| 1.Objective1.1 To investigate the incidence rate of early deterioration of Acute Ischemic Stroke(AIS)in a third-class hospital in shenzhen.To explore early neurological deterioration(END)of influencing factors and provide the basis for intervention strategy AIS patients.1.2 To verify the predictive value of WORSEN rating scale and clinical indicators on patients with END,in order to provide reliable clinical indicators for the sensitivity of early warning model.1.3 To build early warning model of AIS deterioration,to improve the situation of clinical AIS patients END evaluation,promote the development of END early warning assessment tool research.2.MethodsThis is a retrospective study,Conveniently selected 402 patients with acute ischemic stroke(AIS)from a Shenzhen third-class hospital neurology department during January 2014 to December 2017.The 300 patients formed modeling group during January 2014 to December2016,the 100 patients formed validation group during January 2017 to December 2017.In the first stage,a research team was set up to design a questionnaire and extract clinical data.Set up project team and self-designed questionnaire including general information,vital signs of admission,baseline score of National Institutes of Health Stroke Scale(NIHSS),medical history laboratory indicators,cervical ultrasound and imaging results.The project leader and neurology experts conducted unified training for the members of the research team,so that the researchers can understand the collected indicators in this study and fill in the case review information collection form independently according to the unified standards in the process of data collection.The researchers screened patients with the first hospital diagnosed as cerebral infarction or ICD coding for the international classification of diseases I63 retrospectively,checked the discharge summary record,NIHSS score and the rescue records.According to the records,WORSEN score was performed. In the second stage,the clinical data of 402 patients were analyzed and mined.Univariate analysis was performed to compare the clinical data of two groups,and multivariate regression analysis was performed to analyze independent risk factors of END occurrence.To verify externally prediction ability of WORSEN score.Analysis the association between clinical index,WORSEN score and END through spearman correlation analysis,constructed the warning model through the Logistic regression model;Through using the Area Under the receiver-operating characteristic curve(Area Under the ROC Curve,AUC)to estimate the differentiation degree of the model.Kolmogorov-Smirnov test was used to compare the incidence of END in different score segments of the two groups and judge the calibration ability of the model.Akaike Information Criterion(AIC)was used to evaluate the goodness of fit of the model and select the optimal model.3.Results3.1 The incidence of END in AIS patients and its independent influencing factors.A total of 402 subjects were included in this study,68 cases were END,the incidence was16.9%.After adjusting for related confounding factors,admission NIHSS score(OR= 1.31,95%=1.20-1.43),patients with previous history of atrial fibrillation(OR=3.58,95%=1.24-10.35)and moderate or severe carotid artery stenosis(OR=3.75,95%=1.50-9.38)were independent risk factors for END occurrence(P<0.05).However,after adjusting the severity of the patient’s hospitalization(NIHSS score),White Blood Cell(WBC)level had no statistically significant effect on the risk of END(OR=1.64,95%CI = 0.59-4.55,P=0.341).There were saturation threshold effect relationship between Serum triglycerides(TG)levels and END.When TG level was less than 3.8mmol/L,it has a protective effect on the occurrence of END(OR=0.84,95%CI=0.54-1.30,P=0.437);When TG level was more than or equal to3.8mmol/L,the TG level had a more significant impact on the prognosis(OR=1.64,95%CI=1.08-2.49,P=0.020),and the trend was statistically significant(P=0.045).Blood Uric Acid(UA)level was the protective factor of END,and there was a population difference.Compared with the low-score array in model 3,the NIHSS score of patients with the high-score array of UA was reduced by 2.16 points [β=-2.16,95%CI=(-3.53,-0.78),P=0.002],and the risk of END was reduced by 60%(OR=0.40,95%CI= 0.16-0.97,P=0.042),the trend test was statistically significant(P<0.05);Subgroup analysis found that UA level was negatively correlated with the severity of the disease in male patients newly diagnosed with stroke.With the increase of UA level,short-term clinical prognosis was better,and the differences were statistically significant(P<0.05).3.2.The correlation and predictive value between WORSEN score,clinical biochemical indexes and END.The WORSEN scores were positively correlated(r = 0.390,P < 0.001),The sensitivity to predict the occurrence of END was 68.0%,and the specificity was 76.8%,three points was the best cut-off point according to the Youden’s index(0.448).The AUC > 0.75 of WORSEN score predicts the END in two groups,its predicted ability was good and better fitting(x~2 =7.000,P=0.321).Univariate analysis was used in the modeling group,the clinical biochemical index of C-reactive protein(CRP)was positively correlated with END(r=0.363,P<0.001),which AUC was 0.72 when predicting the occurrence of END.Blood uric acid was negatively correlated with END(r=-0.183,P=0.004),which AUC was 0.64 when predicting the occurrence of END.3.3 AIS patients with END early warning model constructionLogistic stepwise regression was used to establish three warning models,model score was formed finally according to the regression coefficient,Model 1 =-3.564 +0.672* WORSEN +0.044*CRP(AIC=63.51),Model2=-2.183+0.782*WORSEN-0.003*UA(AIC=195.15);Model 3=-3.049+0.776 * WORSEN + 0.105* CRP-0.003 *UA(AIC= 50.73),The optimal Model 3was selected to be superior to other models based on AIC.Model 3(AUC = 0.850)and WORSEN score(AUC = 0.832)was statistically significant difference(Z = 1.989,P = 1.989).4.Conclusion4.1 The incidence of clinical END was 16.9%.Multivariate analysis showed that independent risk factors for END included NIHSS score,previous history of atrial fibrillation,moderate to severe carotid artery stenosis and TG levels was greater than 3.8mmol/L.In male patients with new stroke,high UA levels was the protective factor of END,which guides clinical nursing staff to strengthen condition observation in the acute phase,evaluate the degree of neurological injury of patients every day,and implement super-early stepwise rehabilitation training to promote neurological function recovery after the vital signs are stable.Early admission screening for atrial fibrillation,and actively implement relevant laboratory tests;For patients with moderate or severe stenosis of the artery,we should be vigilant.In the AIS acute stage,the UA levels should not be excessively reduced in male newly diagnosed patients.4.2 WORSEN score was positively correlated with the incidence of END,3 points was the optimal threshold to distinguish the progression of the disease.It provides a theoretical basis for clinical nurses to evaluate patients’ condition change and to avoid the risk factors of disease progression.4.3 Clinical biochemical indicators such as CRP and UA,which can improve the ability of END diagnosis and prediction of END early warning,suggesting that medical staff should pay attention to monitoring the dynamic changes of patients’ biochemical indicators,and provide theoretical basis for the evaluation of patients’ conditions. |
