US20060129427A1 - Systems and methods for predicting healthcare related risk events - Google Patents

Systems and methods for predicting healthcare related risk events Download PDF

Info

Publication number
US20060129427A1
US20060129427A1 US11280611 US28061105A US2006129427A1 US 20060129427 A1 US20060129427 A1 US 20060129427A1 US 11280611 US11280611 US 11280611 US 28061105 A US28061105 A US 28061105A US 2006129427 A1 US2006129427 A1 US 2006129427A1
Authority
US
Grant status
Application
Patent type
Prior art keywords
data
risk
patient
system
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
US11280611
Inventor
David Wennberg
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.)
Health Dialog Services Corp
Original Assignee
Health Dialog Services Corp
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

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F19/00Digital computing or data processing equipment or methods, specially adapted for specific applications
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F19/00Digital computing or data processing equipment or methods, specially adapted for specific applications
    • G06F19/30Medical informatics, i.e. computer-based analysis or dissemination of patient or disease data
    • G06F19/32Medical data management, e.g. systems or protocols for archival or communication of medical images, computerised patient records or computerised general medical references
    • G06F19/328Health insurance management, e.g. payments or protection against fraud
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06QDATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00Administration; Management
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06QDATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q40/00Finance; Insurance; Tax strategies; Processing of corporate or income taxes
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06QDATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q40/00Finance; Insurance; Tax strategies; Processing of corporate or income taxes
    • G06Q40/08Insurance, e.g. risk analysis or pensions
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06QDATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/10Services
    • G06Q50/22Social work
    • 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
    • G16H50/20ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
    • 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
    • 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
    • G16H50/00ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/70ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients

Abstract

A system for predicting healthcare risk events including the process of accessing patient data associated with one or more patents, accessing geographic and healthcare system data, filtering the patient data, geographic data, and healthcare system data into clean data, and applying a predictive risk model to the clean data to generate patient profile data and to identify a portion of the patients susceptible to one or more risk events.

