848109B1 (en) 2019-07-29 2023-12-19 Alchemy Logic Systems, Inc. System and method of determining financial loss for worker's compensation injury claims
US11676727B2 (en) 2019-08-14 2023-06-13 Optum Technology, Inc. Cohort-based predictive data analysis
WO2021034677A1 (en) * 2019-08-16 2021-02-25 OptimDosing, LLC Application for tracking progression and isolating causes of adverse medical conditions
US11367527B1 (en) 2019-08-19 2022-06-21 State Farm Mutual Automobile Insurance Company Senior living engagement and care support platforms
US11581093B2 (en) 2019-09-19 2023-02-14 Merative Us L.P. Automatic detection of mental health condition and patient classification using machine learning
US11521724B2 (en) * 2019-10-04 2022-12-06 International Business Machines Corporation Personalized patient engagement in care management using explainable behavioral phenotypes
CA3154229A1 (en) * 2019-10-09 2021-04-15 Affinity Recovery Management Services Inc. System and method for monitoring system compliance with measures to improve system health
US20220399122A1 (en) * 2019-11-01 2022-12-15 Nec Corporation Risk prediction apparatus, risk prediction method, and computer program
CN110853760A (en) * 2019-11-12 2020-02-28 好人生(上海)健康科技有限公司 Interactive system for providing health guidance based on disease diagnosis
JP2023504398A (en) * 2019-11-26 2023-02-03 デックスコム・インコーポレーテッド Suggestions based on continuous blood glucose monitoring
US20210183481A1 (en) * 2019-12-17 2021-06-17 Mahana Therapeutics, Inc. Method and system for remotely monitoring the psychological state of an application user based on average user interaction data
EP3843103A1 (en) * 2019-12-25 2021-06-30 Softimize Ltd. Systems and methods for increasing adherence for medical devices
US11276501B1 (en) 2020-02-24 2022-03-15 BlueOwl, LLC Systems and methods for determining a risk score using machine learning based at least in part upon collected sensor data
AU2021226736A1 (en) * 2020-02-25 2022-09-01 Becton, Dickinson And Company Systems and methods for dynamic surveillance of medication and antimicrobial resistance trends
US11823782B2 (en) 2020-03-03 2023-11-21 Optum, Inc. Systems, methods, apparatuses, and computer program products for administering pain management solutions remotely
CA3174390A1 (en) 2020-04-03 2021-10-07 Aifred Health Systems and methods for treatment selection
US11587682B2 (en) 2020-05-15 2023-02-21 Medable Inc. Method and system to integrate data, analyze and develop improved care plan for a patient at home
US20210361175A1 (en) * 2020-05-20 2021-11-25 Koninklijke Philips N.V. Sleep-based biometric to predict and track viral infection phases
US20210374873A1 (en) * 2020-05-29 2021-12-02 New Directions Behavioral Health, L.L.C. System and method for case management risk stratification
US11670421B2 (en) * 2020-06-01 2023-06-06 Medable Inc. Method and system enabling digital biomarker data integration and analysis for clinical treatment impact
WO2021245728A1 (en) * 2020-06-01 2021-12-09 サスメド株式会社 Disease estimation system, teacher data acquisition device, and program for acquiring teacher data
US20240006052A1 (en) * 2020-09-25 2024-01-04 Cornell University System implementing generative adversarial network adapted to prediction in behavioral and/or physiological contexts
US11763919B1 (en) 2020-10-13 2023-09-19 Vignet Incorporated Platform to increase patient engagement in clinical trials through surveys presented on mobile devices
CN112489815A (en) * 2020-11-17 2021-03-12 中国电子科技集团公司电子科学研究院 Depression emotion monitoring method and device and readable storage medium
CN112331337B (en) * 2021-01-04 2021-04-16 中国科学院自动化研究所 Automatic depression detection method, device and equipment
US11688516B2 (en) 2021-01-19 2023-06-27 State Farm Mutual Automobile Insurance Company Alert systems for senior living engagement and care support platforms
US20220351822A1 (en) * 2021-05-03 2022-11-03 Intuition Robotics, Ltd. Techniques for executing and modifying transient care plans via an input/output device
TWI787806B (en) * 2021-05-06 2022-12-21 許弘毅 Method for risk assessment of neurological disorder and electronic device using the same
US11710576B2 (en) 2021-05-24 2023-07-25 OrangeDot, Inc. Method and system for computer-aided escalation in a digital health platform
EP4113535A1 (en) * 2021-06-29 2023-01-04 Siemens Healthcare GmbH Remote monitoring methods and systems for monitoring patients suffering from chronical inflammatory diseases
US20230091240A1 (en) * 2021-09-23 2023-03-23 Behavidence Inc. System and method of monitoring mental health conditions
CN114446427A (en) * 2021-12-28 2022-05-06 海信集团控股股份有限公司 Electronic equipment and health data attribution identification method
WO2023170441A1 (en) * 2022-03-07 2023-09-14 Diamond Therapeutics Inc. Systems and methods for affecting well-being
CN114969557B (en) * 2022-07-29 2022-11-08 之江实验室 Propaganda and education pushing method and system based on multi-source information fusion
JP7324979B1 (en) * 2023-03-28 2023-08-14 株式会社Y4.com Information processing equipment

