The Predictive Value Of Commonly Used Scales/scores In Long-term Prognosis Of Ischemic Stroke

Stroke,as a common disease in contemporary society,still constitutes a heavy burden in China despite the great advances in stroke awareness and treatment with increasing medical levels.With a high number of 3.94 million new stroke cases per year in China,the age standardized prevalence is on the rise,reaching 276.7 per 100000 population in 2019.The epidemiological findings in 2019 show that of 28.76 million stroke cases,ischemic stroke is24.18 million,nearly 85% of all strokes are ischemic stroke.Besides,the 2019 epidemiological survey indicates that the number of stroke death cases in China is 2.19 million.And the period between 1990 and 2019 saw mortality increased by 32.3%.Due to its high incidence,disability and lethality the disease has caused heavy burden of care and economy and significant pressure on society and families involved.It is therefore particularly important to provide theoretical support to prevent the disease recurrence,assess prognosis and personalize diagnosis and treatment.And the scientifically predictable analysis of data of the recurrence and prognosis evaluation may enable the patents’ family members informed of early warning information before potential events occur and therefore guide the families of patients to make clinical treatment decisions as soon as possible.Establishing predictive models to evaluate and verify diseases with long-term and far-reaching effects is a crucial task since the recurrent stroke has a high mortality and disability rate and effectively identifying patients with a high risk of recurrent cerebral ischemia will have a profound impact on individuals and society.To date,many risk scores have been developed to predict the risk of recurrence of cardiovascular disease after the first stroke attack.Based on previous studies,the commonly used relapse scales in clinical practice include ESRS and SPI-II scores,which have certain predictive value in previous studies.And the C-statistic(0.53-0.69),which can be used for long-term risk prediction is not effective in predicting short-term risk.Several studies have also shown that ESRS accurately stratifies the associated risk of recurrent stroke or major vascular events.In predicting the risk of short-term recurrent cerebral infarction,ESRS and SPI-II perform poorly,and their effectiveness can be verified within 1-2 years.Further research is needed to demonstrate their effectiveness and clinical practicality over a longer period of time.The significance of accurate prognostic models for predicting acute ischemic stroke patients lies in conveying information to patients and their relatives,thereby providing psychological expectations for prognosis,cultivating hope,confidence,and a positive outlook for the future and life.More importantly,they provide theoretical support for longer-term medical decisions,including management,treatment,and prevention.Previous studies have shown that evaluation scales used to predict the adverse prognosis of ischemic stroke have shown good test efficacy,such as the THRIVE score,i Score score,and ASTRAL score.These scales can quantitatively and qualitatively predict the adverse prognosis of stroke,which play a very important role in clinical work for it can provide clinicians with quantitative assessments beyond diagnosis and treatment experience.However,studies focus on 1-2 years after stroke,but there are fewer validation studies for longer events.This study established an ischemic stroke cohort to verify the effectiveness of the Stroke Prognosis Scale over a longer period of time.This study is divided into three parts,with a summary of each part as follows.Part One Recurrence prediction of Essen Stroke Risk and Stroke Prognostic Instrument-Ⅱ Scores in ischemic stroke: A study of5-year follow-upObjective: To evaluate the predictive accuracy of the Essen Stroke Risk Score and the Stroke Prognostic Instrument-Ⅱ score on the long-term recurrence in Chinese patients with acute ischemic stroke.Methods: Patients with acute ischemic stroke were enrolled and had completed ESRS and SPI-II scores.Patients we divided into low-risk(0-2points)and high-risk(3-9 points)groups according to ESRS,patients were stratified according to Stroke Prognostic Instrument-II score to low-risk(0-3points),medium risk(4-7 points),and high-risk(8-15 points)group.Patients were followed until stroke recurrence and composite endpoint event(stroke recurrence,myocardial infarction or cardiovascular death).We estimated stratified incidence rates and calculated the cumulative risks at 5 years using Kaplan-Meier estimates.We used receiver operating characteristic(ROC)curves to compare the predictive ability of the Essen Stroke Risk Score and Stroke Prognostic Instrument II score.Results:1.A total of 578 patients completed the follow-up.The cumulative IS recurrence rates were 14.7 %(95% CI,11.8%-17.6%)at 1 year,22.9 %(95 %CI,19.4%-26.4%)at 3-year and 32.3%(95% CI,28.2%-36.4%)at5-year.Meanwhile,the cumulative composite vascular event rates were 17.3%(95% CI,14.2%-20.4%)at 1-year,26.7%(95%CI,23.0%-30.4%)at 3-year and 37.9%(95% CI: 33.8%-42.0%)at 5-year,respectively.2.The cumulative risk of all outcomes increased with increasing risk scores.3.AUC for ESRS and