Description

    REFERENCE TO RELATED APPLICATIONS
  • This application claims the benefit of and priority to: U.S. Provisional Application No. 60/628,476, filed on Nov. 16, 2004, entitled “Healthcare Surgical and Diagnostic Event Risk Predictive Modeling Analytic System Incorporating Unwarranted Geographic Treatment Pattern Variation Data”; U.S. Provisional Application No. 60/628,504, filed on Nov. 16, 2004, entitled “Healthcare Financial Risk Predictive Modeling Analytic System Incorporating Unwarranted Geographic Treatment Pattern Variation Data”; and U.S. Provisional Application No. 60/628,212, filed on Nov. 16, 2004, entitled “Healthcare Risk Predictive Modeling Analytic System Incorporating Unwarranted Geographic Treatment Pattern Variation Data.” This application is co-pending with Docket No. FMSF-P01-002, filed Nov. 16, 2005, and entitled “Systems and Methods for Predicting Healthcare Related Financial Risk.” The entire contents and teachings of the above referenced applications are incorporated herein by reference.
  • FIELD OF THE INVENTION
  • The invention relates generally to methods and systems for healthcare system analysis. More particularly, in various embodiments, the invention relates to applying predictive modeling to healthcare information to predict healthcare risk events.
  • BACKGROUND
  • Numerous countries throughout the world are facing an unprecedented rise in healthcare costs affecting both healthcare providers and employers. One major component of healthcare costs is costs associated with surgery. Another component of healthcare costs is costs associated with diagnostics.
  • Healthcare predictive models have been employed that utilize actuarial models of cost predictions based on standard demographic data of patients to predict health care costs. Predictive statistical modeling is a field of data mining that utilizes statistics, machine learning, pattern recognition, and other techniques to analyze information and/or data. Other healthcare predictive models have included timing data associated with the periods when patients are examined for a particular illness to estimate costs. However, prior attempts at predictive healthcare models have focused on resource utilization, rather than the likelihood that an individual will undergo a specific surgical or diagnostic procedure.
  • Accordingly, there remains a widespread need for improved mechanisms for assisting healthcare providers and employers in lowering healthcare costs while providing superior quality of healthcare to patients. For healthcare providers such as health insurers and managed care organizations (“MCOs”), there exists a need for determining which patients are likely to present the highest risk of undergoing a surgical or diagnostic procedure, referred to as event risk, which can assist in developing strategies for managing healthcare programs.
  • SUMMARY OF THE INVENTION
  • The invention, in various embodiments, is directed to systems, methods, and/or devices relating to identifying patients who are likely to incur costs associated with healthcare. According to one feature, the invention defines certain healthcare related risk events that lead to incurring such costs. Risk events includes two broad categories; diagnostic risk events and therapeutic risk events. A diagnostic risk event includes, for example, a medical procedure performed by a healthcare professional to identify a medical condition associated with a patient. A therapeutic risk event includes, for example, a medical procedure and/or treatment performed by a healthcare professional to treat a medical condition associated with a patient. Either a therapeutic or diagnostic risk event may include a surgical risk event. A surgical risk event is any medical procedure provided by a healthcare professional for a patient involving the removal, replacement, and/or examination of an organ or tissue. A surgical risk event may also be defined as the diagnosis or treatment of an injury, deformity, or disease by manual and instrumental means.
  • By predicting risk events, the invention enables interested parties to establish intervention plans to mitigate the occurrence of risk events for patients. Interested parties include, for example, healthcare providers, insurers, and payors. By mitigating the occurrence of risk events, the invention mitigates the costs or financial risks associated with such risk events. Risk events and financial risks may be collectively referred to as healthcare related risk outcomes.
  • More particularly, the invention employs a predictive model to identify patients who are likely to incur costs associated with healthcare. The predictive model may utilize multiple factors and/or variables to predict which patients are most likely to incur the highest, lowest, or a selected range of healthcare costs by predicting which patients are likely to incur certain costly risk events. The number of risk factors utilized can be extensive and include, without limitation, patient data and geographic/healthcare systems data.
  • The patient data may include information such as medical claims data, pharmacy claims data, referral post hospital discharge data, health risk assessment and functional status data, laboratories values, pre-notification or authorization data, and other risk factor data. The geographic/healthcare system factors may include practice pattern variation data, supply-sensitive factor data, healthcare system factor data, and other geographic and healthcare system factor data. In particular, the invention advantageously utilizes unwarranted healthcare/geographical treatment pattern variation data to more accurately predict which patients are more likely to incur certain risk events. An unwarranted healthcare/geographic treatment pattern variation is any variation in treatments across different geographic regions and healthcare systems that is not caused by patient preferences or characteristics.
  • In one feature, the predictive model determines the most significant risk factors associated with a particular type of risk event. Once the significant factors are identified, a logistic regression model is employed to apply a weight to each significant factor based on how closely each factor correlates to a risk event. The invention applies the risk factors and associated weights to a population of patients to establish a total weight or score for each patient of the population. Based on the total weights, the invention identifies a portion of patients associated with a range of susceptibility to a particular risk event.
  • According to one advantage, the invention enables interested parties to predict likely healthcare costs for an upcoming period. According to certain implementations, the invention enables the projection of healthcare costs over a period of at least about 3 months, 6 months, 9 months, 12 months, or greater than about 12 months.
  • According to another advantage, the invention enables interested parties to take action to limit healthcare costs by, for example, providing early intervention plans to patients and/or interested parties that prevent or mitigate the occurrence of risk events. The prevention of certain risk events may encourage interested parties such as healthcare insurers to pay for or cover the cost of such early intervention programs.
  • In various embodiments, the invention provides, without limitation, mathematical models, algorithms, methods, systems, devices, computer program codes, and computer readable mediums for performing the above predictive models to identify healthcare related risk outcomes.
  • In one aspect, the invention employs a software application running on a computer system for predicting healthcare related risk events. The software application may perform functions including: accessing patient data associated with one or more patients; accessing geographic and healthcare system data; filtering the patient data, geographic data, and healthcare system data into clean data; and applying a predictive model to the clean data to generate patient profile data and to identify a portion of the patients with a range of susceptibility to one or more risk events. The range may include an individual patient, a group of patients, or the entire patient population. In one feature, the invention categorizes one or more patients into one or more clinical segments. The segments may be based on preference sensitive conditions, chronic disease, or large medical cases not associated with chronic disease. In another feature, the invention applies the predictive risk model to each clinical segment and/or uses the segments to apply the predictive model.
  • In one configuration, the invention generates one or more facts based on the clean data. In a further feature, the invention reports the portion of patients with a range of susceptibility to one or more risk events to an interested party such as a healthcare provider, a health care insurer, or other payor. In another feature, the invention generates suggested intervention plans for one or more patients based on each patient's susceptibility to a risk event.
  • The patient data may include patient claims data and patient non-claims data. The patient claims data may include medical claims data and/or pharmacy claims data. The patient non-claims data may include referral data, functional status data, laboratory values, patient risk factors, demographics, disease burden, and/or disease complications. The geographic data may include geographic practice pattern variables and/or unwarranted geographic treatment pattern variations. The healthcare system data may include unwarranted healthcare system treatment pattern variables and/or variations.
  • In one configuration, the invention filters the patient data, geographic data, and healthcare system data into clean data by importing patient data files, mapping patient data into standard formats, processing adjustments and duplicates, checking patient data parameters against internal and external normal parameters, identifying and correcting data errors, and creating a table to link patient data to unique patient identifiers.
  • In another configuration, the predictive model includes: separating patient data into a first and second data set; evaluating regressively one or more risk factors in the first data set to determine weights associated with significant risk factors; and applying the weights for each significant risk factor to the second data set to validate the predictive risk model performance. In one embodiment, the predictive model evaluates regressively one or more risk factors in the second data set to determine weights associated with significant risk factors. The predictive model then compares the weights of risk factors of the first data set with the weights of the risk factors of the second data set.
  • In another feature, the invention includes applying, without limitation, a linear regression model, a non-linear regression model, a logistic regression model, a Bayesian network model, a Markov model, or a propensity score to evaluate the risk factors. In one configuration, the invention adds the weights of the risk factors associated with one or more patients to generate risk totals associated with the one or more patients. The invention identifies a portion of the patients with a range of susceptibility to one or more risk events by identifying a portion of the one or more patients with the highest risk totals. A weight may include a beta weight. The beta weight may used to derive a risk score and/or risk total. The risk score may be within a range of 0 to 1. The weight may include a cost associated with a risk factor and/or risk event. Certain risk factors may be considered insignificant for a particular risk event, having a value of 0.
  • In one feature, validating includes applying a best-fit test and/or goodness-of-fit measure. In another feature, the invention refines and validates the risk predictive model by comparing a first portion of the patient profile data with a second portion of the patient profile data. The refining and validating process may include dividing the patient profile into a model development data set and a model validation data set and applying weights to the model development data set to profile the model validation set. The data may include one or more variables. A condition and/or predictor flag may be associated with a patient clinical condition and/or occurrence.
  • In one configuration, the model development data may include patient claims data associated with a first period of time and the model validation data may be associated with a second period of time.
  • In another configuration, the portion of the patients with a range of susceptibility to the one or more risk events may include a selected percentage of the patients that are most susceptible to the one or more risk events.
  • In a further configuration, the portion of the patients with a range of susceptibility to the one or more risk events may include a selected percentage of the patients that are least susceptible to the one or more risk events.
  • In yet a further configuration, the portion of patients with a range of susceptibility to the one or more risk events may include a portion of the patients representing a selected spectrum of susceptibility to the one or more risk events.
  • The invention will now be described with reference to various illustrative embodiments.
  • BRIEF DESCRIPTION OF THE DRAWINGS
  • The foregoing and other objects, features, advantages, and illustrative embodiments of the invention will now be described with reference to the following drawings in which like reference designations refer to the same parts throughout the different views. These drawings are not necessarily to scale, emphasis instead being placed upon illustrating principles of the invention.
  • FIG. 1 is a conceptual block diagram of a healthcare risk predictive modeling analytic system according to an illustrative embodiment of the invention.
  • FIG. 2 is a functional block diagram of a computer for performing a predictive analysis according to an illustrative embodiment of the invention.
  • FIG. 3 is a flow diagram of an exemplary healthcare system risk analysis process according to an illustrative embodiment of the invention.
  • FIG. 4 is a conceptual block diagram of the healthcare risk analysis process according to an illustrative embodiment of the invention.
  • FIGS. 5A and 5B include an exemplary list of selected geographic practice pattern variables according to an illustrative embodiment of the invention.
  • FIG. 6 is an exemplary flow diagram of an exemplary data cleaning process according to an illustrative embodiment of the invention.
  • FIG. 7 is an exemplary receiver-operating characteristic (ROC) curve of sensitivity versus specificity according to an illustrative embodiment of the invention.
  • FIG. 8 is an exemplary graph of relative risk versus cutpoint according to an illustrative embodiment of the invention.
  • FIG. 9 is an exemplary graph of sensitivity versus positive predictive value for an exemplary back surgery predictive model according to an illustrative embodiment of the invention.
  • FIG. 10 is an exemplary graph of average cost in Year 2 versus predicted average cost in Year 1 according to an illustrative embodiment of the invention.
  • DESCRIPTION OF ILLUSTRATIVE EMBODIMENTS
  • As described above in the summary, the invention is generally directed to systems and methods that identify patients who have a predicted susceptibility and/or level of risk (e.g., more risk or less risk or a selected range of risk) to certain event risks and/or are associated with certain levels of financial risk, thereby facilitating the establishment of intervention plans to mitigate the healthcare event risks for patients and financial risks for patients and/or healthcare providers.
  • In one embodiment, predictive modeling is employed because it has the potential to reduce healthcare costs and/or spending while assisting patients by helping MCOs, insurers, or other providers identify patients who are most likely to incur future surgical and/or diagnostic events, and target specialized interventions to assist such existing patients, or new enrollees. Predictive modeling may also allow MCOs, insurers, or other providers to identify which patients will likely consume the most resources in the future as a result of such event risks and/or financial risks. Predictive modeling may further enable healthcare providers to identify high risk patients, and get interventions to them, as medical, biotech and drug treatments have grown more sophisticated and expensive.
  • In another embodiment, the invention is directed to systems and methods relating to surgical and diagnostic event risk predictive modeling (herein surgical, treatment, and diagnostic procedures may be referred to as events). In one feature, a system according to the invention assists in the determination of event risk—the risk of undergoing a surgical or diagnostic procedure, not the risk of adverse outcomes from the procedure—within a defined patient population. In another feature, a system according to the invention assists in the determination of financial risk to a healthcare provider associated with a patient population.
  • In a further embodiment, the invention helps identify individuals within a population who are at the highest risk of incurring such event risks, preferably within about a 3-9 month period. In yet another embodiment, a system according to the invention assists in the determination of diagnostic event risk—the risk of undergoing any diagnostic procedure (such as, for example, a Magnetic Resonance Imaging study; a coronary angiography catheterization; or an echocardiogram of the heart), not the risk of adverse outcomes from the procedure—within a defined patient population, and helps identify individuals within a population who are at the highest risk of incurring such diagnostic event risks, preferably within about a 3-9 month period.
  • The invention may also be applied to predicting the risk of any surgical procedure including, but not limited to, those associated with the knees, the hip, the back, uterine fibroids and uterine bleeding, and cardiac event risks (including, for example, coronary artery bypass graft (“CABG”), and Percutaneous Transluminal Coronary Angioplasty (“PTCA”) or other modalities of catheter based treatments of the coronary arteries).
  • In one embodiment, the invention employs statistical predictive modeling and clinical segmentation analytics to combine data associated with unwarranted geographic treatment pattern variations with relevant patient claim and non-claim information to determine future event risks and/or financial risks (collectively referred to as “predicted risk outcomes”) of the aforementioned types within patient populations. According to another feature, the invention also identifies individuals within a population who are at the highest risk of incurring such event risks or who are susceptible to incurring, generating, or otherwise experiencing a certain level of healthcare costs. This identification enables healthcare organizations to engage in intervention and health coaching of high risk individuals to lower their event risks and/or financial risks or costs.
  • While prior attempts at predictive healthcare models have focused on resource utilization, one configuration of the invention includes an analysis of the likelihood that an individual will undergo a specific surgical, treatment, or diagnostic procedure. According to one advantage, the systems and methods of the invention recognize the importance of unwarranted geographic treatment pattern variation data, and rely upon various information including, without limitation, medical research, a given patient's medical claims experience, and non-claims factors to identify relationships between healthcare utilization and event risks and/or financial risks.
  • FIG. 1 is a conceptual block diagram of a healthcare risk predictive modeling analytic system 100 according to an illustrative embodiment of the invention. The analytic system 100 includes computer system 102, local healthcare database 106, network 108, remote information system 110, and remote healthcare databases 112, 114, and 116. The computer system 102 also includes predictive risk modeling application 104.
  • FIG. 2 shows a functional block diagram of general purpose computer system 200 for performing the functions of the computer 102 according to an illustrative embodiment of the invention. The exemplary computer system 200 includes a central processing unit (CPU) 202, a memory 204, and an interconnect bus 206. The CPU 202 may include a single microprocessor or a plurality of microprocessors for configuring computer system 200 as a multi-processor system. The memory 204 illustratively includes a main memory and a read only memory. The computer 200 also includes the mass storage device 208 having, for example, various disk drives, tape drives, etc. The main memory 204 also includes dynamic random access memory (DRAM) and high-speed cache memory. In operation and use, the main memory 204 stores at least portions of instructions and data for execution by the CPU 202.
  • The mass storage 208 may include one or more magnetic disk or tape drives or optical disk drives, for storing data and instructions for use by the CPU 202. At least one component of the mass storage system 208, preferably in the form of a disk drive or tape drive, stores the database used for processing the predictive modeling of system 100 of the invention. The mass storage system 208 may also include one or more drives for various portable media, such as a floppy disk, a compact disc read only memory (CD-ROM), or an integrated circuit non-volatile memory adapter (i.e. PC-MCIA adapter) to input and output data and code to and from the computer system 200.
  • The computer system 200 may also include one or more input/output interfaces for communications, shown by way of example, as interface 210 for data communications via the network 212. The data interface 210 may be a modem, an Ethernet card or any other suitable data communications device. To provide the functions of a computer 102 according to FIG. 1, the data interface 210 may provide a relatively high-speed link to a network 212, such as an intranet, internet, or the Internet, either directly or through an another external interface. The communication link to the network 212 may be, for example, optical, wired, or wireless (e.g., via satellite or cellular network). Alternatively, the computer system 200 may include a mainframe or other type of host computer system capable of Web-based communications via the network 212.
  • The computer system 200 also includes suitable input/output ports or may use the interconnect bus 206 for interconnection with a local display 216 and keyboard 214 or the like serving as a local user interface for programming and/or data entry, retrieval, or manipulation purposes. Alternatively, server operations personnel may interact with the system 200 for controlling and/or programming the system from remote terminal devices via the network 212.
  • The computer system 200 may run a variety of application programs and store associated data in a database of mass storage system 208. One or more such applications may enable the receipt and delivery of messages to enable operation as a server, for implementing server functions relating to predicting risk event and/or financial risks using application 104 of FIG. 1.
  • The components contained in the computer system 200 are those typically found in general purpose computer systems used as servers, workstations, personal computers, network terminals, portable devices, and the like. In fact, these components are intended to represent a broad category of such computer components that are well known in the art. Certain aspects of the invention may relate to the software elements, such as the executable code and database for the server functions of the predictive risk modeling application 104.
  • Returning to FIG. 1, the predictive risk modeling application 104, in various embodiments, may combine predictive statistical modeling with clinical segmentation analytics to determine future risk of having identified surgeries or diagnostic procedures within a defined population and/or set of patients, and/or the associated financial risks.
  • FIGS. 3 and 4 are a flow and conceptual diagrams, respectively, of an exemplary healthcare system risk analysis process 300 and system 400 according to an illustrative embodiment of the invention. In operation in certain embodiments, the software application 104 performs the following. The application 104 accesses and/or retrieves patient data 402 associated with one or more patients (Step 302). The patient data 402 may reside within an internal database 208, local database 106, or a remote database 112, 114, and 116. The remote databases 112, 114, and 116 may be accessible via a communications network 108 including, for example, any one or combination of the Internet, an internet, an intranet, a local area network (LAN), wide area network (WAN), a wireless network, and the public switched telephone network (PSTN). Each of the remote databases 112, 114, and 116 may be associated with a public and/or private healthcare database including patient specific information, general healthcare information, general demographic information, and/or other information relevant to the financial and/or risk event analysis process 300.
  • The application 104 may also access and/or retrieve geographic and healthcare system data 404 (Step 304). Then, the application 104 filters the patient, geographic, and healthcare system data into clean data using a data cleaning/fact generation component 408 (Step 306). Based on the clean data, the application 104 generates one or more facts using the data cleaning/fact generation component 408 (Step 308). Then, the application 104 applies a predictive statistical model 412 to the clean data to generate and/or define patient profile data and/or to identify a portion of the patients associated with a range of predicted financial risk and/or to identify a portion of the patients with a range of susceptibility to one or more risk events (Step 310). The application 104 may output the financial risk predictions and/or risk event predictions 414 in the form of a data file that may be delivered to a local user interface and/or display 216 or to a remote information system 110 for further processing and/or viewing. In one embodiment, the clean data processed by the prediction model includes the facts generated from the clean data in Step 308 and/or any information that can be correlated to a predicted risk outcome.
  • In further illustrative embodiments, the application 104 also identifies individuals within such a population who are at the highest risk of incurring risk events. According to one advantage, the application 104 applies predictive statistical modeling, in combination with clinical segmentation analytics, to patient data, and also takes into account geographic factors. In one embodiment, the invention enables the aforementioned determination for about a 3-9 month period following the determination. In other embodiments, other periods may be used, such as, and without limitation, about a 1-3 month period, about a 1-6 month period, about a 1-9 month period, about a 1-12 month period, and greater than about 12 month period following the determination. In another illustrative embodiment, the application 104 performs clinical segmentation across patients exhibiting one or more of the following: preference sensitive conditions; chronic diseases; and large medical cases that do not fall into any chronic disease category. Chronic diseases may include, but are not limited to, asthma, chronic obstructive pulmonary disease (“COPD”), coronary artery disease (“CAD”), congestive heart failure (“CHF”), and diabetes. Clinical segmentation is described in greater detail below.
  • As shown in FIG. 4, the various categories of data that may be used by the application 104 of predictive modeling analytic system 100 include, but are not limited to, the following: patient data 402 including, e.g., claims-related data and non-claims related data; and geographic and timely medical research data on geographic and healthcare system factors 404 including, e.g., healthcare system factor data, unwarranted geographic treatment pattern variation data, and data addressing clinical care gaps in healthcare systems patient care delivery.
  • Clinical care gaps include gaps in patient care where a patient has not received tests, medications, or treatments in accordance with established evidence-based clinical guidelines specific to a patient's condition, or where a patient is receiving tests, treatments or medications in contrast to established evidence-based clinical guidelines. Clinical care gaps are used to identify unwarranted variation in the area of effective care. Health care system factor data includes information about the health care system where an individual receives care, e.g., geographic area, a hospital referral region (HRR), and/or the hospital system (a hospital and the physicians who practice in it). These factors may include the number of hospital beds per capita, number of specialists per capita, and/or diagnostic testing intensity associated with the health care system. Health care system factor data has been shown to be a strong determinant of the number of supply sensitive services a patient receives.
  • Patient claims data may include, but are not limited to, the following classes of data: medical claims data and pharmacy claims data. Patient non-claims data may include, but are not limited to, referral data, pre-notification or authorization data, post-hospital discharge data, health risk assessment and functional status data, laboratory values (such as, for example, prostate-specific antigen (PSA) values), and other patient risk factors that include, but are not limited to, demographics, disease burden, and disease complications.
  • FIGS. 5A and 5B include an exemplary list of selected geographic and/or healthcare practice pattern variables according to an illustrative embodiment of the invention. Geographic and/or healthcare practice pattern variables for which geographic rates of practice pattern variation are analyzed in the model include, but are not limited to, those provided in FIGS. 5A and 5B. Unwarranted healthcare system and geographic treatment pattern variation is defined as variation in treatments across different healthcare systems and geographic regions that is not driven by patient preferences or characteristics. Patient demographics may include, but are not limited to, a patient's unique member identification, date of birth, sex, enrollment and membership information, and geographic data such as ZIP code, Hospital Service Area (“HSA”), and/or Hospital Referral Region (“HRR”). Pharmacy claims data may be maintained, for example, at drug class or NDC-level, with groupings by class; this data may include, but is not limited to, number of dispensed prescriptions; days supply; measures of adherence; and most recent fill date.
  • The application 104, according to certain illustrative embodiments, employs statistical modeling to capture relevant relationships based on a patient's history. According to a further embodiment, the application 104 also incorporates timely medical research on geographic and healthcare system factors 404 into the healthcare event risk and/or financial risk predictive model, which yields results that are both predictive and clinically relevant. In one preferred illustrative embodiment, the timely medical research data 404 includes clinical research data supported by institutions such as the Foundation for Informed Decision Making, Dartmouth Center for the Evaluative Clinical Sciences, Harvard Medical School, or Maine Medical Center's Center for Outcomes Research and Evaluation which may reside, for example, within one or more databases such as databases 112, 114, and 116.
  • The incorporation of unwarranted geographic and healthcare system treatment pattern variation data into the risk event and/or financial risk model of the application 104 allows interested parties such as insurers, healthcare plans, employers, or other providers or payors to identify and reduce unwarranted variations in the real-world delivery of healthcare, and to identify a higher likelihood of successful health coaching opportunities. According to one illustrative embodiment, the application 104 incorporates research on disease burden, geographic practice patterns, and supply-sensitive factors into the risk event and/or financial risk models. By way of example, supply-sensitive factors may include, but are not limited to, admission and re-admission frequency; total hospital days; frequency of emergency room visits; frequency of physician visits; date of most recent encounter; number of unique providers seen; and visit frequency among: PCPs, nurse practitioners, chiropractors, OB/GYNs, specialists overall, and other provider types; and relative value units (“RVU”), total and sub-group, which may be used as a standardized resource utilization measure instead of cost.
  • Prospective modeling employed by the illustrative application 104 may be used to predict risk events and/or financial cost for a given time period based on risk factors identified in a prior time period. In one embodiment, that application 104 requires members and/or patients used in the modeling process to have continuous eligibility in a healthcare system for both time periods. Such risk factors may include incurred costs. Since the values of incurred costs do not tend to follow a linear pattern, costs may be transformed using data transformation factors within the statistical modeling component 412 of the application 104 that include, for example, the natural logarithm. Various conventional statistical modeling systems may be employed in the application 104, including, but not limited to, Markov and Bayesian statistical modeling systems. The application 104 may be implemented with and applied to computer-based modeling using any suitable computer language. In one illustrative embodiment, the computer language used is SAS. Other computer languages may include, without limitation, C, C++, JAVA®, COBAL, BASIC, HTML, XML, SGML, and like computer languages.
  • Preferably, patient data 402 includes claim and non-claim data and patient-related risk factors. The patient data 402 may be processed including geographic factors in the risk event and/or financial risk predictive model of the application 104 as shown in FIG. 4. These patient-related risk factors may include, but may not be limited to, age; gender; significant medical events; chronic conditions including, but not limited to, asthma, chronic obstructive pulmonary disease (“COPD”), coronary artery disease (“CAD”), congestive heart failure (“CHF”), and diabetes; co-morbidities; complications; utilization; clinical diagnostic groupings; code groupings; procedural groupings, pharmaceutical claims, and geographic region. According to an illustrative embodiment, the data for the risk event and/or financial risk predictive model of the application 104 may then be subjected to a data cleaning process, and may also be subjected to a process to identify and categorize relevant facts and relationships among facts (a fact generation process, described below) by data cleaning/fact generation component 408, to create one or more patient profiles 410.