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070094048A1 (en) * 2005-10-25 2007-04-26 Caterpillar Inc. Expert knowledge combination process based medical risk stratifying method and system
US20110118555A1 (en) * 2009-04-29 2011-05-19 Abhijit Dhumne System and methods for screening, treating, and monitoring psychological conditions
US20130297536A1 (en) * 2012-05-01 2013-11-07 Bernie Almosni Mental health digital behavior monitoring support system and method

Family Cites Families (109)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US96634A (en) 1869-11-09 Improved enamel or glaze for pottery, brick, tiles
US4845323A (en) 1987-08-28 1989-07-04 Tactilitics, Inc. Flexible tactile switch
ATE250229T1 (en) 1998-07-08 2003-10-15 Univ Manchester SAMPLE ANALYSIS
US6884792B2 (en) 1999-01-08 2005-04-26 Marvin B. Bacaner Bretylium compositions and kits and their use in preventing and treating cardiovascular conditions
US6356940B1 (en) 1999-05-26 2002-03-12 Brian Robert Short Method and system of electronically logging remote user dietary information, and generating and automatically sending suggested dietary modifications
US7395216B2 (en) 1999-06-23 2008-07-01 Visicu, Inc. Using predictive models to continuously update a treatment plan for a patient in a health care location
US7376700B1 (en) 1999-08-23 2008-05-20 Wellcoaches Corporation Personal coaching system for clients with ongoing concerns such as weight loss
US6827670B1 (en) 1999-10-11 2004-12-07 Izex Technologies, Inc. System for medical protocol management
US7630986B1 (en) 1999-10-27 2009-12-08 Pinpoint, Incorporated Secure data interchange
US20010027384A1 (en) 2000-03-01 2001-10-04 Schulze Arthur E. Wireless internet bio-telemetry monitoring system and method
EP1275042A2 (en) 2000-03-06 2003-01-15 Kanisa Inc. A system and method for providing an intelligent multi-step dialog with a user
US7818185B2 (en) 2000-06-02 2010-10-19 Qualitymetric Incorporated Method, system and medium for assessing the impact of various ailments on health related quality of life
WO2002005622A1 (en) 2000-07-17 2002-01-24 Burley Blaine F Agricultural earth working apparatus
WO2002017602A1 (en) 2000-08-22 2002-02-28 Symbian Limited Method of and apparatus for communicating user related information using a wireless information device
US8488761B2 (en) 2001-02-27 2013-07-16 Verizon Data Services Llc Methods and systems for a call log
GB2373218B (en) 2001-03-13 2004-12-15 Autoliv Dev Improvements in or relating to a safety arrangement
AU2002254463A1 (en) 2001-03-28 2002-10-15 Televital, Inc. Real-time monitoring assessment, analysis, retrieval, and storage of physiological data
US7328842B2 (en) 2001-08-14 2008-02-12 Ikan Technologies Inc. Networked waste processing apparatus
US20110295621A1 (en) 2001-11-02 2011-12-01 Siemens Medical Solutions Usa, Inc. Healthcare Information Technology System for Predicting and Preventing Adverse Events
US7761309B2 (en) * 2002-03-22 2010-07-20 Thinksharp, Inc. Method and system of mass and multiple casualty triage
JP3754394B2 (en) 2002-05-29 2006-03-08 横河電機株式会社 Sleep stage estimation apparatus and apparatus using signals output from sleep stage estimation apparatus
US20040143453A1 (en) 2002-12-17 2004-07-22 International Business Machines Corporation Behavior based life support with default behavior patterns
US20040122702A1 (en) 2002-12-18 2004-06-24 Sabol John M. Medical data processing system and method
US20040122703A1 (en) 2002-12-19 2004-06-24 Walker Matthew J. Medical data operating model development system and method
US20040225340A1 (en) 2003-03-10 2004-11-11 Evans James W. Light/breath/meditation device
