RU2336825C1 - Method for forecasting of insult acuity outcome - Google Patents

Abstract

FIELD: medicine.

SUBSTANCE: probability of case of fatality rate in percentage is defined. Thus sex of patient is defined (SEX), thus male is designated as 1, female - as 2; age (AGE), thus the age of less than 40 years is designated as 1, from 40 till 59 years - as 2, from 60 to 80 as 3, more than 80 years - as 4; presence of arterial hypertension (AH) in the anamnesis, ischemic heart trouble (CVD), atrial fibrillation (AF), myocardium heart attack (Ml), dislipidemia (DLP), diabetes (DM), stress (STRESS), insult (PRESTR), thus presence is designated as 1, absence - as 2. Received anamnestic data and results of examination should be substitute into mathematical model of forecasting of survival rate for patients receiving treatment ambulatory: S(t)=e ^{(0.077971)*((0.705192)^(1+t*(0.539213))-1)*e^Y} where: S(t) - function of survival rate depending on time; e - constant equal to 2.7183 (basis of hyperbolic logarithm); t - time from the beginning of insult acuity (in days), Y_ambulatory=0.023865*SEX+0.553548*AGE+(-0.059296)*AH++0.184393 *CVD++0.387131*AF+(-0.325802)*MI+0.007298*DLP+(-0.019912)^ileDM+0.303960*STRESS++0.068226*PRESTR, then into mathematical model of forecasting of survival rate for patients receiving treatment in hospital Y_hospital=0.23865*SEX+0.553548*AGE+(-0.059296)*AH++0.184393*CVD++0.387131*AF+(-0.325802)*MI+0.007298*DLP+(-0.019912)*DM+0.303960*STRESS++0.068226*PRESTR, where: Y_hospital=0.119477*SEX+0.350032*AGE+0.847160*AH+(-0.282700)*CVD++0.548215*AF+0.820587*MI+(-0.048081)*DLP+0.177812*DM+0.561128*STRESS+0.631101*PRESTR and also probability of survival for 1-28 days of acuity in ambulatory and stationary conditions is defined.

EFFECT: working out of simple and fast means of forecasting of insult acuity outcome.

6 dwg, 4 tbl, 3 ex

Description

The invention relates to medicine, namely to neurology, and can be used to predict the outcome of an acute period of stroke from the 1st to the 28th day.

Methods have been developed and continue to be developed for predicting the long-term outcomes of cerebrovascular diseases with determining the likelihood of the development of primary and repeated cerebrovascular disorders. (Skvortsova V.I., Chazova I.E., Stakhovskaya L.V. Secondary prevention of stroke. - Moscow, 2002, p.10)

The computer program of the Perm State Medical Academy “Brain stroke - risk degree” (Razuvaev VV, Zolotarev OV Journal of Neurology and Psychiatry named after S.S. Korsakov, 1999, No. 6, p. 38-42), A method for predicting the course and outcome of a disease in patients with an acute period of ischemic stroke, including an assessment of the patient's condition (Ch. P. Vorlow, M. Dennis et al. Stroke. A practical guide for managing patients. - St. Petersburg: Polytechnic, 1998, p. .76-93). The disadvantages of these methods include the low reliability of the forecast.

Methods for predicting the adverse outcome of an acute stroke period using neuroimaging for the presence of leukoaraiosis (Thein SS, Hamidon BB, Teh HS, Raymond AA, Leukoaraiosis as a predictor for mortality and morbidity after an acute ischaemic stroke, Singapore Med J., 2007, 48 ( 5), p.396-9). The disadvantage of this method is the use of an expensive technique.

A known method for predicting the course and outcome of an acute period of ischemic stroke by determining the dynamics of succinate dehydrogenase, glycerol-3-phosphate dehydrogenase, acid phosphatase in lymphocytes of peripheral blood on days 1-3, 4-7, 8-14 (application RU 2002134027, published 2004.07. twenty). The authors of the methodology determined an unfavorable or favorable outcome of the disease by changing the content of these enzymes. The disadvantage of this method is its complexity.

A known method for predicting the outcome of ischemic stroke, including magnetic resonance imaging of the brain, proton magnetic resonance spectroscopy of identified pathological and unchanged areas of the brain, determining the content of N-acetylaspartate, choline, creatine and lipids in them (patent RU 2233116, 2004). With a ratio of the content of these metabolites in the pathological focus to their content in the unchanged zone of more than 0.6, a favorable outcome is predicted, with at least one of these relations less than 0.6, an unfavorable outcome. The disadvantage of this method is the use of expensive methods of examination, the ability to conduct research in a hospital or outpatient only in mild condition.