  • FIG. 6 is a flow diagram of an exemplary data cleaning process 600 according to an illustrative embodiment of the invention. In this illustrative embodiment, data cleaning may include, for example, one or more of: (1) Importing client data files (Step 602); (2) Mapping client data to standard formats (Step 604); (3) Processing adjustments and duplicates (Step 606); (4) Checking client data against internal and external norms (Step 608); (5) Identifying and correcting data errors (Step 610); and (6) creating a patient-level system of tables to link client data to unique patient identifiers (Step 612). In one embodiment, the application 104 excludes members with certain disease states, trauma, dialysis or heart transplants from the predictive model because such procedures may entail high financial risks.
  • As noted above, data may also be subject to a fact generation process in connection with preparing a patient profile. This is a process aimed at consistently extracting and employing clinical, financial, utilization and/or individual information from healthcare data.
  • The facts (including relationships among facts) resulting from this process can be defined at multiple levels, such as:
      • First-level facts—these are the most basic facts, and are used to identify certain types of claims (e.g., claims pertaining to an asthma diagnosis). They are based directly on clinical codes (such as CPT4, ICD9, revenue codes, and specialty codes). For example, cardiac catheterization (CATH), Echo, and Percutaneous Transluminal Coronary Angioplasty (PTCA) are all first-level procedure facts.
      • Second- or higher-level facts—these are based on Boolean (e.g., AND, OR) association of first- or other lower-level facts. For example, left ventricular ejection fraction (LVEF) is a higher-level procedure fact based on lower-level CATH, Echo and PTCA facts.
      • Complex facts—these are based on an association of facts through complex logical relationships. Identification facts are an example of complex facts, as is the final output leading to a risk event and/or financial risk prediction fact.
  • In one preferable embodiment, a risk predictive modeling analytic system 100 includes, but is not limited to, approximately 1500 such facts. The facts identified in the analysis process can be used in the application 104 for various purposes, such as:
      • Development and refinement of the statistical model;
      • Reporting of event risk and/or financial risk predictions; and/or
      • Generating suggested intervention campaigns based on event risk and/or financial risk predictions.
  • In this illustrative embodiment, the patient profile data 410 resulting from this process of data cleaning and fact generation is then fed into the predictive risk statistical modeling component 412 of the application 104. At least one data file 410 is separated (randomly or otherwise) into two data sets. The first set is used for model development and each potential data point/risk factor (“risk factor”) is evaluated against client specific data. In one embodiment, a stepwise linear regression is used to filter out non-significant risk factors. In another embodiment, a non-linear regression model is used. The resulting estimates and/or, for example, weights, for each significant risk factor are applied to the second data set to validate the model's performance. Various conventional validation and “goodness of fit” tests may be employed. In one embodiment, model validation and goodness of fit tests are based on measures published by the Society of Actuaries (such as A Comparative Analysis of Methods of Health Risk Assessment, Daniel L., Alice Rosenblatt, Deborah A. Taira, et. al., Schaumburg, Ill., Society of Actuaries, 1996).
  • In one embodiment, the predictive modeling of the application 104 employs surgery indicators that have occurred about 3-15 months before the study date. This eliminates about 3 months before surgery to account for claims lag as well as the likely inability to intervene effectively in such a period before surgery.
  • The predictive model of the application 104 may also be refined and validated using one or more geographic factors such as clinical care gaps and geographic rates of practice pattern variation (see examples in FIGS. 5A and 5B), including various supply-sensitive factors described above. This population-specific data captures local practice patterns and the capacity of the acute care sector in the client's geographic regions that result in variations in the delivery of healthcare. Each model may be reviewed to incorporate additional clinical criteria and revisions to the model based on additional data, which may include timely medical research data 404. In one preferred embodiment, the model is reviewed quarterly. In other embodiments, the model is reviewed and updated monthly, fortnightly, weekly, or daily to incorporate additional clinical criteria and revisions to the model based on additional data.
  • In one embodiment, the following example of model development is employed. The model includes a split sample methodology wherein 50% of the profile data 410 is randomly selected for a model development sample and the other 50% of the data is selected for a model validation sample. A model is then developed that uses patient data 402 and geographic and healthcare system factor information 404 for a first period (such as 3-15 months) represented in the profile data 410 of the development sample to identify data points/risk factors associated with risk events and/or financial costs in a second period represented in the profile data 410 (such as the next 3 to 9 months). In an illustrative embodiment, the model used includes a logistic regression. In another, the model used includes a propensity score. Other models may also be used. Numerous variables may be used in developing the model, including without limitation age, gender, log transformed cost for the first period, number of specialist visits, primary care capacity, flags for conditions such as renal disease (not on dialysis), Rheumatologic disease, Cancer (non-metastatic), and treatment with anti-depressants, Ca+ channel blockers, ACE inhibitors, diuretics, anti-anginals, anti-anxiety medications, and cardiac glycosides. In one embodiment, the operating characteristics of the development data set are measured using R2 metrics. For example, in one embodiment, the operating characteristics of the development data set are in the range R2=0.34−0.45.
  • The predictive value of the model may be validated by applying beta weights developed from the model development sample to profile data from the validation sample, calculating predicted likely risk events and/or financial cost in a second period represented in this validation sample data based on information from a first period represented in the data, and comparing the predicted risk events and/or financial costs to the actual surgical or diagnostic events and/or actual healthcare costs experienced in the second period. In one embodiment, the operating characteristics of the validation data set are measured using ROC metrics. In one embodiment, the operating characteristics of the validation data set are in the range R2=0.33−0.44.
  • Clinical Segmentation
  • According to a further illustrative embodiment, the application 104 incorporates clinical segmentation into the predictive statistical model component 412. In one embodiment, the application 104 performs clinical segmentation across the following categories: preference sensitive conditions; chronic disease; and large medical cases that do not fall into chronic disease categories. This segmentation can be used to identify information about the individuals' disease conditions and comorbidities, effective care opportunities and utilization profiles. By way of example, comorbidities include, but are not limited to, such conditions as AIDS, asthma, ischemic heart disease, chronic heart failure, diabetes, depression, hypertension, and other conditions. Effective care opportunities include, but are not limited to, pneumonia vaccination; influenza vaccination; controller medication in asthma; lipid profile; microalbuminuria testing; HgbA1C testing; diabetic eye exam; CAD/CHF beta blocker treatment; follow-up care; and other measures.
  • The application 104 may, for example, complete an extensive statistical analysis within each of these categories to predict future surgical or diagnostic events, or future financial costs/expenses. In one embodiment, the application 104 identifies the top about 10% of patients that represent the predicted highest impactable event risk, highest financial risk to the health plan or other payor, and/or the highest level of cost to the health plan or other payor. Within this about 10% the application 104 segments the top about 4% for highest level intensity of interventions. According to other embodiments, the application 104 segments the top about 1%, 2%, 3%, 5%, 6%, 7%, 8%, or 9% for highest level intensity of interventions. In a similar fashion, the application 104 may also segment the lowest about 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9% or 10% of predicted impactable event risk and/or financial cost for lowest level intensity of interventions.
  • Preference Sensitive Conditions (“PSCs”)
  • PSCs typically involve conditions where the patient has some choice of the treatment and/or action to address a medical condition, e.g., elective surgery. The application 104 may create a monthly or other period-based prediction of members that are in a “decision window” for a significant surgical or diagnostic event, or for a significant cost. These files may be revised monthly or on another periodic basis to capture the most current data on members and/or patients. (Patients with PSCs may represent, for example, 15-20% of members and 25-40% of total medical costs, with the highest risk members in this category representing excellent opportunities for health coaching and management.)
  • Chronic Diseases
  • In one preferred embodiment, the application 104 identifies patients with one of five chronic diseases: asthma, chronic obstructive pulmonary disease (“COPD”), coronary artery disease (“CAD”), congestive heart failure (“CHF”), and diabetes. In certain embodiments, these diseases may be selected because these five disease states present the greatest opportunity for intervention and reduction in unwarranted variation that can reduce medical cost while improving the quality of patient care.
  • Large Medical Cases
  • The application 104 may also provide event risk and/or financial cost estimates of those members within large medical cases that are not members of chronic disease categories. In an illustrative embodiment, these may include, for example, patients with otherwise non-identified cancers. For example, in one illustrative embodiment, large cases may be identified by Clinical Complication Scores (“CCS”), a methodology developed and continuously updated by the Agency for Healthcare Research and Quality (“AHRQ”).
  • Building the Member Profile
  • In one embodiment, the results of the application 104's statistical modeling and clinical segmentation may include, but are not limited to, individual-level records that may include, for example, event risks and/or financial risks, key drivers of utilization, clinical care gaps, acute clinical events, and the probability of facing discrete Preference Sensitive Condition events. Information in each member profile may then be used to further segment members within high-risk groups. This further segmentation may be used to direct specific interventions for each member and/or patient.
  • In one embodiment, the application 104 uses six primary factors: excessive or unneeded hospitalizations; unneeded surgeries; clinical care gaps; overuse of emergency room service; overuse of high-cost pharmaceuticals; and under-use of appropriate preventive pharmaceuticals (condition-specific) to segment each member by their costs and utilization in each of the clinical segmentation categories mentioned above. This allows for effective member interventions across categories (e.g., interventions to reduce potential hospitalizations should be managed differently for members who have extremely high utilization of chemotherapy and other oncology drugs). Thus, the application 104 can allow for impactable financial and/or event opportunities, such as clinical opportunities. In one embodiment, the application 104 uses a combination of clinical trials and observational data to identify these opportunities.
  • In the application 104, an individual's predicted event risks and/or financial risks may be combined with expected results from specific, proven interventions (e.g., telephonic, mailing, others) and a healthcare organization's specific goals for the program (e.g., to reduce costs, improve Health Plan Employer Data and Information Set (“HEDIS”) scores, reduce absenteeism).
  • One product of the application 104 may be a risk score for future predicted event risks and/or financial costs. The risk score may be presented as a percentage likelihood of event risk and/or cost in a defined period of time (e.g., a 3-9 month period in the future). Event risks and/or financial risks may be established by a predictive model of the application 104 to a range between 0 to 100%. In addition to producing a risk score for future risk events and/or costs, the application 104 may identify key points in the care process that can be used as ‘leverage points’ to change expected event risks and/or financial costs. These may include information about recent utilization (e.g., consultations to specialists, emergency room admissions, hospitalizations, etc.). This information may be used in care management outreach efforts.
  • The application 104 may also identify key selected effective care opportunities for the chronic population. Effective care opportunities (also known as “evidence based medicine”) include interventions that are of known clinical effectiveness. When systematically applied, they reduce morbidity, mortality and costs. These interventions include use of specific preventive services (e.g., diabetic retinal exam) and pharmaceutical interventions (e.g., lipid lowering agents in patient with coronary artery disease). These care opportunities may be incorporated as individual and total patient clinical scores (weighted by risk) into the patient level output and that may be sent, for example, to a health plan provider or other payor.
  • Based on the analysis allowed by the predictive statistical model 412 of the application 104, a database of actionable information may be created such as database 208 and/or 106. In one embodiment, such actionable information may be accessed by call centers staffed by healthcare professionals such as licensed RNs, dieticians, and other clinicians to support outbound interventions to better manage high event risk and/or high cost individuals for managed care organizations. The actionable information may be stored within a database of information system 110 among other databases.
  • Different embodiments of the predictive modeling application 104 described herein may be implemented as computer software, hardware, or a combination of software and hardware. In certain embodiments, patient population profile information, risk event predictions, and financial risk predictions may be displayed and/or outputted to a computer graphical user interface (“GUI”) such as display 216 for user viewing, or outputted to another system such as information system 110 for further processing.
  • FIGS. 7-9 provide graphical illustrations of how, in one embodiment, the predictive model of the application 104 may be applied to predict surgical or diagnostic Risk—in this example, the risk of having back surgery. Alternatively, the predictive model of application 104 may be applied to predict the financial risk associated with back surgery. One purpose of the predictive model for back surgery is to identify beneficiaries at risk of back surgery using, for example, medical and pharmacy claims incurred within an actionable time period. Such a predictive model may identify members with back pain who can benefit from shared decision making enabled by the predictive model.
  • In one embodiment, the predictive model of application 104 for back surgery may be implemented according to an approach generally including 1) creating an analytic file of patient profile data 410 containing data for variables that may be related to back surgery and tend to occur more than 3 months before surgery (e.g., pharmacy claims for back pain medication, which may be likely to commence before the 3 months preceding surgery, as opposed to MRIs, which may be more likely to occur within the 3 months preceding surgery), and 2) running regression analyses to identify meaningful predictors of back surgery and develop a statistical model that tends to yield a meaningful predictive result when applied to patient profile data 410. For a particular data set, these predictors/risk factors may include certain patterns of physician visits coupled with pharmacy claims for medications associated with the treatment of back pain. In addition, various geographic and healthcare system factors 404 (such as healthcare system factor data, unwarranted geographic treatment pattern variations data, supply-sensitive factor data, and clinical care gaps) may be identified as significant in the development and validation of a predictive model for back surgery, such as the number of back surgeries per 1000 enrollees, the number of neurosurgeons or orthopedic surgeons per 100,000 residents, the number of physical medicine/rehabilitation facilities or practitioners per 100,000 residents, the number of unique providers seen, and visit frequency among PCPs and chiropractors. (See, e.g., FIGS. 5A and 5B.)
  • In one embodiment, a provider selects and/or identifies a set of predictors or independent variables that may be indicative of predicting a risk event and/or financial risk. The predictors may be selected manually based on provider experience and understanding of the medical conditions, medications, symptoms, and any other medical indicators of a future risk event. The predictors may then be programmed, submitted, and/or entered into a database such as database 106 to be accessed by the application 104. The predictors may be selected based on an analysis of existing patient data where common variables associated with a group or groups of patient that have experienced a particular risk event are identified, either by manual review or by an automated database 112, 114, 116, and 106 search and analysis. The application 104 may include a data mining process and/or subroutine to identify predictors associated with predicting a particular risk event. The data mining process may also be a separate application of the computer system 200 and/or another information system such as information system 110. Periodically, a panel of medical professionals may review and/or modify the set of predictors associated with a predicted risk event.
  • Once the predictors associated with a particular risk event such as back surgery are determined, the application 104 develops a predictive model based on patient data 402 extracted from one or more populations of patients and geographic/healthcare system data 404. A predictor may also be described as an independent variable or risk factor. In one exemplary embodiment, the application 104 employs one or more logistic regression models and/or stepwise logistic regression models to develop the predictive model for a back surgery risk event.
  • A logistic regression model may be preferable in relation to other regression models, such as a linear regression model, because the values of the dependent variable Y (e.g., the predicted risk event) are binary (0,1). In certain embodiments, the application 104 employs a logistic regression model to generate a logistic curve that relates an independent variable and/or predictor X to a rolling mean of the dependent variable Y as shown in the following formulas:
    P=exp (a+bX)/1+exp (a+bX)  (1)
    P=1/1+exp−(a+bX)  (2)
      • Where: P=the probability of a 1 (the proportion of 1s, the mean of Y)
      • exp=the base of the natural logarithm (approximately 2.718)
      • a=parameter, yields P when X=0
      • b (beta weight)=adjusts changes in P with changes in X (note: b may be standardized or unstandardized)
        Because the relationship or correlation between X and P is non-linear, the beta weight b does not have a linear interpretation as in ordinary linear regression modeling.
  • The application 104 preferably employs a maximum likelihood estimation (MLE) to determine the loss function for the above logistic curve. A likelihood may be considered a conditional probability of Y given X. Thus, in certain embodiments, the application 104 chooses parameters (a and b) of the logistic curve to estimate the best fitting curve that characterizes the relationship of the predictors X to the resulting risk event Y. In one embodiment, the parameters are chosen randomly. In another embodiment, the parameters are chosen based on a trial-and-error method. The estimates are called maximum likelihood estimates because the parameters are chosen to maximize the conditional probability of the relationship between X and Y.
  • In one embodiment, the application employs a numerical analysis to find the maximum likelihood estimates (MLEs). For example, the application 104 makes an initial estimate of the parameters. The application 104 computes the likelihood of the logistic curve based on the parameters. Then, the application 104 improves and/or adjusts the parameter estimates to a certain degree and re-calculates the likelihood of the logistic curve fitting the data. The application 104 continuously performs this likelihood estimation for a number of iterations and/or until the parameter changes are below a minimum amount. In certain embodiments, the maximum number of iterations is greater than or equal to about 50, about 100, about 200, about 500, and about 1000. In certain embodiments, the minimum amount of parameter change is less than about 0.1, about 0.01, and about 0.001.
  • By applying an MLE to determine the best fitting logistic curve of the relationship between each predictor X and risk event Y, the application 104 determines the beta weight b associated with each predictor and/or risk factor. Table 1 provides an exemplary table of the determined beta weights of predictors associated with predicting a lumbar back surgery risk event.
    TABLE 1
    Exemplary Predictors and beta Weights associated
    with Lumbar Back Surgery Risk Event
    Predictor Category Beta
    (Element) Predictor Weight
    Demographic age 0.0321
    male 0.2030
    Diagnosis Lumbar Back Pain 1.1439
    Unspecified Back Pain 0.7158
    Sciatica 1.1300
    Procedure Unspecified Back Myelography 1.1555
    Back Lumbar CAT Scan 1.0010
    Back Lumbar X-ray 1.4451
    Back Cervical X-ray −0.3148
    Unspecified Back X-ray 0.3018
    Disk Excision 0.7762
    Anesthetic Injections 0.9280
    Physical Therapy 0.6247
    Pharmacy Narcotic Analgesics 0.3188
    Anti-inflammatory Analgesics 0.4372
    Anti-depressents 0.3296
    Specialist Orthopedic Visit 0.2692
    Geographic/Healthcare HSA_primary 0.0229
    Cost Log_year1_cost 0.4742
  • Table 1 includes predictors, predictor categories and/or elements, and beta weights associated with patient data 402 and geographic/healthcare system data 404 for a lumbar back surgery risk event. The application 104 may derive the beta weights of predictors associated with the demographics, diagnosis, procedures, pharmacy, and specialist categories based on the patient data 402. The application 104 may derive the beta weights of the predictors associated with geographic and cost categories from the geographic/healthcare system data 404 and/or other information sources.
  • In one embodiment, the geographic/healthcare system beta weight is derived from the rate of a medical condition and/or occurrence (or predictor) within a geographic and/or health care system area, such as a hospital service area (HSA), multiplied by the beta coefficient that associates the particular condition and/or predictor with the predicted risk event and/or financial risk. For example, the HSA primary predictor in Table 1 may be derived from the beta weight that correlates the number of physical medicine/rehabilitation facilities or practitioners with lumbar surgery risk events multiplied by the number of physical medicine/rehabilitation facilities or practitioners per 100,000 residents within the HSA of a particular patient.
  • Table 2 provides an exemplary listing of the beta coefficient and adjusted rate for the geographic/healthcare system predictor Medical Discharges (DRG) within, for example, a portion of the state of Massachusetts. In certain embodiments, adjusted rates are derived for health care regions and/or HSAs within any states, provinces, regions, territories, countries, and other geographic and/or healthcare system segments to account for variations in the occurrence of particular risk events and/or financial risks. The variations may be unwarranted geographic treatment variations.
    TABLE 2
    Exemplary List of beta Weight Determination for Geographic/Healthcare
    System Variable Medical Discharges (DRG)
    Age/Sex Rate*Beta
    Home Service Area HSA number adjusted rate coefficient
    MA-Athol 22002 275.8403472 0.090486667
    MA-Attleboro 22003 255.7602388 0.083899589
    MA-Ayer 22004 264.9613393 0.086917918
    MA-Beverly 22005 224.8310859 0.073753589
    MA-Boston 22006 282.136566 0.092552079
    MA-Brockton 22007 298.7619729 0.098005878
    MA-Burlington 22008 227.4512058 0.074613094
    MA-Cambridge 22009 261.7359147 0.085859849
    MA-Clinton 22010 300.1428059 0.098458846
    MA-Concord 22011 215.9323691 0.070834454
    MA-Everett 22013 345.3774771 0.113297628
    MA-Fall River 22014 285.1100436 0.093527499
    MA-Falmouth 22015 215.4991889 0.070692354
    US-United States 99999 248.9081168
  • Table 2 provides an exemplary list of the Medical Discharge (DRG) Rate per 1,000 Medicare Enrollees within a group of HSAs. Because the beta coefficient=0.00032804 for DRGs in this example, the beta weights are determined by multiplying the Adjusted Rate for a particular HSA with the beta coefficient for DRGs. The resulting beta weight for a particular HSA (shown in the fourth column) may then be summed with other beta weights to determine whether a particular patient has a susceptibility to a risk event and/or financial risk. Each HSA may be assigned an HSA number as shown in Table 2 to enable identification of a particular HSA. The application 104 may identify the HSA and/or other geographic/healthcare segment associated with a particular patient by determining the ZIP code of the patient's home address and/or the ZIP code of the location in which medical care is being administered. The Age/Sex Adjusted Rate (e.g., selected year period) indicates the number of medical discharges for that particular HSA over a period of time, e.g., the year 2001. The period of time employed may be at least about 3 months, 6 months, 12 months, and 24 months. In the circumstance where there are negative or suppressed rates associated with a particular HSA, an average value for the adjusted rate for a particular set of HSAs may be employed. For example, the average value for the adjusted rate for United States of America may be employed.
  • In certain embodiments, the application 104 uses the predictors and associated beta weights to predict risk events and/or financial risks. In another embodiment, the application 104 uses categories of predictors to predict risk events and/or financial risks. In one embodiment, the application 104 uses the adjusted geographic/healthcare beta weights of Table 2 for predicting risk events and/or financial risks. In another embodiment, the application 104 uses non-adjusted geographic/healthcare beta weights and/or beta coefficients for predicting risk events and/or financial risks.
  • In one embodiment, the application 104 determines the odds that a risk event and/or financial risk occurs according to the following formula:
    Odds=P/1−P  (3)
      • where: P=the probability that a risk event occurs
      • 1−P=the probability that a risk event does not occur
        In logistic regression, the dependent variable Y (e.g., risk event) can be express as a logit which is dependent on the Odds of the dependent variable. Thus, the application 104 may express the logistic regression by the following formula:
        Log (Odds)=logit (P)=ln (P/1−P)  (4)
        Therefore, because the logit is a log of the Odds and the Odds are a function of P (P=1), the logistic regression may be expressed as a linear expression and/or curve by the following formula:
        Log (Odds)=Logit (P)=a+bX  (5)
        Thus, while the log Odds are linearly related to the predictors X, the probability P of a risk event is non-linearly related to the predictors X. Because the Log (Odds)=a+bX, Equation (2) can be expressed as:
        RS (risk score)=P=1/1+exp −(Log (Odds))  (6)
        Assuming that the predictor X may be expressed using a binary 1 (“predictor flag”) when the predictor is present in a patient profile, the Log (Odds) including multiple predictors may be expressed as:
        Log (Odds)=sum of (predictor flag * beta weights)  (7)
        Accordingly, a risk score RS and/or probability P of Equation (6) may be employed to determine a risk score associated with a portion of the patients in a population to determine which patients are most susceptible to a particular risk event. Using Equations (6) and (7), the application 104, in certain embodiments, calculates a risk score based on the predictors identified in a patient profile 410. For example, if a patient profile 410 for patient A includes predictor and/or condition flags associated with the predictors age, back lumbar X-ray, and MA primary care, Equation (7) is calculated as:
        Log (Odds)=0.321+1.4451+0.0229=1.789
        Therefore, the risk score RS for patient A to incur a lumbar back surgery event is calculated as:
        RS=1/1+exp−(1.789)=0.857 or as P=85.7%
        The risk score RS, in this instance, is the probability P that the risk event, e.g., lumbar back surgery, occurs due to the three predictors and/or risk factors identified in Patient A profile 410. Patient B may have three other predictors associated with the lumbar back surgery risk event and have a RS=0.812. Patient C may have one predictor associated with the lumbar back surgery risk event and have a RS=0.321. In one embodiment, the application 104 determines the RS for each patient within a group of patients. The application 104 then determines a portion of patients with a range of susceptibility to the lumbar back surgery risk event based on the range of risk scores of the patients. The most susceptible patients have the highest risk scores and the least susceptible patients have the lowest risk scores.
  • In certain embodiments, the application 104 may express financial risk associated with a particular risk event as follows:
    PC (predicted cost)=exp (Log (Odds))  (8)
    An associated financial risk score (FRS) may also be express as follows:
    FRS (financial risk score)=PC/C  (9)
      • where: C is a constant such as 20000.
        For example, the PC and FRS for patient A above are calculated as:
        PC=exp (1.789)=5.984
        FRS=5.984/20000=0.00299
        With regard to the risk score, patient A may be ranked among a group of patients to determine which portion of patients posses a certain range of financial risk associated with lumbar back surgery. In other embodiments, financial risk may be expressed in the form of a probability and/or risk score according to the formula (6). Other values and/or risk totals may be employed to rank and/or identify one or more patients within a patient population with a range of susceptibility to a risk event and/or financial risk. The risk totals may include, without limitation, standardized and/or normalized beta weights, unstandardized and/or unnormalized beta weights, probabilities, propensity scores, financial costs, and like weight scores. The risk total score range may be adjusted and/or normalized to a particular range such as 0-1, 0-10, 0-50, 0-100, 0-1000 and like numerical ranges. The score may be expressed as a percentage, fraction, chart, scale, bar graph, plot, and any other image that depicts the predicted risk event and/or financial risk.
  • In one embodiment, the application 104 employs one or more chi-square tests such as a model chi-square and/or likelihood ratio test to statistically test whether a particular predictor reduces the badness-of-fit measure or increases the goodness-of-fit measure. If the chi-square is significant, a predictor is considered to be a significant predictor in the model which is analogous to the beta weight in a simultaneous logistic regression. The chi-square test may include a stepwise and/or sequential logistic regression. In one embodiment, the likelihood ratio, also known as the likelihood ratio chi-square has a probability value of greater than about 0.05 for a well-fitting model.
  • Certain patients may be excluded from intervention plans because they are less likely to benefit from shared decision making for preference sensitive decisions (PSD). This could either be due to the presence of a condition that takes priority over the PSD or a clinical reason and/or designation indicating that the risk event and/or surgical decision is not preference sensitive. Accordingly, the application 104 may exclude certain patients from the prediction model analysis based on designated exclusion facts. The following exemplary list of patient profile 410 facts may be used to exclude certain patients and/or healthcare system members from PSD targeting for lumbar back surgery.
      • Spine Trauma
      • Osteomyelitis
      • Lumbar Back Surgery in previous 12 months
      • AIDS
      • Any Cancer
        Certain facts may be specific to certain models. For example, the AIDS and cancer exclusions, in one instance, may be specific to the lumbar back model.
  • In certain embodiments, the application 104 may determine and output certain performance benchmarks. Table 2 lists some performance metrics for a predictive model such as the lumbar back predictive model. The measures may be based on validation reports derived from empirical data regarding actual outcomes from one or more interested parties. Each measure may be based on patients that are identified as being in the top 10 risk groups. There may be a significant amount of variation in model performance between different interested parties and/or payors which may be taken into account when predicting risk events and/or financial risk, identifying intervention opportunities, and defining target lists of patients for intervention.
    TABLE 3
    Exemplary Predict Model Performance Metrics
    Percent of Relative
    Interested Party Population Risk Sensitivity PPV
    Payor A (top 10) 4% 8.04 25% 6.9%
    Payor B (top 10) 3% 9.58 20% 4.8%
    Payor C (top 10) 10%  4.59 34% 4.2%
    Payor (top 10) 0% 4.96  2% 4.7%