US7706871B2 (en) * 2003-05-06 2010-04-27 Nellcor Puritan Bennett Llc System and method of prediction of response to neurological treatment using the electroencephalogram
US7912528B2 (en) 2003-06-25 2011-03-22 Siemens Medical Solutions Usa, Inc. Systems and methods for automated diagnosis and decision support for heart related diseases and conditions
US8068917B2 (en) 2003-08-29 2011-11-29 Medtronic, Inc. Fail-safe programming for implantable medical device
US20050069936A1 (en) 2003-09-26 2005-03-31 Cornelius Diamond Diagnostic markers of depression treatment and methods of use thereof
US20050108051A1 (en) * 2003-11-07 2005-05-19 Andrew Weinstein Method of promoting patient adherence to a treatment regimen
US7914468B2 (en) 2004-09-22 2011-03-29 Svip 4 Llc Systems and methods for monitoring and modifying behavior
JP4781710B2 (en) * 2005-05-12 2011-09-28 シスメックス株式会社 Treatment effect prediction system and program thereof
US7995717B2 (en) 2005-05-18 2011-08-09 Mattersight Corporation Method and system for analyzing separated voice data of a telephonic communication between a customer and a contact center by applying a psychological behavioral model thereto
EP1747855A1 (en) 2005-07-28 2007-01-31 Felco S.A. Assembly of two parts, assembly kit and process to realize an assembly
US8566121B2 (en) * 2005-08-29 2013-10-22 Narayanan Ramasubramanian Personalized medical adherence management system
US8055707B2 (en) 2005-11-30 2011-11-08 Alcatel Lucent Calendar interface for digital communications
US20070226012A1 (en) 2005-12-13 2007-09-27 Naryx Pharma, Inc. Methods of measuring symptoms of chronic rhinosinusitis
US8425583B2 (en) 2006-04-20 2013-04-23 University of Pittsburgh—of the Commonwealth System of Higher Education Methods, devices and systems for treating insomnia by inducing frontal cerebral hypothermia
US8684922B2 (en) 2006-05-12 2014-04-01 Bao Tran Health monitoring system
US20070288266A1 (en) 2006-06-02 2007-12-13 Suzanne Sysko System and methods for chronic disease management and health assessment
AU2007269784B2 (en) 2006-06-30 2012-07-05 Light Sciences Oncology, Inc. Compositions and methods of making a chlorin e6 derivative as a photoactive agent
US8726195B2 (en) 2006-09-05 2014-05-13 Aol Inc. Enabling an IM user to navigate a virtual world
US7728730B2 (en) 2006-09-06 2010-06-01 Casella Waste Systems, Inc Systems and methods for measuring the purity of bales of recyclable materials
US8768718B2 (en) * 2006-12-27 2014-07-01 Cardiac Pacemakers, Inc. Between-patient comparisons for risk stratification of future heart failure decompensation
WO2008096634A1 (en) * 2007-02-06 2008-08-14 Nec Corporation Health management system, health managing method, and health management program
US20080201429A1 (en) 2007-02-16 2008-08-21 Siemens Medical Solutions Usa, Inc. System Supporting Live Electronic Messaging Communication Between A Healthcare Worker And A Patient
US8447285B1 (en) 2007-03-26 2013-05-21 Callwave Communications, Llc Methods and systems for managing telecommunications and for translating voice messages to text messages
WO2008142611A2 (en) * 2007-05-16 2008-11-27 Koninklijke Philips Electronics, N.V. Apparatus and methods for medical patient role playing / simulation activity
US8257612B2 (en) 2007-07-05 2012-09-04 Cabot Corporation Compositions having multiple responses to excitation radiation and methods for making same
JP5425793B2 (en) 2007-10-12 2014-02-26 ペイシェンツライクミー, インコーポレイテッド Personal management and comparison of medical conditions and outcomes based on patient community profiles
US7999688B2 (en) 2007-12-20 2011-08-16 Intel-Ge Care Innovations Llc Embedded assessment of refuse for activity monitoring
