CN115223706A - Sepsis early stage screening model suitable for mobile monitoring equipment - Google Patents

Abstract

The invention discloses a sepsis early stage screening model suitable for mobile monitoring equipment, and relates to the technical field of sepsis screening. The early screening model of pyemia suitable for mobile monitoring equipment, mobile monitoring equipment include consciousness assessment module, electrocardio guardianship equipment, body temperature check out test set, urine volume check out test set and leucocyte count equipment, and mobile monitoring equipment still includes data processing equipment, and data processing equipment is respectively with consciousness assessment module, electrocardio guardianship equipment, body temperature check out test set, urine volume check out test set and leucocyte count equipment electric connection, and the data processing equipment of telling carries out SIRS and calculates the grade. The invention adopts SIRS scoring to diagnose sepsis with different weights in each item; the diagnosis model which is constructed by collecting noninvasive physiological parameters and using logistic regression can be used for early screening of sepsis, the accuracy of the diagnosis model is superior to SIRS scoring, and the monitoring equipment is moved to assist in continuous monitoring, so that the screening effect of sepsis can be improved, and the cost of manpower can be saved to a certain extent.

Description

Sepsis early stage screening model suitable for mobile monitoring equipment

Technical Field

The invention relates to the technical field of sepsis screening, in particular to a sepsis early-stage screening model suitable for mobile monitoring equipment.

Background

Systemic inflammatory response syndrome scores have been widely used clinically in sepsis diagnostics, and have been limited by their higher sensitivity and poorer specificity. The 2016 international conference on sepsis consensus reformulated the sepsis diagnostic criteria and proposed a rapid sequential organ failure score for rapid screening of sepsis. However, the global consensus meeting for sepsis in 2021, alone, was not recommended for sepsis screening due to the poor sensitivity of the qsfa score.

Different weights of each parameter in the SIRS score for screening sepsis were compared by retrospectively analyzing sepsis patients within 24 hours of the ICU in the existing critical care medical information database iv. A new scoring system is constructed by non-invasive physiological parameters suitable for mobile monitoring equipment, and the accuracy of early screening of sepsis is analyzed.

Disclosure of Invention

The present invention aims to provide an early sepsis screening model suitable for mobile monitoring devices to solve the problems set forth in the background above.

In order to achieve the purpose, the invention provides the following technical scheme: the early sepsis screening model is suitable for mobile monitoring equipment, the mobile monitoring equipment comprises an awareness evaluation module, electrocardiogram monitoring equipment, body temperature detection equipment, urine volume detection equipment and leukocyte counting equipment, the mobile monitoring equipment further comprises data processing equipment, the data processing equipment is electrically connected with the awareness evaluation module, the electrocardiogram monitoring equipment, the body temperature detection equipment, the urine volume detection equipment and the leukocyte counting equipment, the data processing equipment calculates SIRS scores, the SIRS scores are weighted by adopting logistic regression analysis SIRS scores, new parameters are introduced to establish newSIRS scores by using logistic regression, and the SIRS scores and the newSIRS scores are compared.

Still further, the electrocardiographic monitoring device comprises a motor patch for detecting an electrocardiogram; the heart rate blood oxygen hand clip is used for monitoring heart rate and blood oxygen concentration, and the blood pressure meter is used for monitoring systolic pressure.

Furthermore, the electrocardiographic monitoring device further comprises a display screen for displaying electrocardiographic data, heart rate data, blood oxygen data and systolic pressure data collected by the electrode patch, the heart rate blood oxygen hand clip and the sphygmomanometer.

Furthermore, the data processing device comprises an information receiving module, an information storage module is connected to the lower layer of the information receiving module, the data processing module further comprises an information processing host used for processing data inside the information storage module, and an information output module used for outputting the data processed by the information processing module is arranged.

Furthermore, the information storage module comprises an information storage database, a data table is established in the information storage database, and fields of the data table comprise consciousness change, blood oxygen, systolic pressure, body temperature grading, urine volume grading and white blood cell number.

Still further, the new parameters introduced by the newSIRS score include changes in consciousness, body temperature, systolic blood pressure, and urine volume.

Still further, the parameters of the SIRS score include white blood cell count and body temperature.