    In the exemplary Table 3, the relative risk is the ratio of surgery rates and/or predicted risk events for identified patients to the general population. A relative risk greater than 1 may indicate that the identified group has a higher surgery and/or risk event rate. Larger values may indicate a greater difference. Sensitivity indicates the proportion of all surgeries and/or risk event that were captured by the group. The Positive Predictive Value (PPV) is the percent of identified patients that incurred a risk event.
  • While the R2 value may be appropriate for certain regression models such as linear regression models, the R2 value may not provide a sufficiently accurate measure when the application 104 employs a logistic regression model. However, the R2 value may be utilized with a logistic regression if desired with an understanding of its limitations. As an alternative to the R2 value, the application 104 may employ, without limitation, the Cox and Snell's R-square, the Nagelkerke R-square, the R1-square, and R1a-square, and like values.
  • In one embodiment, the application 104 employs a linear regression model to predict risk outcomes such as risk events and/or financial risk. The application 104 may employ other model including, without limitation, the log-linear, logit, and/or probit regression models to predict risk events and/or financial risk. A predictive model may include multiple regression models. In one embodiment, a separate regression model is applied to each variable.
  • In one illustrative embodiment, the predictive model component 412 of application 104 also includes stratifications wherein regression models produce continuous risk scores. Stratification permits the size of the targeted population to be set based on various factors, such as resources or objectives.
  • FIG. 7 is an exemplary receiver-operating characteristic (ROC) curve of sensitivity versus specificity according to an illustrative embodiment of the invention and provides an illustration of potential ROC Curve metrics for the operating characteristics of a predictive model of the application 104 for back surgery. It illustrates an association of predicted probabilities and observed responses (in this example, 85.7% concordant; 1.9% discordant; 12.4 % tied; c=0.919).
  • FIG. 8 is an exemplary graph of relative risk versus cutpoint according to an illustrative embodiment of the invention and provides an illustration of further potential metrics for the operating characteristics of a predictive model of the application 104 for back surgery. It illustrates relative risk as a function of different cut-points that may be selected.
  • FIG. 9 is an exemplary graph of sensitivity versus positive predictive value for an exemplary back surgery predictive model according to an illustrative embodiment of the invention and illustrates examples of the potential flexibility and tradeoffs available with a predictive model of the application 104 for back surgery by comparing sensitivity against positive predictive value. As indicated in FIG. 8, different cut-points can be selected for different operational aims, goals or constraints (such as objectives, resources, and expected return).
  • FIG. 10 is an exemplary graph of average cost in Year 2 versus predicted average cost based on Year 1 claims according to an illustrative embodiment of the invention and provides a graphical illustration of the predicted cost as modeled by an exemplary financial risk model of the application 104. In one embodiment, an R-square value of 0.30 is reflective of standard model parameters. In this exemplary illustration, the estimates are based on a total population of 2 million members. However, other population sizes may be employed.
  • As evidenced by the foregoing discussion and illustrations, the predictive modeling analytic systems and methods of the invention are useful in a wide range of applications. While this invention has been particularly shown and described with reference to preferred embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the scope of the invention encompassed by the appended claims. In certain embodiments, the term ‘event risk’ may mean and/or refer to a “risk event,” and/or refer to the risk of the occurrence of a risk event.
  • It will be apparent to those of ordinary skill in the art that methods involved in the present invention may be embodied in a computer program product that includes a computer usable and/or readable medium. For example, such a computer usable medium may consist of a read only memory device, such as a CD ROM disk or conventional ROM devices, or a random access memory, such as a hard drive device or a computer diskette, or flash memory device having a computer readable program code stored thereon.