SPI-II risk scores were 0.613(95% CI: 0.565-0.661)and 0.613(95% CI: 0.564-0.662)for 5-year stroke recurrence respectively and correspondingly 0.622(95% CI: 0.576-0.668)and 0.627(95% CI:0.581-0.674)for composite endpoint events.Conclusions:1.The incidences at 5 years were approximately two times of those at 1year.2.Both Essen Stroke Risk Score and Stroke Prognostic Instrument-II scores could stratify the risk of 5-year recurrent stroke and combined vascular events.3.In Chinese patients with acute ischemic stroke,Essen Stroke Risk Score and Stroke Prognostic Instrument-II scores could stratify the risk of5-year recurrent stroke and combined vascular events equally.Part Two Comparative study of ASTRAL and THRIVE scores in predicting 5-year unfavorable functional outcome in patients with acute ischemic strokeObjective: ASTRAL and THRIVE scores are of certain value in predicting the 3-month prognosis and 1-year adverse prognosis of patients with acute ischemic stroke,which have been verified externally for many times.However,there are still questions about the applicability of the above prediction scales over a longer period of time.The purpose of this study was to evaluate the accuracy of ASTRAL and THRIVE scores for 5-year poor prognosis.Methods: A prospective database was established.From December 1,2012 to October 31,2014,612 patients with AIS who were hospitalized in the Department of Neurology of the Aerospace Center Hospital(Peking University Aerospace Clinic School of Medicine)were collected.Within one week after admission,baseline data such as age,gender,self-care ability before admission,severity of stroke,level of consciousness,past medical history,blood glucose were registered,and ASTRAL and THRIVE scores were recorded.The end point events were unfavorable functional outcome(defined as modified Rankin Scale 3 to 6)and 5-year mortality.The 5-year follow-up was completed in November 2019.The predictive effects of the two scales were compared by plotting the receiver operating characteristic curve(ROC)of subjects with ASTRAL and THRIVE scores of AIS patients and calculating the area under the curve(AUC),Kaplan-Meier analysis was used to estimate the 5-year survival probability after grouping.COX regression analysis was also conducted to investigate whether each scale is an independent predictor of prognosis.Results: Of 578 patients included in the final count,258(44.6%)had adverse outcomes at 5-year follow-up,including 138(23.8%)deaths.The AUC values of ASTRAL and THRIVE scores in predicting 5-year poor prognosis of AIS were 0.810(95%CI: 0.776~0.841),0.758(95%CI: 0.721~0.792),respectively.The predicted 5-year death AUC values were 0.741(95%CI: 0.704~0.777)and 0.721(95%CI: 0.682~0.757).Pairwise comparison showed that there was a statistically significant difference between the AUC of THRIVE score and ASTRAL score in predicting poor prognosis(P=0.005).When predicting death events in AIS patients,there was no significant difference between the AUC values of ASTRAL and THRIVE scores(P=0.360).HR of ASTRAL for 5-year mortality was 1.080(95% CI: 1.063~1.097).Conclusions: ASTRAL score is a reliable predictor of 5-year adverse prognosis in patients with acute ischemic stroke,which is superior to THRIVE score,and ASTRAL score is an independent predictor of 5-year mortality risk.Part Three Study on the evaluation of i Score score for 5-year prognosis of ischemic strokeObjective: After external validation,the i Score score showed significant consistency in the short-term prognosis of patients with acute ischemic stroke.The purpose of this study was to evaluate the accuracy of the i Score score in predicting 5-year prognosis.Method: A prospective database was established to collect the basic clinical information of 612 AIS patients hospitalized in the Department of Neurology of the Aerospace Center Hospital(Peking University Aerospace Clinic School of Medicine)from December 1,2012 to October 31,2014.IScore score was recorded are and baseline data for patients were collected.The end point events were the poor outcome(defined as modified Rankin Scale,m RS 3-6 point)and death of the patient.For each outcome,the area under the curve was calculated;in addition,a Cox proportional risk analysis was conducted to determine whether the i Score score is an predictor of mortality outcomes.The Kaplan-Meier product limit method was used to assess the 5-year survival probability for each i Score score level.Results:1.The AUC values of the i Score score predicting 5-year poor functional outcomes and 5-year mortality were 0.804(95%,0.768~0.840)and 0.771(0.726~0.816),respectively.2.In Cox proportional risk analysis,the i Score score was independently correlated with 5-year mortality by 1.026(95% CI,1.021-1.030).When analyzed by quintile,compared to quintile 1,the HR for quintiles 2,3,4,and 5was 2.455(95% CI,1.431~4.210),5.771(95%CI,3.408~9.773),8.062(95%CI,4.697~13.839),13.108(95%CI,7.063~24.328),respectively3.During the 5-year follow-up period,the survival probability showed a downward trend with the increase of the i Score score(Log-Rank test,P<0.001).Conclusions:1.The i Score score can reliably predict 5-year functional outcomes and mortality in patients with acute ischemic stroke.2.A high risk rating on the i Score score indicates a high risk of death.