Compared with the prior art, the invention has the beneficial effects that:

the sepsis early-stage screening model suitable for the mobile monitoring equipment adopts SIRS scoring to diagnose different sepsis weights for each item; the diagnosis model which is constructed by collecting noninvasive physiological parameters and using logistic regression can be used for early screening of sepsis, the accuracy of the diagnosis model is superior to SIRS scoring, and the monitoring equipment is moved to assist in continuous monitoring, so that the screening effect of sepsis can be improved, and the cost of manpower can be saved to a certain extent.

Drawings

FIG. 1 is a schematic diagram of a mobile monitoring device according to the present invention;

FIG. 2 is a schematic view of the electrocardiographic monitoring apparatus of the present invention;

FIG. 3 is a schematic diagram of an information processing apparatus according to the present invention;

FIG. 4 is a flow chart of a study in an embodiment;

FIG. 5 is a profile of a SIRS score for a patient in accordance with an embodiment;

FIG. 6 is a ROC curve for screening sepsis and predicting 28-day mortality in a newSIRS scoring model in an exemplary embodiment.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without making any creative effort based on the embodiments in the present invention, belong to the protection scope of the present invention.

It should be noted that in the description of the present invention, the terms "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. indicate orientations or positional relationships based on the orientations or positional relationships shown in the drawings, which are only for convenience of description and simplification of description, and do not indicate or imply that the referred device or element must have a specific orientation, be configured in a specific orientation, and be operated, and thus, should not be construed as limiting the present invention.

Further, it will be appreciated that the dimensions of the various elements shown in the figures are not drawn to scale, for ease of description, e.g., the thickness or width of some layers may be exaggerated relative to other layers.

It should be noted that like reference numerals and letters refer to like items in the following figures, and thus, once an item is defined or illustrated in one figure, it will not need to be further discussed or illustrated in detail in the description of the following figure.

Examples

Retrospective study of infected patients in the intensive care medical information database IV who were transferred to the ICU for 24 hours, the patients were classified into sepsis and non-sepsis groups according to the presence or absence of sepsis. Patient baseline information and prognosis were collected and a SIRS score calculated 24 hours into the ICU. Weights for diagnosing sepsis using logistic regression analysis SIRS scores for each parameter. And (3) incorporating 4 parameters of change of consciousness, body temperature, systolic pressure and urine volume, and establishing a newSIRS sepsis screening model by using logistic regression. Calculate newSIRS score to screen sepsis and predict specificity and sensitivity for 28-day mortality, plot the subject working characteristics, and compare to the SIRS score.

A total of 53150 hospitalizations were screened and included in 23681 patients infected within 24 hours of the ICU and 18277 patients with sepsis, of which 2471 patients died on day 28. And analyzing the SIRS score by adopting a logistic regression model, wherein the maximum weight scoring items are body temperature and white blood cell count. The newSIRS logistic regression formula for the incorporated, body temperature, systolic pressure, urine volume is Σ β iXi =2.286+1.204 (change of consciousness: 0/1) -0.022 (systolic pressure: continuous variable) +0.288 (body temperature fractionation: 0/1) +0.224 (urine volume fractionation: 0/1/2), sepsis risk =1/[1+ [ EXP (- ∑ β iXi) ]. Its diagnostic sepsis ROC area is greater than the SIRS score [0.700 (95% CI 0.692, 0.708) vs.0.573 (95% CI 0.565, 0.582), P < 0.05]; the optimal cut-off was 0.77, sensitivity 71.49%, specificity 58.55%. The 28-day mortality ROC area was predicted to be greater than the SIRS score [0.710 (95% CI 0.699, 0.720) vs.0.607 (95% CI 0.596, 0.618), P < 0.05].

SIRS scores differ from item to item in the diagnosis of sepsis; the diagnosis model which is constructed by collecting noninvasive physiological parameters and using logistic regression can be used for early screening of sepsis, has accuracy superior to SIRS score, and can be used for continuous monitoring by mobile monitoring equipment.