Claims (55)

  1. 1. A system for predicting healthcare risk events comprising:
    a computer;
    a computer readable medium, operatively coupled to the computer, the computer readable medium program codes performing functions comprising:
    accessing patient data associated with one or more patients,
    accessing geographic and healthcare system data,
    filtering the patient data, geographic data, and healthcare system data into clean data, and
    applying a predictive model to the clean data to generate patient profile data and to identify a portion of the patients with a range of susceptibility to one or more risk events.
  2. 2. The system of claim 1 comprising categorizing one or more patients into one or more clinical segments and applying the predictive model to each clinical segment.
  3. 3. The system of claim 2, wherein the segments are based on at least one of preference sensitive conditions, chronic diseases, and large medical cases not associated with chronic disease.
  4. 4. The system of claim 1, comprising generating one or more facts based on the clean data.
  5. 5. The system of claim 1 comprising reporting the portion of patients with a range of susceptibility to one or more risk events to at least one of a healthcare provider, a healthcare insurer, and a payor.
  6. 6. The system of claim 1 comprising generating a suggested intervention plan for one or more patients based on the susceptibility to a risk event.
  7. 7. The system of claim 1, wherein the patient data includes patient claims data and patient non-claims data.
  8. 8. The system of claim 7, wherein patient claims data includes at least one of medical claims data and pharmacy claims data.
  9. 9. The system of claim 7, wherein patient non-claims data includes at least one of referral data, functional status data, laboratory values, patient risk factors, demographics, disease burden, and disease complications.
  10. 10. The system of claim 1, wherein the geographic and healthcare system data include geographic and healthcare practice pattern variables.
  11. 11. The system of claim 1 wherein the geographic and healthcare system data include unwarranted treatment pattern variation data.
  12. 12. The system of claim 1, wherein filtering the patient data, geographic data, and healthcare system data into clean data includes:
    importing patient data files,
    mapping patient data into standard formats,
    processing adjustments and duplicates,
    checking patient data parameters against internal and external normal parameters,
    identifying and correcting data errors, and
    creating a table to link patient data to unique patient identifiers.
  13. 13. The system of claim 1, wherein applying the predictive model includes:
    separating patient data into a first and second data set,
    evaluating regressively one or more risk factors in the first data set to determine weights associated with significant risk factors, and
    applying the weights for each significant risk factor to the second data set to validate the predictive model performance.
  14. 14. The system of claim 1, wherein applying the predictive model includes:
    separating patient data into a first and second data set,
    evaluating regressively one or more risk factors in the first data set to determine weights associated with significant risk factors,
    evaluating regressively one or more risk factors in the second data set to determine weights associated with significant risk factors, and
    comparing the weights of risk factors of the first data set with the weights of the risk factors of the second data set.
  15. 15. The system of claim 13 comprising applying at least one of a linear regression model, a non-linear regression model, a logistic regression model, a Bayesian network model, a Markov model, and a propensity score to evaluate the risk factors.
  16. 16. The system of claim 13 comprising adding the weights associated with the risk factors of one or more patient to generate risk totals associated with the one or more patients.
  17. 17. The system of claim 16 comprising identifying a portion of the patients susceptible to one or more risk events by identifying a portion of the one or more patients with the highest risk totals.
  18. 18. The system of claim 13, wherein the weight includes a beta weight.
  19. 19. The system of claim 18 comprising deriving a risk score based at least in part on the beta weight.
  20. 20. The system of claim 19, wherein the risk score includes a value in the range of 0 to 1.
  21. 21. The system of claim 13, wherein validating includes applying a goodness-of-fit measure.
  22. 22. The system of claim 1 comprising refining and validating the risk predictive model by comparing patient profile data for a first portion of patients with patient profile data for a second portion of patients.
  23. 23. The system of claim 22, wherein the refining and validating include:
    dividing the patient profile data into a model development data set and a model validation data set,
    applying weights to the model development data set to profile the model validation set.
  24. 24. The system of claim 23, wherein the data includes one or more variables.
  25. 25. The system of claim 24, wherein a variable includes one of age, gender, log transformed cost for the first period, number of specialist visits, primary care capacity, and a condition flag.
  26. 26. The system of claim 25, wherein a condition flag is associated with a patient clinical condition.
  27. 27. The system of claim 23, wherein the model development data includes patient claims data associated with a first period of time and the model validation data is associated with a second period of time.
  28. 28. The system of claim 1, wherein the portion of the patients with a range of susceptibility to the one or more risk events includes a selected percentage of the patients that are most susceptible to the one or more risk events.
  29. 29. The system of claim 1, wherein the portion of the patients with a range of susceptibility to the one or more risk events includes a selected percentage of the patients that are least susceptible to the one or more risk events.
  30. 30. The system of claim 1, wherein the portion of patients with a range of susceptibility to the one or more risk events includes patients representing a selected spectrum of susceptibility to the one or more risk events.
  31. 31. A method for predicting healthcare risk events comprising:
    accessing patient data associated with one or more patents,
    accessing geographic and healthcare system data,
    filtering the patient data, geographic data, and healthcare system data into clean data, and
    applying a predictive model to the clean data to generate patient profile data and to identify a portion of the patients with a range of susceptibility to one or more risk events.
  32. 32. The method of claim 31 comprising categorizing one or more patients into one or more clinical segments and applying the predictive model to each clinical segment.
  33. 33. The method of claim 32, wherein the segments are based on at least one of preference sensitive conditions, chronic diseases, and large medical cases not associated with chronic disease.
  34. 34. The method of claim 31 comprising generating one or more facts based on the clean data.
  35. 35. The method of claim 31 comprising reporting the portion of patients with a range of susceptibility to one or more risk events to at least one of a healthcare provider, a healthcare insurer, and a payor.
  36. 36. The method of claim 31 comprising generating a suggested intervention plan for one or more patients based on the susceptibility to a risk event.
  37. 37. The method of claim 31, wherein geographic data includes at least one of geographic practice pattern variables and unwarranted geographic treatment pattern variations.
  38. 38. The method of claim 31, wherein healthcare system data includes unwarranted healthcare system treatment pattern variations.
  39. 39. The method of claim 31, wherein filtering the patient data, geographic data, and healthcare system data into clean data includes:
    importing patient data files,
    mapping patient data into standard formats,
    processing adjustments and duplicates,
    checking patient data parameters against internal and external normal parameters,
    identifying and correcting data errors, and
    creating a table to link patient data to unique patient identifiers.
  40. 40. The method of claim 31, wherein applying the predictive model includes:
    separating patient data into a first and second data set,
    evaluating regressively one or more risk factors in the first data set to determine weights associated with significant risk factors, and
    applying the weights for each significant risk factor to the second data set to validate the predictive model performance.
  41. 41. The method of claim 31, wherein applying the predictive model includes:
    separating patient data into a first and second data set,
    evaluating regressively one or more risk factors in the first data set to determine weights associated with significant risk factors,
    evaluating regressively one or more risk factors in the second data set to determine weights associated with significant risk factors, and
    comparing the weights of risk factors of the first data set with the weights of the risk factors of the second data set.
  42. 42. The method of claim 40 comprising applying at least one of a linear regression model, a non-linear regression model, a logistic regression model, a Bayesian network model, a Markov model, and a propensity score to evaluate the risk factors.
  43. 43. The method of claim 40 comprising adding the weights associated with the risk factors associated with one or more patient to generate risk totals associated with the one or more patients.
  44. 44. The method of claim 43 comprising identifying a portion of the patients susceptible to one or more risk events by identifying a portion of the one or more patients with the highest risk totals.
  45. 45. The method of claim 40, wherein the weight includes a beta weight.
  46. 46. The method of claim 45 comprising deriving a risk score based at least in part on the beta weight.
  47. 47. The method of claim 46, wherein the risk score includes a value in the range of 0 to 1.
  48. 48. The method of claim 40, wherein validating includes applying a goodness-of-fit measure.
  49. 49. The method of claim 31 comprising refining and validating the risk predictive model by comparing patient profile data of a first portion of patients with patient profile data of a second portion of patients.
  50. 50. The method of claim 49, wherein the refining and validating include:
    dividing the patient profile data into a model development data set and a model validation data set,
    applying weights to the model development data set to profile the model validation set.
  51. 51. The method of claim 50, wherein the model development data includes patient claims data associated with a first period of time and the model validation data is associated with a second period of time.
  52. 52. The method of claim 31, wherein the portion of the patients with a range of susceptibility to the one or more risk events includes a selected percentage of the patients that are most susceptible to the one or more risk events.
  53. 53. The method of claim 31, wherein the portion of the patients with a range of susceptibility to the one or more risk events includes a selected percentage of the patients that are least susceptible to the one or more risk events.
  54. 54. The method of claim 31, wherein the portion of patients with a range of susceptibility to the one or more risk events includes patients representing a selected spectrum of susceptibility to the one or more risk events.
  55. 55. A computer readable medium having computer readable program codes embodied therein for predicting healthcare risk events, the computer readable medium program codes performing functions comprising:
    accessing patient data associated with one or more patients,
    accessing geographic and healthcare system data,
    filtering the patient data, geographic data, and healthcare system data into clean data, and
    applying a predictive model to the clean data to generate patient profile data and to identify a portion of the patients with a range of susceptibility to one or more risk events.
US11280611 2004-11-16 2005-11-16 Systems and methods for predicting healthcare related risk events Abandoned US20060129427A1 (en)

Priority Applications (4)

Application Number Priority Date Filing Date Title
US62821204 true 2004-11-16 2004-11-16
US62847604 true 2004-11-16 2004-11-16
US62850404 true 2004-11-16 2004-11-16
US11280611 US20060129427A1 (en) 2004-11-16 2005-11-16 Systems and methods for predicting healthcare related risk events

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
US11280611 US20060129427A1 (en) 2004-11-16 2005-11-16 Systems and methods for predicting healthcare related risk events

Publications (1)

Publication Number Publication Date
US20060129427A1 true true US20060129427A1 (en) 2006-06-15

Family

ID=36407724

Family Applications (5)

Application Number Title Priority Date Filing Date
US11280611 Abandoned US20060129427A1 (en) 2004-11-16 2005-11-16 Systems and methods for predicting healthcare related risk events
US11281233 Active 2028-09-03 US8095380B2 (en) 2004-11-16 2005-11-16 Systems and methods for predicting healthcare related financial risk
US13313601 Active US8290789B2 (en) 2004-11-16 2011-12-07 Systems and methods for predicting healthcare risk related events
US13617225 Active US8484085B2 (en) 2004-11-16 2012-09-14 Systems and methods for predicting healthcare risk related events
US13936713 Abandoned US20140046693A1 (en) 2004-11-16 2013-07-08 Systems and methods for predicting healthcare risk related events

Family Applications After (4)

Application Number Title Priority Date Filing Date
US11281233 Active 2028-09-03 US8095380B2 (en) 2004-11-16 2005-11-16 Systems and methods for predicting healthcare related financial risk
US13313601 Active US8290789B2 (en) 2004-11-16 2011-12-07 Systems and methods for predicting healthcare risk related events
US13617225 Active US8484085B2 (en) 2004-11-16 2012-09-14 Systems and methods for predicting healthcare risk related events
US13936713 Abandoned US20140046693A1 (en) 2004-11-16 2013-07-08 Systems and methods for predicting healthcare risk related events

Country Status (4)

Country Link
US (5) US20060129427A1 (en)
EP (1) EP1836674A4 (en)
CA (1) CA2587715A1 (en)
WO (1) WO2006055630A3 (en)