A known method for predicting the course and outcome of the disease in the acute period of ischemic stroke, in which the patient is assessed when the patient is admitted to the hospital on days 7 and 14, RR cardiointervalograms are recorded, their time series are processed, standard deviation of the cardiointervalues is evaluated, fractal indices respectively for the central nervous system and autonomic nervous systems, the potential for self-organization of the system, predict a favorable prognosis, deterioration or critical th threshold, incompatible with life, according to the values of the measured parameters (patent RU 2234852, 2004). The disadvantage of this method is the use in hospital patients.

A known method for predicting the course of ischemic stroke by immunophenotyping of lymphocytes and determining the average level of CD95 lymphocytes at 1, 10, 21 days from the time of the development of the disease (patent RU 2236685, 2004). The dynamics of the content of this indicator predict a favorable outcome, a favorable prognosis for life and an unfavorable prognosis for the restoration of a neurological deficit or an unfavorable prognosis for life. The disadvantage of this method is the complexity, considerable time, the use of expensive reagents.

The prototype of the invention is a method for predicting the outcome of ischemic stroke, which consists in the fact that the patient reveals neurological syndromes, assess them in points, determine the clinical coefficient of ischemic stroke as the ratio of the total scores of the cerebral syndrome complex to the total points of the focal, calculate the probability of mortality in percent according to the formula ( patent RU 2297786, 2007). This method, including the assessment of many parameters, is quite time-consuming and lengthy, does not differentiate the outcome of the disease in outpatient and inpatient settings.

The technical result is a simplification of the method, reducing research time, differential prognosis of the outcome of the disease in an outpatient and inpatient setting.

The proposed method for predicting the outcome of an acute period of stroke is as follows. The patient determines the sex (SEX), while the male gender is designated as 1, female - as 2; age (AGE), while the age of less than 40 years is designated as 1, from 40 to 59 years - as 2, from 60 to 80 as 3, older than 80 years - as 4; history of arterial hypertension (AN), coronary heart disease (CVD), atrial fibrillation (AF), myocardial infarction (MI), dyslipidemia (DLP), diabetes mellitus (DM), stress (STRESS), stroke (PRESTR), with this presence is denoted as 1, absence - as 2. The obtained anamnestic data and examination results are substituted into the mathematical model for predicting survival for patients receiving outpatient treatment:

S (t) = e ^{(0,077971) * ((0,705192) ^ (1 + t * (0,539213)) - 1) * e ^ Y} ,

Where

S (t) is the survival function as a function of time;

e is a constant equal to 2.7183 (the base of the natural logarithm);

t is the time from the onset of the acute period of stroke (in days);

Y _ambulatory = 0.023865 * SEX + 0.553548 * AGE + (- 0.059296) * AH +

0.184393 * CVD + 0.387131 * AF + (- 0.325802) * MI +

0.007298 * DLP + (- 0.019912) * DM + 0.303960 * STRESS +

0.068226 * PRESTR,

then into the mathematical model for predicting survival for patients receiving treatment in a hospital

S (t) = e ^{(0.006787) * ((0.910974) ^ (1 + x * (1.32044)) - 1) * e ^ Y} ,

Where

Y _hospital = 0.119477 * SEX + 0.350032 * AGE + 0.847160 * AH +

(-0.282700) * CVD + 0.548215 * AF + 0.820587 * MI +

(-0.048081) * DLP + 0.177812 * DM + 0.561128 * STRESS +

0.631101 * PRESTR.

After counting, they get the probability of survival in percent for a given day (from 1 to 28). With a value of S (t) = 90-95%, the risk of dying is low, with S (t)> 75% and <90%, the risk of dying is average, with S (t) ≤75%, the risk is high. Such a gradation of the degree of risk was carried out taking into account the fact that the early 30-day mortality after a stroke in Russia is on average 34.6% (Skvortsova V.I., Evzelman M.A. Ischemic stroke. - Orel, 2006, p.6- 10).

Thus, the proposed method allows to determine the likelihood of surviving in the acute period of stroke for every day from 1 to 28 taking into account gender, age, place of treatment and the presence / absence of risk factors.