AU2009217184B2 (en) 2008-02-20 2015-03-19 Digital Medical Experts Inc. Expert system for determining patient treatment response
US20090216558A1 (en) * 2008-02-27 2009-08-27 Active Health Management Inc. System and method for generating real-time health care alerts
KR101806432B1 (en) 2008-03-26 2017-12-07 테라노스, 인코포레이티드 Methods and systems for assessing clinical outcomes
WO2010005877A2 (en) * 2008-07-08 2010-01-14 Proteus Biomedical, Inc. Ingestible event marker data framework
CA2680952A1 (en) * 2008-10-01 2010-04-01 Loyaltyone Us, Inc. System and method for providing a health management program
US20100179833A1 (en) 2009-01-12 2010-07-15 Roizen Michael F Automated coaching
US20100198614A1 (en) 2009-01-30 2010-08-05 The Regents Of The University Of Michigan Medical communication system for health care practitioners
US20100203876A1 (en) * 2009-02-11 2010-08-12 Qualcomm Incorporated Inferring user profile properties based upon mobile device usage
US20120016218A1 (en) 2009-04-20 2012-01-19 Resmed Limited Discrimination of cheyne-stokes breathing patterns by use of oximetry signals
US20100280838A1 (en) 2009-05-01 2010-11-04 Adam Bosworth Coaching Engine for a Health Coaching Service
CN101600008A (en) 2009-05-25 2009-12-09 成都途筏达科技有限公司 A kind of electrocardio cellphone of intelligence
CN101658414A (en) * 2009-06-19 2010-03-03 中卫莱康科技发展(北京)有限公司 System and method for monitoring personal health
US9215212B2 (en) 2009-06-22 2015-12-15 Citrix Systems, Inc. Systems and methods for providing a visualizer for rules of an application firewall
GB2471903A (en) 2009-07-17 2011-01-19 Sharp Kk Sleep management system for monitoring sleep quality and making recommendations for improvement
US8500635B2 (en) * 2009-09-17 2013-08-06 Blife Inc. Mobile system and method for addressing symptoms related to mental health conditions
US11562323B2 (en) 2009-10-01 2023-01-24 DecisionQ Corporation Application of bayesian networks to patient screening and treatment
US20110161110A1 (en) 2009-10-06 2011-06-30 Mault James R System And Method For An Online Platform Distributing Condition Specific Programs Used For Monitoring The Health Of A Participant And For Offering Health Services To Participating Subscribers
CN101710371A (en) * 2009-12-25 2010-05-19 张立彬 Technology of human health condition holographic acquisition and computer-aided diagnosis and health risk assessment
CN102971755A (en) * 2010-01-21 2013-03-13 阿斯玛西格诺斯公司 Early warning method and system for chronic disease management
KR20130051922A (en) 2010-03-04 2013-05-21 뉴미트라 엘엘씨 Devices and methods for treating psychological disorders
US9872637B2 (en) 2010-04-21 2018-01-23 The Rehabilitation Institute Of Chicago Medical evaluation system and method using sensors in mobile devices
US20110306028A1 (en) 2010-06-15 2011-12-15 Galimore Sarah E Educational decision support system and associated methods
EP2609533A2 (en) 2010-08-27 2013-07-03 Smartex S.r.l. Monitoring method and system for assessment or prediction of mood trends
US11096577B2 (en) 2010-11-03 2021-08-24 Nxgn Management, Llc Proactive patient health care inference engines and systems
US9154564B2 (en) 2010-11-18 2015-10-06 Qualcomm Incorporated Interacting with a subscriber to a social networking service based on passive behavior of the subscriber
US8744979B2 (en) * 2010-12-06 2014-06-03 Microsoft Corporation Electronic communications triage using recipient's historical behavioral and feedback
US8676607B2 (en) 2011-01-10 2014-03-18 Medimpact Healthcare Systems, Inc. Obtaining patient survey results