Cited By (82)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030187699A1 (en) * 2001-12-31 2003-10-02 Bonissone Piero Patrone System for rule-based insurance underwriting suitable for use by an automated system
US20040010462A1 (en) * 2002-07-15 2004-01-15 Susan Moon Method and system for a multi-purpose transactional platform
US20060149603A1 (en) * 2005-01-04 2006-07-06 Barbara Patterson Method and system for determining healthcare eligibility
US20060149529A1 (en) * 2005-01-04 2006-07-06 Loc Nguyen Method for encoding messages between two devices for transmission over standard online payment networks
US20070282637A1 (en) * 2006-05-30 2007-12-06 Nigel Smith Method and system using combined healthcare-payment device and web portal for receiving patient medical information
US20080010094A1 (en) * 2006-06-21 2008-01-10 Mark Carlson Distribution of health information for providing health related services
US20080140447A1 (en) * 2006-06-08 2008-06-12 Stacy Pourfallah System and method using extended authorization hold period
US20080146893A1 (en) * 2006-12-13 2008-06-19 Advanced Brain Monitoring, Inc. Apnea risk evaluation system - automated prediction of risk for perioperative complications
US20080235049A1 (en) * 2007-03-23 2008-09-25 General Electric Company Method and System for Predictive Modeling of Patient Outcomes
US20080275729A1 (en) * 2007-04-09 2008-11-06 Nina Mithi Taggart System and method for population health management
US20080319794A1 (en) * 2007-06-20 2008-12-25 Mark Carlson Health information services using phone
US20090222290A1 (en) * 2008-02-29 2009-09-03 Crowe Michael K Methods and Systems for Automated, Predictive Modeling of the Outcome of Benefits Claims
US20090248600A1 (en) * 2008-03-26 2009-10-01 Matthew Bochner Kennel Estimating transaction risk using sub-models characterizing cross-interaction among categorical and non-categorical variables
US20090326976A1 (en) * 2008-06-26 2009-12-31 Macdonald Morris Estimating healthcare outcomes for individuals
US20100057621A1 (en) * 2008-06-30 2010-03-04 Faith Patrick L Payment processing system secure healthcare data trafficking
US7698159B2 (en) 2004-02-13 2010-04-13 Genworth Financial Inc. Systems and methods for performing data collection
US20100100484A1 (en) * 2005-01-04 2010-04-22 Loc Nguyen Product level payment network acquired transaction authorization
WO2010051240A2 (en) * 2008-10-31 2010-05-06 Archimedes, Inc. Individualized ranking of risk of health outcomes
US20100185466A1 (en) * 2009-01-20 2010-07-22 Kenneth Paradis Systems and methods for tracking health-related spending for validation of disability benefits claims
US20100227296A1 (en) * 2009-03-05 2010-09-09 Quantum Dental Technologies Method of assessing oral health risk
US7801748B2 (en) 2003-04-30 2010-09-21 Genworth Financial, Inc. System and process for detecting outliers for insurance underwriting suitable for use by an automated system
WO2010108092A2 (en) * 2009-03-19 2010-09-23 Phenotypeit, Inc. Medical health information system
US7813945B2 (en) 2003-04-30 2010-10-12 Genworth Financial, Inc. System and process for multivariate adaptive regression splines classification for insurance underwriting suitable for use by an automated system
US7818186B2 (en) 2001-12-31 2010-10-19 Genworth Financial, Inc. System for determining a confidence factor for insurance underwriting suitable for use by an automated system
US7844476B2 (en) 2001-12-31 2010-11-30 Genworth Financial, Inc. Process for case-based insurance underwriting suitable for use by an automated system
US7844477B2 (en) 2001-12-31 2010-11-30 Genworth Financial, Inc. Process for rule-based insurance underwriting suitable for use by an automated system
US20100332251A1 (en) * 2006-07-31 2010-12-30 Edward Yanak Electronic payment delivery service
WO2011019998A2 (en) * 2009-08-14 2011-02-17 Visa U.S.A. Inc. Influenza vaccine administration payment device processing
US7895062B2 (en) 2001-12-31 2011-02-22 Genworth Financial, Inc. System for optimization of insurance underwriting suitable for use by an automated system
US7899688B2 (en) 2001-12-31 2011-03-01 Genworth Financial, Inc. Process for optimization of insurance underwriting suitable for use by an automated system
US20110054925A1 (en) * 2009-08-25 2011-03-03 Rayid Ghani Claims analytics engine
US20110054289A1 (en) * 2009-09-01 2011-03-03 Adidas AG, World of Sports Physiologic Database And System For Population Modeling And Method of Population Modeling
US20110071363A1 (en) * 2009-09-22 2011-03-24 Healthways, Inc. System and method for using predictive models to determine levels of healthcare interventions
US20110079648A1 (en) * 2009-10-05 2011-04-07 Stacy Pourfallah Portable prescription transaction payment device
US20110079643A1 (en) * 2009-10-05 2011-04-07 Stacy Pourfallah Prescription sample transaction payment card
US20110131076A1 (en) * 2009-12-01 2011-06-02 Thomson Reuters Global Resources Method and apparatus for risk mining
US20110166872A1 (en) * 2009-08-14 2011-07-07 Cervenka Karen L Auto-substantiation for healthcare upon sponsor account through payment processing system
US8005693B2 (en) 2001-12-31 2011-08-23 Genworth Financial, Inc. Process for determining a confidence factor for insurance underwriting suitable for use by an automated system
US8214314B2 (en) 2003-04-30 2012-07-03 Genworth Financial, Inc. System and process for a fusion classification for insurance underwriting suitable for use by an automated system
US20120296845A1 (en) * 2009-12-01 2012-11-22 Andrews Sarah L Methods and systems for generating composite index using social media sourced data and sentiment analysis
US20120316916A1 (en) * 2009-12-01 2012-12-13 Andrews Sarah L Methods and systems for generating corporate green score using social media sourced data and sentiment analysis
US20130124217A1 (en) * 2011-11-11 2013-05-16 Debra Thesman Health plan rating system improvement program
US8538773B2 (en) 2009-05-27 2013-09-17 Archimedes, Inc. Healthcare quality measurement
US20130282395A1 (en) * 2013-06-18 2013-10-24 Naryan L. Rustgi Medical registry
US8571937B2 (en) 2010-10-20 2013-10-29 Playspan Inc. Dynamic payment optimization apparatuses, methods and systems
US8577803B2 (en) 2011-06-03 2013-11-05 Visa International Service Association Virtual wallet card selection apparatuses, methods and systems
US20130324861A1 (en) * 2012-06-04 2013-12-05 Fujitsu Limited Health condition determination method and health condition determination system
US20140019161A1 (en) * 2012-07-13 2014-01-16 Koninklijke Philips Electronics 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
US8660862B2 (en) 2005-09-20 2014-02-25 Visa U.S.A. Inc. Determination of healthcare coverage using a payment account
US20140095201A1 (en) * 2012-09-28 2014-04-03 Siemens Medical Solutions Usa, Inc. Leveraging Public Health Data for Prediction and Prevention of Adverse Events
US20140108044A1 (en) * 2012-10-12 2014-04-17 Jayaram Reddy Methods and systems for analyzing health risk score and managing healthcare cost
US20140108042A1 (en) * 2012-10-11 2014-04-17 Jayaram Reddy Methods and systems for determining health risk score, risk of hospitalization and forecasting healthcare cost
US20140142978A1 (en) * 2012-11-21 2014-05-22 Welvie, Llc Systems and methods for assessing a likelihood of elective surgery
US20140164007A1 (en) * 2012-12-12 2014-06-12 Debra Thesman Methods for administering preventative healthcare to a patient population
US20140249829A1 (en) * 2013-03-01 2014-09-04 3M Innovative Properties Company Configurable resource utilization determinator and estimator
US8939356B2 (en) 2009-06-08 2015-01-27 Visa International Service Association Portable prescription payment device management platform apparautses, methods and systems
WO2015112832A1 (en) * 2014-01-24 2015-07-30 Children's Hospital Medical Center System for transforming patient medical record data into a visual and graphical indication of patient safety risk
US20150235001A1 (en) * 2014-02-19 2015-08-20 MedeAnalytics, Inc. System and Method for Scoring Health Related Risk
US9117225B2 (en) 2011-09-16 2015-08-25 Visa International Service Association Apparatuses, methods and systems for transforming user infrastructure requests inputs to infrastructure design product and infrastructure allocation outputs
US20150286792A1 (en) * 2014-04-03 2015-10-08 HCMS Group LLC Systems and methods for health risk determination
US20150305688A1 (en) * 2014-04-25 2015-10-29 Wipro Limited Method of determining discharge readiness condition for a patient and system thereof
US20150358341A1 (en) * 2010-09-01 2015-12-10 Phillip King-Wilson Assessing Threat to at Least One Computer Network
US20160055412A1 (en) * 2014-08-20 2016-02-25 Accenture Global Services Limited Predictive Model Generator
US20160063213A1 (en) * 2014-09-03 2016-03-03 Optum, Inc. Healthcare similarity engine
US20160110516A1 (en) * 2014-10-20 2016-04-21 Liwei Ma Dynamic monitoring early cancer risk
US9355393B2 (en) 2011-08-18 2016-05-31 Visa International Service Association Multi-directional wallet connector apparatuses, methods and systems
US9370689B2 (en) 2008-01-07 2016-06-21 The Quantum Group, Inc. System and methods for providing dynamic integrated wellness assessment
US9589266B2 (en) 2011-04-01 2017-03-07 Visa International Service Association Restricted-use account payment administration apparatuses, methods and systems
US9646291B2 (en) 2011-05-11 2017-05-09 Visa International Service Association Electronic receipt manager apparatuses, methods and systems
US9652765B2 (en) 2008-08-26 2017-05-16 Visa International Service Association System and method for implementing financial assistance programs
US9710807B2 (en) 2011-08-18 2017-07-18 Visa International Service Association Third-party value added wallet features and interfaces apparatuses, methods and systems
US9734146B1 (en) 2011-10-07 2017-08-15 Cerner Innovation, Inc. Ontology mapper
US9760871B1 (en) 2011-04-01 2017-09-12 Visa International Service Association Event-triggered business-to-business electronic payment processing apparatuses, methods and systems
US9773212B2 (en) 2011-02-28 2017-09-26 Visa International Service Association Secure anonymous transaction apparatuses, methods and systems
US9830328B2 (en) 2012-02-02 2017-11-28 Visa International Service Association Multi-source, multi-dimensional, cross-entry, multimedia merchant analytics database platform apparatuses, methods and systems
US20170351717A1 (en) * 2016-06-02 2017-12-07 International Business Machines Corporation Column weight calculation for data deduplication
US9953378B2 (en) 2012-04-27 2018-04-24 Visa International Service Association Social checkout widget generation and integration apparatuses, methods and systems
US9953334B2 (en) 2011-02-10 2018-04-24 Visa International Service Association Electronic coupon issuance and redemption apparatuses, methods and systems
US9996838B2 (en) 2011-03-04 2018-06-12 Visa International Service Association Cloud service facilitator apparatuses, methods and systems
US10074081B1 (en) 2009-08-14 2018-09-11 Citicorp Credit Services, Inc. (Usa) Methods and systems for use of a prepaid payment device
US10096022B2 (en) 2011-12-13 2018-10-09 Visa International Service Association Dynamic widget generator apparatuses, methods and systems
US10115087B2 (en) 2017-09-11 2018-10-30 Visa International Service Association Event-triggered business-to-business electronic payment processing apparatuses, methods and systems