As shown in fig. 1, the present invention provides a technical solution: the early sepsis screening model is suitable for mobile monitoring equipment, the mobile monitoring equipment comprises an awareness evaluation module, electrocardio monitoring equipment, body temperature detection equipment, urine volume detection equipment and leukocyte counting equipment, the mobile monitoring equipment further comprises data processing equipment, the data processing equipment is respectively electrically connected with the awareness evaluation module, the electrocardio monitoring equipment, the body temperature detection equipment, the urine volume detection equipment and the leukocyte counting equipment, the data processing equipment calculates SIRS scores, the SIRS scores are analyzed through logistic regression, weights of all parameters of the SIRS scores are introduced, new parameters are introduced, logical regression is used for establishing newSIRS scores, and the SIRS scores and the newSIRS scores are compared.

The electrocardiographic monitoring device shown in fig. 2 and 3 comprises a motor patch for detecting electrocardiogram; good including the heart rate blood oxygen hand clamp that is used for monitoring rhythm of the heart and blood oxygen concentration, and be used for monitoring the sphygmomanometer of systolic pressure, electrocardio guardianship equipment still includes the display screen, the display screen is used for showing by the electrode paster, the heart rate blood oxygen hand clamp and the electrocardiogram data that the sphygmomanometer was collected, heart rate data, blood oxygen data and systolic pressure data, data processing equipment includes information receiving module, information receiving module lower floor is connected with information storage module, data processing module still includes the information processing host computer and is used for handling the inside data of information storage module, and be provided with information output module and export the data after information processing module handles, information storage module includes information storage database, the inside data sheet that has established of information storage database, the field of data sheet is including consciousness change, the blood oxygen, systolic pressure, body temperature is hierarchical, urine volume is hierarchical, the leucocyte number.

New parameters introduced by the newSIRS score include changes in consciousness, body temperature, systolic blood pressure and urine volume, and parameters of the SIRS score include white blood cell count and body temperature.

Application study

Patients with infections or suspected infections within 24 hours of the ICU were housed retrospectively during 2008-2019 as included in the MIMIC-IV database. All patient-related information in the MIMIC-IV database used in this study was anonymous and used with the mobile monitoring device described above.

Inclusion criteria were: infected or suspected infected patients were within 24 hours of entering the ICU. Definition of infection or suspected infection: the patient uses antibiotics and takes a specimen for culture.

Exclusion criteria: patient case data is missing and SIRS score cannot be collected; patients received mechanical ventilation therapy 24 hours into the ICU; patients were admitted to the ICU for 24 hours with vasoactive drugs.

Research method

Grouping: the groups were classified into septic group and non-septic group according to whether the patient was on ICU for 24 hours and whether the diagnostic criteria of Sepsis3.0 were met.

Collecting basic information of the patient such as age, sex, chalcone syndrome index, acute physiological score III score, hospitalization time before ICU admission and the like, and recording the prognosis result of the patient, including 28-day mortality, ICU admission time and hospitalization time. Calculating the SIRS score in 24 hours by an ICU, establishing a logistic regression model according to SIRS score parameters, and analyzing the weight of each parameter for diagnosing the sepsis.

The newSIRS sepsis screening model was established using logistic regression incorporating 4 parameters of change of consciousness, body temperature, systolic blood pressure, urine volume. The area under the working characteristic curve for the subjects screening for sepsis and 28-day mortality was calculated, their specificity and sensitivity calculated, and compared to the SIRS score.

The normal distribution data of the measurement data is described by using an arithmetic mean plus or minus standard deviation, and the non-normal distribution data is described by using a median (quartile interval). The comparison of the rates was performed by the X2 test. The diagnostic model is built using logistic regression, and regression coefficients are calculated. The ROC curve was used to compare the diagnostic test accuracy and the optimal cut-off was determined using the Jordan index. The data were statistically analyzed using SPSS 20.0, plotted using Prism 8.0, and differences of P < 0.05 were statistically significant.

Results

The baseline data and prognosis were recorded in 53150 hospitalizations, which were included in 23681 patients infected within 24 hours of ICU, 5404 patients non-septic patients, and 18277 patients septic. The consciousness change, body temperature, systolic pressure and urine volume of two groups of patients are different (P is less than 0.001). The 28-day mortality rate was significantly higher in sepsis patients than in non-sepsis patients (13.5%/vs.5.1%, P < 0.001), as shown in Table 1 and FIG. 4.