Using the results of prediction, it is possible to identify patients with a high probability of death, to determine the combination of risk factors at which the probability of survival has significant differences depending on the place of treatment (hospital or outpatient conditions). This will allow you to carefully select the treatment tactics of both the stroke itself and associated diseases.

The proposed method is illustrated by the following figures: figure 1 shows a graph of the survival of patients of the second and third age groups, figure 2 is a graph of the survival of patients depending on gender, figure 3 is a graph of survival depending on the presence of arterial hypertension, in fig. 4 is a graph of patient survival depending on the presence of atrial fibrillation, FIG. 5 is a graph of intensity function h (t) for outpatient treatment conditions, and FIG. 6 is a graph of intensity function h (t) for inpatient treatment conditions.

Information Collection Technique

According to the method proposed by the National Association for the Control of Stroke NABI, each new stroke case was recorded among the permanent residents of the territory selected for maintaining the register (study population of 128,000 people).

Data was recorded by filling out paper registration forms, as well as by maintaining an electronic register form (program proposed by NABI)

Information in the register came from the following sources:

1. ambulance stations (call logs, patient records),

2. from clinics (home call logs, outpatient cards, statistical coupons),

3. from hospitals (journals of admission and discharge of patients, medical history, autopsy protocols)

4. from the bureau of forensic medical examination (in case of sudden death),

5. from the registry office (death certificate, in which any cerebrovascular pathology is indicated as the cause of death)

6. the results of the conversation with the patient’s relatives and with the doctors in contact with the patient.

To be included in the register, a case must meet the criteria for a stroke and occur in a person permanently residing in the study area. The diagnosis of a stroke is established if the patient had a sudden appearance of focal neurological and / or cerebral symptoms, which persisted for more than 24 hours or led to the death of the patient in a shorter period of time.

Types of stroke are established in accordance with the international classification of diseases of the ICD X.

Criteria for exclusion from the register:

1. transient ischemic attacks;

2. stroke caused by blood diseases, brain tumors, brain metastases, stroke caused by trauma;

3. deaths if there is no information in the medical documentation confirming the diagnosis of a stroke;

4. lack of patient passport data (date of birth, address, gender).

When registering, the patient’s passport data, the date of the onset of the stroke, the nature of the stroke, the presence of a stroke in history, the outcome of a stroke in the acute period, place of treatment, risk factors, neuroimaging were taken into account.

A stroke was considered repeated after 28 days from the start of the first stroke. A fatal stroke was recorded if the patient died within 28 days from the onset of the disease.

According to the NABI method, information is recorded on the following risk factors: arterial hypertension, smoking, cardiovascular disease (coronary heart disease), a history of myocardial infarction, atrial fibrillation, diabetes mellitus, dyslipidemia, stress.

The main epidemiological indicators are calculated: incidence, mortality, mortality (in the studied population, these indicators are higher than the national average). A low proportion of patients receiving inpatient treatment (no more than 31%).

Given the above, it becomes relevant to predict the outcome of the acute period of stroke, especially among patients receiving treatment on an outpatient basis.

Using the Kaplan-Mayer survival analysis (Statistica 7.0 application package), the impact on the survival of sex, age, and risk factors listed above was evaluated.

The following results are obtained.

Depending on age, patients are divided into four age groups: group 1 - less than 40 years old, group 2: 40-59 years old, group 3: 60-80 years old, group 4 - older than 80 years. Comparison of patient survival depending on age group, gender, presence of risk factors, place of treatment (hospital, outpatient conditions) is performed.

Statistically significant survival worsen over 60 years of age (Fig. 1), female gender (Fig. 2), outpatient treatment, the presence of such risk factors as heart disease (CHD), atrial fibrillation (Fig. 4), dyslipidemia, stress. An insignificant statistically reduced survival is caused by arterial hypertension (Fig. 3), diabetes mellitus, a history of myocardial infarction. The effect of smoking is mixed and requires further study.

The lack of influence on the survival of diabetes mellitus, myocardial infarction can be explained by active follow-up and adequate treatment of these conditions. The lack of influence of arterial hypertension is probably due to insufficient control of the level of blood pressure.

Patient survival prediction model (Cox model).

A model was developed for predicting patient survival for inpatient and outpatient settings depending on gender, risk factors, and age group.