Families Citing this family (51)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CA2587715A1 (en) 2004-11-16 2006-05-26 David E. Wennberg Systems and methods for predicting healthcare related risk events and financial risk
US8589274B2 (en) * 2005-07-08 2013-11-19 Open Market Partners, Inc. System and method for managing healthcare costs
US7765117B2 (en) * 2005-07-08 2010-07-27 Open Market Partners, Inc. System and method for managing healthcare costs
CA2624705A1 (en) * 2005-10-03 2007-04-12 Health Dialog Services Corporation Systems and methods for analysis of healthcare provider performance
US20080033852A1 (en) * 2005-10-24 2008-02-07 Megdal Myles G Computer-based modeling of spending behaviors of entities
US20070156551A1 (en) * 2005-12-30 2007-07-05 Smith Thomas L Method of creating and utilizing healthcare related commodoties
US7801812B2 (en) * 2006-03-10 2010-09-21 Vantagescore Solutions, Llc Methods and systems for characteristic leveling
US7711636B2 (en) * 2006-03-10 2010-05-04 Experian Information Solutions, Inc. Systems and methods for analyzing data
US20070226007A1 (en) * 2006-03-27 2007-09-27 Takanobu Osaki Health service support system
US7739129B2 (en) * 2006-04-10 2010-06-15 Accenture Global Services Gmbh Benefit plan intermediary
US20070239492A1 (en) * 2006-04-10 2007-10-11 Sweetland Christopher L Estimating benefit plan costs
US20080015891A1 (en) * 2006-07-12 2008-01-17 Medai, Inc. Method and System to Assess an Acute and Chronic Disease Impact Index
US8036979B1 (en) 2006-10-05 2011-10-11 Experian Information Solutions, Inc. System and method for generating a finance attribute from tradeline data
US7657569B1 (en) 2006-11-28 2010-02-02 Lower My Bills, Inc. System and method of removing duplicate leads
US7778885B1 (en) 2006-12-04 2010-08-17 Lower My Bills, Inc. System and method of enhancing leads
US8606626B1 (en) 2007-01-31 2013-12-10 Experian Information Solutions, Inc. Systems and methods for providing a direct marketing campaign planning environment
US8606666B1 (en) 2007-01-31 2013-12-10 Experian Information Solutions, Inc. System and method for providing an aggregation tool
US7975299B1 (en) 2007-04-05 2011-07-05 Consumerinfo.Com, Inc. Child identity monitor
US7742982B2 (en) 2007-04-12 2010-06-22 Experian Marketing Solutions, Inc. Systems and methods for determining thin-file records and determining thin-file risk levels
US8180713B1 (en) 2007-04-13 2012-05-15 Standard & Poor's Financial Services Llc System and method for searching and identifying potential financial risks disclosed within a document
WO2008147918A3 (en) 2007-05-25 2009-01-22 Experian Information Solutions System and method for automated detection of never-pay data sets
US20080319802A1 (en) * 2007-06-22 2008-12-25 Abraham Jonathan P Long term care underwriting system and method
US20090030756A1 (en) * 2007-07-27 2009-01-29 Bank Of America Corporation Managing Risk Associated with Various Transactions
US8301574B2 (en) 2007-09-17 2012-10-30 Experian Marketing Solutions, Inc. Multimedia engagement study
US20090089190A1 (en) * 2007-09-27 2009-04-02 Girulat Jr Rollin M Systems and methods for monitoring financial activities of consumers
US9690820B1 (en) 2007-09-27 2017-06-27 Experian Information Solutions, Inc. Database system for triggering event notifications based on updates to database records
US7996521B2 (en) * 2007-11-19 2011-08-09 Experian Marketing Solutions, Inc. Service for mapping IP addresses to user segments
US20090144082A1 (en) * 2007-11-30 2009-06-04 Steven Selbst Physician practice management software for maximizing reimbursement rates from payer contracts
WO2009099448A1 (en) * 2008-02-06 2009-08-13 Vantagescore Solutions, Llc Methods and systems for score consistency
US7991689B1 (en) 2008-07-23 2011-08-02 Experian Information Solutions, Inc. Systems and methods for detecting bust out fraud using credit data
US20110173018A1 (en) * 2010-01-11 2011-07-14 Hoffner Ron M System and method for use of prediction market data to generate real-time predictive healthcare models
US20120209620A1 (en) * 2011-02-16 2012-08-16 International Business Machines Corporation Detecting unexpected healthcare utilization by constructing clinical models of dominant utilization groups
US20130282390A1 (en) * 2012-04-20 2013-10-24 International Business Machines Corporation Combining knowledge and data driven insights for identifying risk factors in healthcare
US20130325502A1 (en) * 2012-06-05 2013-12-05 Ari Robicsek System and method for providing syndrome-specific, weighted-incidence treatment regimen recommendations
US20140081659A1 (en) 2012-09-17 2014-03-20 Depuy Orthopaedics, Inc. Systems and methods for surgical and interventional planning, support, post-operative follow-up, and functional recovery tracking
US20140278472A1 (en) * 2013-03-15 2014-09-18 Archimedes, Inc. Interactive healthcare modeling with continuous convergence
CA2909774A1 (en) * 2013-05-01 2014-11-06 3M Innovative Properties Company Determination of potentially preventable healthcare events
US10055547B2 (en) * 2013-06-14 2018-08-21 HCA Holdings, Inc. Intermediate check points in medical care lines for automatically identifying care deficiencies
US20150032467A1 (en) * 2013-07-26 2015-01-29 The Advisory Board Company Systems and methods for performing multidimensional queries on healthcare provider institutional data
US20150106109A1 (en) * 2013-10-15 2015-04-16 Gary And Mary West Health Institute Systems and Methods for Lowering the Cost of Health Care
US10102536B1 (en) 2013-11-15 2018-10-16 Experian Information Solutions, Inc. Micro-geographic aggregation system
EP3084659A4 (en) * 2013-12-17 2017-04-26 Atigeo LLC Method and system for estimating values derived from large data sets based on values calculated from smaller data sets
US20150193699A1 (en) 2014-01-08 2015-07-09 Civitas Learning, Inc. Data-adaptive insight and action platform for higher education
US9576030B1 (en) 2014-05-07 2017-02-21 Consumerinfo.Com, Inc. Keeping up with the joneses
US10007757B2 (en) 2014-09-17 2018-06-26 PokitDok, Inc. System and method for dynamic schedule aggregation
US20170249426A1 (en) * 2014-09-29 2017-08-31 Cornell University A system and methods for managing healthcare resources
US20160180040A1 (en) * 2014-12-23 2016-06-23 Cerner Innovation, Inc. Predicting glucose trends for population management
WO2016118619A1 (en) * 2015-01-20 2016-07-28 PokitDok, Inc. Health lending system and method using probabilistic graph models
US10013292B2 (en) 2015-10-15 2018-07-03 PokitDok, Inc. System and method for dynamic metadata persistence and correlation on API transactions
US10102340B2 (en) 2016-06-06 2018-10-16 PokitDok, Inc. System and method for dynamic healthcare insurance claims decision support
US10108954B2 (en) 2016-06-24 2018-10-23 PokitDok, Inc. System and method for cryptographically verified data driven contracts

Citations (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5778345A (en) * 1996-01-16 1998-07-07 Mccartney; Michael J. Health data processing system
US6061657A (en) * 1998-02-18 2000-05-09 Iameter, Incorporated Techniques for estimating charges of delivering healthcare services that take complicating factors into account
US6223164B1 (en) * 1994-06-23 2001-04-24 Ingenix, Inc. Method and system for generating statistically-based medical provider utilization profiles
US6266645B1 (en) * 1998-09-01 2001-07-24 Imetrikus, Inc. Risk adjustment tools for analyzing patient electronic discharge records
US20010016821A1 (en) * 1999-08-24 2001-08-23 Debusk Brian C. Modular tracking and profiling system
US20020087358A1 (en) * 1998-12-16 2002-07-04 Gilbert Edward H. System, method, and computer program product for processing diagnostic, treatment, costs, and outcomes information for effective analysis and health care guidance
US20020099686A1 (en) * 2000-07-27 2002-07-25 Schwartz Eric L. Method and apparatus for analyzing a patient medical information database to identify patients likely to experience a problematic disease transition
US6484144B2 (en) * 1999-03-23 2002-11-19 Dental Medicine International L.L.C. Method and system for healthcare treatment planning and assessment
US20030050794A1 (en) * 2001-09-07 2003-03-13 Marjorie Keck Hospital emergency department resource utilization and optimization system
US6611846B1 (en) * 1999-10-30 2003-08-26 Medtamic Holdings Method and system for medical patient data analysis
US20030195772A1 (en) * 2000-03-08 2003-10-16 The Haelan Corporation Healthcare management system and method of predicting high utilizers of healthcare services
US20040039600A1 (en) * 2002-08-23 2004-02-26 Kramer Marilyn Schlein System and method for predicting financial data about health care expenses
US20040039710A1 (en) * 2002-08-23 2004-02-26 Mcmillan Benjamin System and method for health care costs and outcomes modeling with timing terms
US6802810B2 (en) * 2001-09-21 2004-10-12 Active Health Management Care engine
US20060173663A1 (en) * 2004-12-30 2006-08-03 Proventys, Inc. Methods, system, and computer program products for developing and using predictive models for predicting a plurality of medical outcomes, for evaluating intervention strategies, and for simultaneously validating biomarker causality

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5018067A (en) 1987-01-12 1991-05-21 Iameter Incorporated Apparatus and method for improved estimation of health resource consumption through use of diagnostic and/or procedure grouping and severity of illness indicators
US5652842A (en) 1994-03-01 1997-07-29 Healthshare Technology, Inc. Analysis and reporting of performance of service providers
US7222079B1 (en) * 1994-06-23 2007-05-22 Ingenix, Inc. Method and system for generating statistically-based medical provider utilization profiles
US6314556B1 (en) 1997-11-07 2001-11-06 Deroyal Business Systems, Llc Modular health care information management system utilizing reusable software objects
US7853456B2 (en) 2004-03-05 2010-12-14 Health Outcomes Sciences, Llc Systems and methods for risk stratification of patient populations
CA2587715A1 (en) 2004-11-16 2006-05-26 David E. Wennberg Systems and methods for predicting healthcare related risk events and financial risk

Patent Citations (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6223164B1 (en) * 1994-06-23 2001-04-24 Ingenix, Inc. Method and system for generating statistically-based medical provider utilization profiles
US5778345A (en) * 1996-01-16 1998-07-07 Mccartney; Michael J. Health data processing system
US6061657A (en) * 1998-02-18 2000-05-09 Iameter, Incorporated Techniques for estimating charges of delivering healthcare services that take complicating factors into account
US6266645B1 (en) * 1998-09-01 2001-07-24 Imetrikus, Inc. Risk adjustment tools for analyzing patient electronic discharge records
US20020087358A1 (en) * 1998-12-16 2002-07-04 Gilbert Edward H. System, method, and computer program product for processing diagnostic, treatment, costs, and outcomes information for effective analysis and health care guidance
US6484144B2 (en) * 1999-03-23 2002-11-19 Dental Medicine International L.L.C. Method and system for healthcare treatment planning and assessment
US20010016821A1 (en) * 1999-08-24 2001-08-23 Debusk Brian C. Modular tracking and profiling system
US6611846B1 (en) * 1999-10-30 2003-08-26 Medtamic Holdings Method and system for medical patient data analysis
US20030195772A1 (en) * 2000-03-08 2003-10-16 The Haelan Corporation Healthcare management system and method of predicting high utilizers of healthcare services
US20020099686A1 (en) * 2000-07-27 2002-07-25 Schwartz Eric L. Method and apparatus for analyzing a patient medical information database to identify patients likely to experience a problematic disease transition
US20030050794A1 (en) * 2001-09-07 2003-03-13 Marjorie Keck Hospital emergency department resource utilization and optimization system
US6802810B2 (en) * 2001-09-21 2004-10-12 Active Health Management Care engine
US20040039600A1 (en) * 2002-08-23 2004-02-26 Kramer Marilyn Schlein System and method for predicting financial data about health care expenses
US20040039710A1 (en) * 2002-08-23 2004-02-26 Mcmillan Benjamin System and method for health care costs and outcomes modeling with timing terms
US20060173663A1 (en) * 2004-12-30 2006-08-03 Proventys, Inc. Methods, system, and computer program products for developing and using predictive models for predicting a plurality of medical outcomes, for evaluating intervention strategies, and for simultaneously validating biomarker causality