TABLE 1 Baseline data for patients with sepsis and non-sepsis

The logistic regression model analyzed the SIRS scores, with the most heavily weighted scoring items being body temperature (T) and white blood cell count (WBC) and the least heavily weighted scoring items being Respiratory Rate (RR) and Heart Rate (HR). The logistic regression formula is Σ β iXi =0.598+0.093 (HR: 0/1) +0.175 (RR: 0/1) +0.333 (WBC: 0/1) +0.460 (T: 0/1), and the sepsis risk =1/[1+ exp (- ∑ β iXi) ] is shown in table 2 and fig. 5.

TABLE 2SIRS model scoring system

Scoring item	Assignment intervals	P value	Coefficient of regression	Weight scoring
					HR	Greater than 90/min	＜0.001	0.093	1
RR	Greater than 20/min	＜0.001	0.175	2
					WBC	＞12×10 9 L or < 4X 10 9 /L	＜0.001	0.333	4
T	More than 38 ℃ or less than 36 DEG C	＜0.001	0.460	5

* Logistic regression formula:

∑β _i X _i ＝0.598+0.093(HR：0/1)+0.175(RR：0/1)+0.333(WBC：0/1)+0.460(T：0/1)

sepsis risk =1/[1+ exp (- ∑ β) _i X _i )]

The newSIRS logistic regression formula for inclusion of change of consciousness, body temperature, systolic blood pressure, urine volume is Σ β iXi =2.286+1.204 (change of consciousness: 0/1) -0.022 (SBP: continuous variable) +0.288 (grade of body temperature: 0/1) +0.224 (grade of urine volume: 0/1/2), sepsis risk =1/[1+ EXP (- ∑ β iXi) ], as in Table 3.

TABLE 3newSIRS model scoring system

* Logistic regression formula:

∑β _i X _i =2.286+1.204 (change of consciousness: 0/1) -0.022 (systolic blood pressure: continuous variable) +0.288 (body temperature grading: 0/1) +0.224 (urine volume grading: 0/1/2) sepsis risk =1/[1+ exp (— ∑ β) _i X _i )]

Sepsis screening accuracy: the newSIRS model screens for sepsis with a ROC area greater than the SIRS score-diagnosis sepsis with a ROC area greater than the SIRS score [0.700 (95% CI 0.692, 0.708) vs.0.573 (95% CI 0.565, 0.582), P < 0.05]. The optimal cut-off was determined to be 0.77 based on the john index, with 71.49% (70.84%, 72.14%) sensitivity and 58.55% (57.23%, 59.86%) specificity as shown in table 4.

TABLE 4 comparison of SIRS score and newSIRS score models for sepsis screening and 28-day mortality prediction accuracy

For 28-day mortality, the newSIRS model ROC area was greater than the SIRS score [0.710 (0.699, 0.720) vs.0.607 (0.596, 0.618), P < 0.05]. With 0.74 as the optimal cut-off value, 28-day mortality sensitivity was predicted to be 87.56 (86.27%, 88.74%), specificity 32.42% (31.79%, 33.06%).

This protocol assumes that each parameter in the SIRS score is weighted differently in diagnosing sepsis. And (4) after all parameters of the SIRS scoring tool are brought into the SIRS scoring tool again, establishing a logistic regression model, and finding that the scoring items with the maximum weight are body temperature and white blood cells. Thus, patients who are positive for heart rate and respiratory rate, and patients who are positive for body temperature and white blood cell count, who also meet the SIRS score of 2, will have a greater difference in their likelihood of developing sepsis. It is necessary to redesign the scoring tool according to different scoring item weights. The data from this study explain the use of the SIRS score to screen for the causes of the major differences in sepsis when oriented to different patient groups.

The SIRS and qSOFA scoring adopts a set cut-off value, and continuous variables are converted into ordered two-classification variables for scoring. The method is beneficial to simplifying a scoring system and is convenient for a clinician to operate. But will discard more information from the patient. For example, septic patients have systolic pressure drops to 90 or 60mmHg, all at 1 point in the qSOFA score. However, it is not clear to the clinician that the severity of the above two diseases is high or low. Due to the popularization of intelligent mobile terminals in recent years, continuous variables can be directly input into a model for calculation, binary transformation is not needed, and the diagnosis and screening accuracy can be improved. Thus, when redesigning the screening model, the present study provides a balance between making full use of patient information and ease of operation.