KR101212714B1 (en) 2011-01-26 2013-01-22 계명대학교 산학협력단 Clinical diagnosis support method based on hierarchical fuzzy inference technology, and system
US8660864B2 (en) 2011-02-28 2014-02-25 Hartford Fire Insurance Company Systems and methods for intelligent underwriting based on community or social network data
US8617067B2 (en) 2011-05-13 2013-12-31 Fujitsu Limited Continuous monitoring of stress using environmental data
US8905648B2 (en) 2011-06-29 2014-12-09 Cinch Connectivity Solutions, Inc. Expanded beam fiber optic connector
US8423387B1 (en) 2011-07-11 2013-04-16 Muhammad Mirza Direct physician delivery of patient care over a network
JP2013042315A (en) 2011-08-12 2013-02-28 Sony Corp Information processing device and information processing method
US9495331B2 (en) 2011-09-19 2016-11-15 Personetics Technologies Ltd. Advanced system and method for automated-context-aware-dialog with human users
EP2575064A1 (en) 2011-09-30 2013-04-03 General Electric Company Telecare and/or telehealth communication method and system
US9934361B2 (en) 2011-09-30 2018-04-03 Univfy Inc. Method for generating healthcare-related validated prediction models from multiple sources
US20130117040A1 (en) 2011-11-03 2013-05-09 Omada Health, Inc. Method and System for Supporting a Health Regimen
US9019106B2 (en) 2011-12-15 2015-04-28 Google Technology Holdings LLC Adaptive wearable device for controlling an alarm based on user sleep state
US20130179178A1 (en) 2012-01-06 2013-07-11 Active Health Management System and method for patient care plan management
KR101248587B1 (en) 2012-04-18 2013-04-02 조희진 Utility iron
US9763592B2 (en) 2012-05-25 2017-09-19 Emotiv, Inc. System and method for instructing a behavior change in a user
JP6075973B2 (en) 2012-06-04 2017-02-08 富士通株式会社 HEALTH STATE JUDGING DEVICE AND ITS OPERATION METHOD
US10593000B2 (en) 2012-07-13 2020-03-17 Koninklijke Philips N.V. System and method for determining thresholds or a range of values used to allocate patients to a treatment level of a treatment program
US20140039914A1 (en) 2012-08-02 2014-02-06 Dana Dansereau System and method for promoting wellness
US20140039907A1 (en) 2012-08-03 2014-02-06 AxelaCare Health Solutions, Inc. Computer program, method, and system for collecting patient data with a portable electronic device
WO2014028888A2 (en) 2012-08-16 2014-02-20 Ginger.io, Inc. Method for modeling behavior and health changes
US10068060B2 (en) 2012-08-16 2018-09-04 Ginger.io, Inc. Method for modeling behavior and psychotic disorders
US10068670B2 (en) * 2012-08-16 2018-09-04 Ginger.io, Inc. Method for modeling behavior and depression state
US20140257439A1 (en) 2013-03-06 2014-09-11 Verilux, Inc. Adjustable therapeutic lights
US9514490B2 (en) 2013-06-13 2016-12-06 Elbex Video Ltd. Method for propagating combination of signals for operating a closed circuit e-commerce
US9999744B2 (en) 2013-06-21 2018-06-19 Fitbit, Inc. Monitoring device and cognitive behavior therapy
EP3014821A4 (en) 2013-06-28 2017-02-22 Intel Corporation Mechanism to control resource utilization with adaptive routing
WO2015003247A1 (en) 2013-07-10 2015-01-15 1Datapoint Diagnostics Inc. Systems and methods relating to subject monitoring, assessment and treatment
TWI551267B (en) 2013-12-30 2016-10-01 瑞軒科技股份有限公司 Sleep aid system and operation method thereof
US9875903B2 (en) 2014-04-25 2018-01-23 Conduent Business Services, Llc Store resource event detection
US9342884B2 (en) 2014-05-28 2016-05-17 Cox Enterprises, Inc. Systems and methods of monitoring waste
WO2015193900A1 (en) 2014-06-19 2015-12-23 Shlomo Uri Haimi Recycling system
US11216774B2 (en) 2017-06-22 2022-01-04 Beneship Llc Systems and methods for packing optimization and visualization