Cited By (113)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8793146B2 (en) 2001-12-31 2014-07-29 Genworth Holdings, Inc. System for rule-based insurance underwriting suitable for use by an automated system
US8005693B2 (en) 2001-12-31 2011-08-23 Genworth Financial, Inc. Process for determining a confidence factor for insurance underwriting suitable for use by an automated system
US7895062B2 (en) 2001-12-31 2011-02-22 Genworth Financial, Inc. System for optimization of insurance underwriting suitable for use by an automated system
US7899688B2 (en) 2001-12-31 2011-03-01 Genworth Financial, Inc. Process for optimization of insurance underwriting suitable for use by an automated system
US7818186B2 (en) 2001-12-31 2010-10-19 Genworth Financial, Inc. System for determining a confidence factor for insurance underwriting suitable for use by an automated system
US20030187699A1 (en) * 2001-12-31 2003-10-02 Bonissone Piero Patrone System for rule-based insurance underwriting suitable for use by an automated system
US7844476B2 (en) 2001-12-31 2010-11-30 Genworth Financial, Inc. Process for case-based insurance underwriting suitable for use by an automated system
US7844477B2 (en) 2001-12-31 2010-11-30 Genworth Financial, Inc. Process for rule-based insurance underwriting suitable for use by an automated system
US20040010462A1 (en) * 2002-07-15 2004-01-15 Susan Moon Method and system for a multi-purpose transactional platform
US8412623B2 (en) 2002-07-15 2013-04-02 Citicorp Credit Services, Inc. Method and system for a multi-purpose transactional platform
US9818100B2 (en) 2002-07-15 2017-11-14 Citicorp Credit Services, Inc. (Usa) Method and system for a multi-purpose transactional platform
US7801748B2 (en) 2003-04-30 2010-09-21 Genworth Financial, Inc. System and process for detecting outliers for insurance underwriting suitable for use by an automated system
US7813945B2 (en) 2003-04-30 2010-10-12 Genworth Financial, Inc. System and process for multivariate adaptive regression splines classification for insurance underwriting suitable for use by an automated system
US8214314B2 (en) 2003-04-30 2012-07-03 Genworth Financial, Inc. System and process for a fusion classification for insurance underwriting suitable for use by an automated system
US7698159B2 (en) 2004-02-13 2010-04-13 Genworth Financial Inc. Systems and methods for performing data collection
US8560446B2 (en) 2005-01-04 2013-10-15 Visa U.S.A. Inc. Product level payment network acquired transaction authorization
US8688581B2 (en) 2005-01-04 2014-04-01 Visa U.S.A. Inc. Product level payment network acquired transaction authorization
US20100100484A1 (en) * 2005-01-04 2010-04-22 Loc Nguyen Product level payment network acquired transaction authorization
US20060149603A1 (en) * 2005-01-04 2006-07-06 Barbara Patterson Method and system for determining healthcare eligibility
US20060149529A1 (en) * 2005-01-04 2006-07-06 Loc Nguyen Method for encoding messages between two devices for transmission over standard online payment networks
US8660862B2 (en) 2005-09-20 2014-02-25 Visa U.S.A. Inc. Determination of healthcare coverage using a payment account
US8788284B2 (en) 2006-05-30 2014-07-22 Visa U.S.A. Inc. Method and system using combined healthcare-payment device and web portal for receiving patient medical information
US20070282637A1 (en) * 2006-05-30 2007-12-06 Nigel Smith Method and system using combined healthcare-payment device and web portal for receiving patient medical information
US8660855B2 (en) 2006-06-08 2014-02-25 Visa U.S.A. Inc. System and method using extended authorization hold period
US20080140447A1 (en) * 2006-06-08 2008-06-12 Stacy Pourfallah System and method using extended authorization hold period
US20080010094A1 (en) * 2006-06-21 2008-01-10 Mark Carlson Distribution of health information for providing health related services
US8417543B2 (en) 2006-07-31 2013-04-09 Visa U.S.A. Inc. Electronic payment delivery service
US20100332251A1 (en) * 2006-07-31 2010-12-30 Edward Yanak Electronic payment delivery service
US20080146893A1 (en) * 2006-12-13 2008-06-19 Advanced Brain Monitoring, Inc. Apnea risk evaluation system - automated prediction of risk for perioperative complications
US8333696B2 (en) * 2006-12-13 2012-12-18 Watermark Medical, Inc. Systems and methods for automated prediction of risk for perioperative complications based on the level of obstructive sleep apnea
US20080235049A1 (en) * 2007-03-23 2008-09-25 General Electric Company Method and System for Predictive Modeling of Patient Outcomes
WO2008118571A1 (en) * 2007-03-23 2008-10-02 General Electric Company Method and system for predictive modeling of patient outcomes
US20080275729A1 (en) * 2007-04-09 2008-11-06 Nina Mithi Taggart System and method for population health management
US20080319794A1 (en) * 2007-06-20 2008-12-25 Mark Carlson Health information services using phone
US9370689B2 (en) 2008-01-07 2016-06-21 The Quantum Group, Inc. System and methods for providing dynamic integrated wellness assessment
US20120059677A1 (en) * 2008-02-29 2012-03-08 The Advocator Group, Llc Methods and systems for automated, predictive modeling of the outcome of benefits claims
US20090222290A1 (en) * 2008-02-29 2009-09-03 Crowe Michael K Methods and Systems for Automated, Predictive Modeling of the Outcome of Benefits Claims
US8078569B2 (en) * 2008-03-26 2011-12-13 Fair Isaac Corporation Estimating transaction risk using sub-models characterizing cross-interaction among categorical and non-categorical variables
US20090248600A1 (en) * 2008-03-26 2009-10-01 Matthew Bochner Kennel Estimating transaction risk using sub-models characterizing cross-interaction among categorical and non-categorical variables
US8930225B2 (en) * 2008-06-26 2015-01-06 Evidera Archimedes, Inc. Estimating healthcare outcomes for individuals
US8224665B2 (en) 2008-06-26 2012-07-17 Archimedes, Inc. Estimating healthcare outcomes for individuals
US20090326976A1 (en) * 2008-06-26 2009-12-31 Macdonald Morris Estimating healthcare outcomes for individuals
US20120245953A1 (en) * 2008-06-26 2012-09-27 Macdonald Morris Estimating healthcare outcomes for individuals
US20100057621A1 (en) * 2008-06-30 2010-03-04 Faith Patrick L Payment processing system secure healthcare data trafficking
US9652765B2 (en) 2008-08-26 2017-05-16 Visa International Service Association System and method for implementing financial assistance programs
US8694300B2 (en) 2008-10-31 2014-04-08 Archimedes, Inc. Individualized ranking of risk of health outcomes
WO2010051240A2 (en) * 2008-10-31 2010-05-06 Archimedes, Inc. Individualized ranking of risk of health outcomes
WO2010051240A3 (en) * 2008-10-31 2010-08-12 Archimedes, Inc. Individualized ranking of risk of health outcomes
US20100198571A1 (en) * 2008-10-31 2010-08-05 Don Morris Individualized Ranking of Risk of Health Outcomes
US20100185466A1 (en) * 2009-01-20 2010-07-22 Kenneth Paradis Systems and methods for tracking health-related spending for validation of disability benefits claims
US8224678B2 (en) 2009-01-20 2012-07-17 Ametros Financial Corporation Systems and methods for tracking health-related spending for validation of disability benefits claims
US8439681B2 (en) 2009-03-05 2013-05-14 Quantum Dental Technologies Method of assessing oral health risk
US20100227296A1 (en) * 2009-03-05 2010-09-09 Quantum Dental Technologies Method of assessing oral health risk
WO2010099619A1 (en) * 2009-03-05 2010-09-10 Quantum Dental Technologies Inc. Method of assessing oral health risk
WO2010108092A3 (en) * 2009-03-19 2011-01-13 Phenotypeit, Inc. Medical health information system
WO2010108092A2 (en) * 2009-03-19 2010-09-23 Phenotypeit, Inc. Medical health information system
US20100249531A1 (en) * 2009-03-19 2010-09-30 Hanlon Alaina B Medical health information system
US8538773B2 (en) 2009-05-27 2013-09-17 Archimedes, Inc. Healthcare quality measurement
US8939356B2 (en) 2009-06-08 2015-01-27 Visa International Service Association Portable prescription payment device management platform apparautses, methods and systems
US20110166872A1 (en) * 2009-08-14 2011-07-07 Cervenka Karen L Auto-substantiation for healthcare upon sponsor account through payment processing system
WO2011019998A3 (en) * 2009-08-14 2011-04-28 Visa U.S.A. Inc. Influenza vaccine administration payment device processing
WO2011019998A2 (en) * 2009-08-14 2011-02-17 Visa U.S.A. Inc. Influenza vaccine administration payment device processing
US20110145008A1 (en) * 2009-08-14 2011-06-16 Cervenka Karen L Influenza vaccine administration payment device processing
US10074081B1 (en) 2009-08-14 2018-09-11 Citicorp Credit Services, Inc. (Usa) Methods and systems for use of a prepaid payment device
US8762180B2 (en) * 2009-08-25 2014-06-24 Accenture Global Services Limited Claims analytics engine
US20110054925A1 (en) * 2009-08-25 2011-03-03 Rayid Ghani Claims analytics engine
US20110054289A1 (en) * 2009-09-01 2011-03-03 Adidas AG, World of Sports Physiologic Database And System For Population Modeling And Method of Population Modeling
US20110071363A1 (en) * 2009-09-22 2011-03-24 Healthways, Inc. System and method for using predictive models to determine levels of healthcare interventions
US8413905B2 (en) 2009-10-05 2013-04-09 Visa U.S.A. Inc. Portable prescription transaction payment device
US20110079643A1 (en) * 2009-10-05 2011-04-07 Stacy Pourfallah Prescription sample transaction payment card
US20110079648A1 (en) * 2009-10-05 2011-04-07 Stacy Pourfallah Portable prescription transaction payment device
US20120296845A1 (en) * 2009-12-01 2012-11-22 Andrews Sarah L Methods and systems for generating composite index using social media sourced data and sentiment analysis
US20120316916A1 (en) * 2009-12-01 2012-12-13 Andrews Sarah L Methods and systems for generating corporate green score using social media sourced data and sentiment analysis
US20110131076A1 (en) * 2009-12-01 2011-06-02 Thomson Reuters Global Resources Method and apparatus for risk mining
US9288224B2 (en) * 2010-09-01 2016-03-15 Quantar Solutions Limited Assessing threat to at least one computer network
US20150358341A1 (en) * 2010-09-01 2015-12-10 Phillip King-Wilson Assessing Threat to at Least One Computer Network
US9757644B2 (en) 2010-10-20 2017-09-12 Playspin Inc. Dynamic payment optimization apparatuses, methods and systems
US8571937B2 (en) 2010-10-20 2013-10-29 Playspan Inc. Dynamic payment optimization apparatuses, methods and systems
US9953334B2 (en) 2011-02-10 2018-04-24 Visa International Service Association Electronic coupon issuance and redemption apparatuses, methods and systems
US9773212B2 (en) 2011-02-28 2017-09-26 Visa International Service Association Secure anonymous transaction apparatuses, methods and systems
US9996838B2 (en) 2011-03-04 2018-06-12 Visa International Service Association Cloud service facilitator apparatuses, methods and systems
US9760871B1 (en) 2011-04-01 2017-09-12 Visa International Service Association Event-triggered business-to-business electronic payment processing apparatuses, methods and systems
US9589266B2 (en) 2011-04-01 2017-03-07 Visa International Service Association Restricted-use account payment administration apparatuses, methods and systems
US9646291B2 (en) 2011-05-11 2017-05-09 Visa International Service Association Electronic receipt manager apparatuses, methods and systems
US8577803B2 (en) 2011-06-03 2013-11-05 Visa International Service Association Virtual wallet card selection apparatuses, methods and systems
US9355393B2 (en) 2011-08-18 2016-05-31 Visa International Service Association Multi-directional wallet connector apparatuses, methods and systems
US9710807B2 (en) 2011-08-18 2017-07-18 Visa International Service Association Third-party value added wallet features and interfaces apparatuses, methods and systems
US9959531B2 (en) 2011-08-18 2018-05-01 Visa International Service Association Multi-directional wallet connector apparatuses, methods and systems
US9117225B2 (en) 2011-09-16 2015-08-25 Visa International Service Association Apparatuses, methods and systems for transforming user infrastructure requests inputs to infrastructure design product and infrastructure allocation outputs
US9734146B1 (en) 2011-10-07 2017-08-15 Cerner Innovation, Inc. Ontology mapper
US20130124217A1 (en) * 2011-11-11 2013-05-16 Debra Thesman Health plan rating system improvement program
US10096022B2 (en) 2011-12-13 2018-10-09 Visa International Service Association Dynamic widget generator apparatuses, methods and systems
US10013423B2 (en) 2012-02-02 2018-07-03 Visa International Service Association Multi-source, multi-dimensional, cross-entity, multimedia analytical model sharing database platform apparatuses, methods and systems
US9830328B2 (en) 2012-02-02 2017-11-28 Visa International Service Association Multi-source, multi-dimensional, cross-entry, multimedia merchant analytics database platform apparatuses, methods and systems
US9953378B2 (en) 2012-04-27 2018-04-24 Visa International Service Association Social checkout widget generation and integration apparatuses, methods and systems
US20130324861A1 (en) * 2012-06-04 2013-12-05 Fujitsu Limited Health condition determination method and health condition determination system
US20140019161A1 (en) * 2012-07-13 2014-01-16 Koninklijke Philips Electronics 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
US20140095201A1 (en) * 2012-09-28 2014-04-03 Siemens Medical Solutions Usa, Inc. Leveraging Public Health Data for Prediction and Prevention of Adverse Events
US20140108042A1 (en) * 2012-10-11 2014-04-17 Jayaram Reddy Methods and systems for determining health risk score, risk of hospitalization and forecasting healthcare cost
US20140108044A1 (en) * 2012-10-12 2014-04-17 Jayaram Reddy Methods and systems for analyzing health risk score and managing healthcare cost
US20140142978A1 (en) * 2012-11-21 2014-05-22 Welvie, Llc Systems and methods for assessing a likelihood of elective surgery
US20140164007A1 (en) * 2012-12-12 2014-06-12 Debra Thesman Methods for administering preventative healthcare to a patient population
US20140249829A1 (en) * 2013-03-01 2014-09-04 3M Innovative Properties Company Configurable resource utilization determinator and estimator
US20130282395A1 (en) * 2013-06-18 2013-10-24 Naryan L. Rustgi Medical registry
WO2015112832A1 (en) * 2014-01-24 2015-07-30 Children's Hospital Medical Center System for transforming patient medical record data into a visual and graphical indication of patient safety risk
US20150235001A1 (en) * 2014-02-19 2015-08-20 MedeAnalytics, Inc. System and Method for Scoring Health Related Risk
US20150286792A1 (en) * 2014-04-03 2015-10-08 HCMS Group LLC Systems and methods for health risk determination
US20150305688A1 (en) * 2014-04-25 2015-10-29 Wipro Limited Method of determining discharge readiness condition for a patient and system thereof
US20160055412A1 (en) * 2014-08-20 2016-02-25 Accenture Global Services Limited Predictive Model Generator
US20160063213A1 (en) * 2014-09-03 2016-03-03 Optum, Inc. Healthcare similarity engine
US20160110516A1 (en) * 2014-10-20 2016-04-21 Liwei Ma Dynamic monitoring early cancer risk
US20170351717A1 (en) * 2016-06-02 2017-12-07 International Business Machines Corporation Column weight calculation for data deduplication
US10115087B2 (en) 2017-09-11 2018-10-30 Visa International Service Association Event-triggered business-to-business electronic payment processing apparatuses, methods and systems

Also Published As

Publication number Publication date Type
US20060129428A1 (en) 2006-06-15 application
US8290789B2 (en) 2012-10-16 grant
CA2587715A1 (en) 2006-05-26 application
US20140046693A1 (en) 2014-02-13 application
WO2006055630A3 (en) 2009-04-02 application
EP1836674A2 (en) 2007-09-26 application
US8484085B2 (en) 2013-07-09 grant
US20130013336A1 (en) 2013-01-10 application
US8095380B2 (en) 2012-01-10 grant
US20120078656A1 (en) 2012-03-29 application
WO2006055630A2 (en) 2006-05-26 application
EP1836674A4 (en) 2009-12-16 application

Similar Documents

Publication Publication Date Title
Cox et al. Good research practices for comparative effectiveness research: approaches to mitigate bias and confounding in the design of nonrandomized studies of treatment effects using secondary data sources: the International Society for Pharmacoeconomics and Outcomes Research Good Research Practices for Retrospective Database Analysis Task Force Report—Part II
Jencks et al. Quality of medical care delivered to Medicare beneficiaries: a profile at state and national levels
Finkelstein et al. Underuse of controller medications among Medicaid-insured children with asthma
Daviglus et al. Relation of body mass index in young adulthood and middle age to Medicare expenditures in older age
Booth et al. Relationship between avoidable hospitalizations for diabetes mellitus and income level
Payne Identifying and managing inappropriate hospital utilization: a policy synthesis.
Rubin et al. Clinical and economic impact of implementing a comprehensive diabetes management program in managed care
Allison et al. School-based health centers: improving access and quality of care for low-income adolescents
González et al. Depression care in the United States: too little for too few
Tamblyn et al. Association between licensure examination scores and practice in primary care
Jong et al. Prognosis and determinants of survival in patients newly hospitalized for heart failure: a population-based study
Mojtabai et al. Medication costs, adherence, and health outcomes among Medicare beneficiaries
Motheral et al. A checklist for retrospective database studies—report of the ISPOR Task Force on Retrospective Databases
Dorsey et al. Diagnosis, evaluation, and treatment of childhood obesity in pediatric practice
Wharam et al. Emergency department use and subsequent hospitalizations among members of a high-deductible health plan
Ho The welfare effects of restricted hospital choice in the US medical care market
US5835897A (en) Computer-implemented method for profiling medical claims
Masoudi et al. The complexity and cost of drug regimens of older patients hospitalized with heart failure in the United States, 1998-2001
US5976082A (en) Method for identifying at risk patients diagnosed with congestive heart failure
Gilmer et al. The medicaid R x model: pharmacy-based risk adjustment for public programs
Burt et al. Which physicians and practices are using electronic medical records?
US20080126131A1 (en) Predictive Modeling And Risk Stratification Of A Medication Therapy Regimen
Mapel et al. Health care utilization in chronic obstructive pulmonary disease: a case-control study in a health maintenance organization
US20080103834A1 (en) Method and apparatus of providing a radiation scorecard
McGlynn Six challenges in measuring the quality of health care

Legal Events

Date Code Title Description
AS Assignment

Owner name: HEALTH DIALOG SERVICES CORPORATION, MASSACHUSETTS

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:WENNBERG, DAVID E.;REEL/FRAME:019378/0876

Effective date: 20060918