The study designed stricter inclusion and exclusion criteria, excluding vasoactive drugs and mechanically ventilated patients when admitted to the ICU. This is because the screening model designed using non-invasive physiological parameters is applicable to pre-hospital, emergency and general ward settings, whereas patients already using vasoactive drugs and mechanical ventilation are necessarily receiving more clinical attention and need to be diagnosed rather than screened. Effective drug and respiratory therapy will improve patient physiological parameters where data does not reflect the severity of sepsis in the patient, and therefore such patients cannot be used in constructing models.

This study combines 5 noninvasive physiological parameters in SIRS and qsfa scores, plus 24-hour urine volume to build a regression model. The heart rate and respiratory rate were found to be weighted very low in the regression model and not statistically different, and were eventually excluded. Systolic blood pressure has been continuously monitored on a variety of mobile monitors, and therefore continuous variables are directly used. For parameters which are inconvenient to carry out continuous monitoring, the body temperature still continues to use the SIRS scoring standard; the urine volume is used as an ordered classification variable according to a current KIDGO scoring standard; the assessment of consciousness state is done using a binary assessment, with or without change of consciousness, without the use of the glasgow coma score or the wound alertness scale score, which is also operable for use by non-medical professionals with simple training.

In the present study, the ROC area for screening sepsis by SIRS score was less than 0.6, and the specificity was poor, and the sensitivity was comparable, lower than the previous study data. Logistic regression analysis finds that the AUC of sepsis screened by the newSIRS model can reach 0.7, and when the cutoff value is 0.77, the sensitivity can reach 71%, and the specificity is 58%, which are both higher than the SIRS score. If more accurate physiological parameters are obtained through prospective research, the logistic regression coefficients of all parameters are corrected, and the screening efficiency of the newSIRS model can be further improved.

Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.

Claims (7)

1. The early screening model of pyemia suitable for removing guardianship equipment, remove guardianship equipment and include consciousness evaluation module, electrocardio guardianship equipment, body temperature check out test set, urine volume check out test set and leucocyte count equipment, its characterized in that: the mobile monitoring equipment further comprises data processing equipment, the data processing equipment is respectively and electrically connected with the consciousness evaluation module, the electrocardio monitoring equipment, the body temperature detection equipment, the urine volume detection equipment and the leucocyte counting equipment, the data processing equipment carries out SIRS score calculation, each parameter weight of the SIRS score is analyzed by adopting logistic regression, new parameters are introduced, the newSIRS score is established by using logistic regression, and the SIRS score and the newSIRS score are compared.

2. An early screening model of sepsis applicable to mobile monitoring devices in accordance with claim 1, wherein: the electrocardiogram monitoring equipment comprises a motor patch for detecting electrocardiogram; it also includes heart rate oximetry hand grips for monitoring heart rate and blood oxygen concentration, and a sphygmomanometer for monitoring systolic blood pressure.

3. An early screening model of sepsis applicable to mobile monitoring devices in accordance with claim 2, wherein: the electrocardiogram monitoring equipment further comprises a display screen, and the display screen is used for displaying electrocardiogram data, heart rate data, blood oxygen data and systolic pressure data which are collected by the electrode patch, the heart rate blood oxygen hand clamp and the sphygmomanometer.

4. The early stage screening model of sepsis suitable for use in a mobile monitoring device of claim 3, wherein: the data processing device comprises an information receiving module, an information storage module is connected to the lower layer of the information receiving module, the data processing module further comprises an information processing host used for processing data inside the information storage module, and an information output module is arranged for outputting the data processed by the information processing module.

5. An early screening model of sepsis applicable to mobile monitoring devices in accordance with claim 4, wherein: the information storage module comprises an information storage database, a data table is established in the information storage database, and fields of the data table comprise consciousness change, blood oxygen, systolic pressure, body temperature grading, urine volume grading and white blood cell number.

6. An early screening model of sepsis applicable to mobile monitoring devices in accordance with claim 5, wherein: the new parameters introduced by the newSIRS score include changes in consciousness, body temperature, systolic blood pressure, and urine volume.

7. An early screening model of sepsis applicable to mobile monitoring devices in accordance with claim 5, wherein: various parameters of the SIRS score include white blood cell count and body temperature.

Images