The Hazard function makes it possible for the patient to calculate the risk of death in the acute period of a stroke.

h (t) = [h ₀ (t)] e ^Y , where

Y is the mortality ratio, Y = Σb _i X _i ;

b _i - beta (coefficient of the basic function of intensity), reflects the weight of the factor in relation to mortality;

X _i is the coded value of the risk factor (SEX, AGE, etc.);

h ₀ (t) is the basic intensity function;

h ₀ (t) = h (t) for X _i = 0 (the values of all risk factors are zero).

The risk factor is a qualitative variable, for its use in the basic intensity function, its value is indicated as follows: for all risk factors except gender and age group “0” - there is no risk factor, “1” - there is a risk factor; the variable "gender": "1" is the male gender and "2" is the female gender. The variable "age group" (AGE) has the following values: "1", "2", "3", "4" ("less than 40 years old", "40-59 years old", "60-80 years old", "older than 80 years, respectively)

The intensity function taking into account the above risk factors took the form:

The beta values for each factor were calculated separately for inpatient and outpatient conditions based on the obtained tables of lifetimes (table 1).

In accordance with the values of beta table 1, the mortality ratio "Y" takes the following form:

for outpatient care

Y _ambulatory = 0.023865 * SEX + 0.553548 * AGE + (- 0.059296) * AH + 0.184393 * CVD + 0.387131 * AF + (- 0.325802) * MI + 0.007298 * DLP +

(-0.019912) * DM + 0.303960 * STRESS + 0.068226 * PRESTR

for stationary conditions

Y _hospital = 0.119477 * SEX + 0.350032 * AGE + 0.847160 * AH +

(-0.282700) * CVD + 0.548215 * AF + 0.820587 * MI +

(-0.048081) * DLP + 0.177812 * DM + 0.561128 * STRESS +

0.631101 * PRESTR.

Next, the intensity function h (t) was calculated for outpatient (Fig. 5) and inpatient treatment conditions (Fig. 6).

The survival function S (t) = e ^{-h (t)} .

Survival Prediction Model for Outpatient Patients:

S (t) = e ^{(0,077971) * ((0,705192) ^ (1 + t * (0,539213)) - 1) * e ^ Y} , where

Y _ambulatory = 0.023865 * SEX + 0.553548 * AGE + (- 0.059296) * AH +

0.184393 * CVD + 0.387131 * AF + (- 0.325802) * MI +

0.007298 * DLP + (- 0.019912) * DM + 0.303960 * STRESS +

0.068226 * PRESTR.

Survival Prediction Model for inpatients:

S (t) = e ^{(0.006787) * ((0.910974) ^ (1 + x * (1.32044)) - l) * e ^ Y} , where

Y _hospital = 0.119477 * SEX + 0.350032 * AGE + 0.847160 * AH +

(-0.282700) * CVD + 0.548215 * AF + 0.820587 * MI + (- 0.048081) * DLP + 0.177812 * DM + 0.561128 * STRESS +

0.631101 * PRESTR.

The prediction model has been successfully tested using the Chi-Square criterion. The effectiveness of the model is statistically significant with a level of p = 0.00001 (Chi-Square = 168.6407, df = 10) for outpatients, a value of p = 0.00048 (Chi-Square = 31.54700, df = 10) for inpatients .

We give examples of clinical use of the proposed method.

Example 1

A male patient, 72 years old (age is indicated by the value “3”, as it falls in the range of 60-80 years) with arterial hypertension, ischemic heart disease, dyslipidemia, has the following probability of surviving the 28th day of the acute period of stroke (Table 2). By the end of the acute period, the probability of surviving on an outpatient or inpatient basis is 62% (high risk) and 96.51% (low risk), respectively.

Example 2

A 65-year-old male patient with arterial hypertension, coronary heart disease, atrial fibrillation, dyslipidemia, and diabetes by the end of the first week of the acute period of a stroke (day 7) has a 57% chance of survival (high risk) with outpatient treatment and 95% (low risk) during treatment in a hospital (table 3).