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070094048A1 (en) * 2005-10-25 2007-04-26 Caterpillar Inc. Expert knowledge combination process based medical risk stratifying method and system
US20110118555A1 (en) * 2009-04-29 2011-05-19 Abhijit Dhumne System and methods for screening, treating, and monitoring psychological conditions
US20130297536A1 (en) * 2012-05-01 2013-11-07 Bernie Almosni Mental health digital behavior monitoring support system and method

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20230199112A1 (en) * 2021-12-22 2023-06-22 Optum, Inc. Vulnerable callee monitoring system

Also Published As

Publication number Publication date
EP2884888A4 (en) 2016-04-20
US20180342327A1 (en) 2018-11-29
US11195626B2 (en) 2021-12-07
CN104584017B (en) 2019-02-15
US9836581B2 (en) 2017-12-05
US20140052475A1 (en) 2014-02-20
US10068673B2 (en) 2018-09-04
WO2014028888A2 (en) 2014-02-20
CN109785972B (en) 2023-09-26
US20140052465A1 (en) 2014-02-20
US20140052474A1 (en) 2014-02-20
US20180068085A1 (en) 2018-03-08
CN109785972A (en) 2019-05-21
EP2884888A2 (en) 2015-06-24
JP6185066B2 (en) 2017-08-23
US20180052974A1 (en) 2018-02-22
CN104584017A (en) 2015-04-29
JP2015529359A (en) 2015-10-05
US10068672B2 (en) 2018-09-04
WO2014028888A3 (en) 2014-04-10

Similar Documents

Publication Publication Date Title
US20220059236A1 (en) Method for modeling behavior and health changes
US11769576B2 (en) Method and system for improving care determination
US20200402626A1 (en) Method for providing patient indications to an entity
US11200984B2 (en) Method for modeling behavior and psychotic disorders
US20190147995A1 (en) Systems and methods for clinical decision-making
US10265028B2 (en) Method and system for modeling behavior and heart disease state
TWI557679B (en) System and method for generating real-time health care alerts
US20200020452A1 (en) Computer-implemented system and method for identifying similar patients
US20130110551A1 (en) Systems and methods for managing chronic conditions
US10628047B2 (en) System and method for minimizing computational resources when copying data for a well-being assessment and scoring
US20200402630A1 (en) Systems and methods for clinical decision-making
KR20230113690A (en) treatment recommendations
JP2023504197A (en) Digital therapy system and method
US11185282B2 (en) System and method for monitoring and identifying posology efficacy for an an individual
WO2021113760A1 (en) Digital therapeutic systems and methods
CA3228976A1 (en) Ai based methods and systems for tracking chronic conditions

Legal Events

Date Code Title Description
AS Assignment

Owner name: GINGER.IO, INC., CALIFORNIA

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:MADAN, ANMOL;MOTURU, SAI THEJASVEE;SIGNING DATES FROM 20130826 TO 20130904;REEL/FRAME:058008/0166

STPP Information on status: patent application and granting procedure in general

Free format text: DOCKETED NEW CASE - READY FOR EXAMINATION

AS Assignment

Owner name: PACIFIC WESTERN BANK, NORTH CAROLINA

Free format text: SECURITY INTEREST;ASSIGNOR:GINGER.IO, INC.;REEL/FRAME:059697/0901

Effective date: 20220408

STPP Information on status: patent application and granting procedure in general

Free format text: NON FINAL ACTION MAILED

AS Assignment

Owner name: ORANGEDOT, INC., CALIFORNIA

Free format text: MERGER AND CHANGE OF NAME;ASSIGNORS:ORANGEDOT, INC.;SPICER MERGER SUB, INC.;GINGER.IO, INC.;REEL/FRAME:063910/0219

Effective date: 20210824

AS Assignment

Owner name: GINGER.IO, INC., CALIFORNIA

Free format text: RELEASE BY SECURED PARTY;ASSIGNOR:PACIFIC WESTERN BANK;REEL/FRAME:064084/0631

Effective date: 20230623

Owner name: OXFORD FINANCE LLC, VIRGINIA

Free format text: SECURITY INTEREST;ASSIGNOR:GINGER.IO, INC.;REEL/FRAME:064067/0646

Effective date: 20230623

STCB Information on status: application discontinuation

Free format text: ABANDONED -- FAILURE TO RESPOND TO AN OFFICE ACTION