Table 1 Risk factor Beta value, outpatient treatment Beta value, inpatient care Sex (SEX) 0,023865 0.119477 Age Group (AGE) 0.553548 0.350032 Arterial hypertension (AN) -0,059296 0,847160 Heart Disease (CVD) 0.184393 -0.282700 Atrial fibrillation (AF) 0.387131 0.548215 Myocardial infarction (Ml) -0.325802 0,820587 Dyslipidemia (DLP) 0,007298 -0.048081 Diabetes mellitus (DM) -0.019912 0.177812 Stress (STRESS) 0.303960 0.561128 A history of stroke (PRESTR) 0,068226 0.631101

table 2 Time (day) = 28 Ambulatory treatment Hospital treatment beta factor factor * beta beta factor * beta Floor 0,023865 one 0,023865 0.119477 0.119477 Age group 0.553548 3 1,660644 0.350032 1,050097 Arterial hypertension -0,059296 one -0,0593 0,847160 0.84716 Heart disease 0.184393 one 0.184393 -0.282700 -0.2827 Atrial fibrillation 0.387131 0 0 0.548215 0 Myocardial infarction -0.325802 0 0 0,820587 0 Dyslipidemia 0,007298 one 0,007298 -0.048081 -0.04808 Diabetes -0.019912 0 0 0.177812 0 Stress 0.303960 0 0 0.561128 0 A history of stroke 0,068226 0 0 0.631101 0 ho (t) = 0,077689 0,00659 h (t) = 0.478004 0,035571 S (t) = 0.62002 0.965054

Table 3 Time (day) = 7 Ambulatory treatment Hospital treatment beta factor factor * beta beta factor * beta Floor 0,023865 one 0,023865 0.119477 0.119477 Age group 0.553548 3 1,660644 0.350032 1,050097 Arterial hypertension -0,059296 one -0,0593 0,847160 0.84716 Heart disease 0.184393 one 0.184393 -0.282700 -0.2827 Atrial fibrillation 0.387131 one 0.387131 0.548215 0.548215 Myocardial infarction -0.325802 0 0 0,820587 0 Dyslipidemia 0,007298 one 0,007298 -0.048081 -0.04808 Diabetes -0.019912 one -0.01991 0.177812 0.177812 Stress 0.303960 0 0 0.561128 0 A history of stroke 0,068226 0 0 0.631101 0 ho (t) = 0.938701 0,995833 h (t) = 0.561918 0.046582 S (t) = 0.570114 0.954487

Table 4 Time (day) = 28 Ambulatory treatment Hospital treatment beta factor factor * beta beta factor * beta Floor 0,023865 2 0.04773 0.119477 0.238954 Age group 0.553548 3 1,660644 0.350032 1,050097 Arterial hypertension -0,059296 one -0,0593 0,847160 0.84716 Heart disease 0.184393 one 0.184393 -0.282700 -0.2827 Atrial fibrillation 0.387131 one 0.387131 0.548215 0.548215 Myocardial infarction -0.325802 one -0.3258 0,820587 0,820587 Dyslipidemia 0,007298 one 0,007298 -0.048081 -0.04808 Diabetes -0.019912 one -0.01991 0.177812 0.177812 Stress 0.303960 one 0.30396 0.561128 0.561128 A history of stroke 0,068226 one 0,068226 0.631101 0.631101 ho (t) = 0,077689 0,00659 H (t) = 0.740324 0.620085 S (t) = 0.47696 0.537899

Example 3

A woman aged 55 years (age group from 40 to 59 years, indicated by the value “3”), with all risk factors, can survive the 28th day of the acute period of stroke with a probability of 47.7% when treated as an outpatient and with a probability of 53.79% during treatment in a hospital (Table 4). In both cases, the degree of risk remains high.

Claims (1)

A method for predicting the outcome of an acute period of stroke by determining the probability of mortality in percent, characterized in that the patient determines the sex (SEX), while the male gender is designated as 1, female - as 2; age (AGE), while the age of less than 40 years is designated as 1, from 40 to 59 years - as 2, from 60 to 80 as 3, older than 80 years - as 4; history of arterial hypertension (AN), coronary heart disease (CVD), atrial fibrillation (AF), myocardial infarction (MI), dyslipidemia (DLP), diabetes mellitus (DM), stress (STRESS), stroke (PRESTR), with this presence is designated as 1, absence - as 2, the obtained medical history data and the results of the examination are substituted into the mathematical model for predicting survival for patients receiving outpatient treatment:

where S (t) is the survival function as a function of time; e is a constant equal to 2.7183 (the base of the natural logarithm); t is the time from the onset of the acute period of stroke (in days);

then into the mathematical model for predicting survival for patients receiving treatment in a hospital

Where

and determine the probability of survival on the 1-28 days of the acute period in outpatient and inpatient settings.

Images