| Stroke is the third-leading cause of death in Western industrialized countries and a leading cause of chronic disability. In China, stroke is the second cause for mortality in all diseases. Particularly, ischemic strokes account for 60%～80% of all the strokes. With the rapid increase of aging population in China, stroke has become a critical cause for the health of the old. With the retrospective study based on the hospital, this study showed functional outcome, surviving condition, and analyzed influential factors of ischemic stroke, initially set up both a functional prognosis predictive model and a death predictive model, and validated external truth of models with the help of prospective study. Therefore, scientific evidence would be provided to improve the prognosis and the quality of life for stroke patients.1 Retrospective study of the prognosis and the influential factor of ischemic strokeWe retrospectively reviewed consecutive patients with ischemic stroke admitted to the Neurology Department of Xijing Hospital between January, 2002 and June, 2005. Demographic, clinical and laboratory data were collected with the use of a specially formulated questionnaire by the trained neurologists. Follow-up assessments were performed strictly blinded to the patients'clinical features by telephone interviews or letters. The follow-up was finished in January 2006. The clinical outcome was assessed by two index as the following①the modified Rankin Scale (mRS) was categorized as good (score 0-2) or poor (score 3-6) outcomes.②death. Univariate and multivariate logistic regression analyses were performed to explore functional predictors of ischemic stoke, and a predictive model was set up. The receiver operating characteristic curves was drawn and the cut-off point was calculated by linear interpolation method. Kaplan-Meier methods were used for survival description and comparing the survival rate. Monovariant and multivariate Cox's proportional hazard regression model were used to analyze prognostic factors on mortality. Thus, a prognosis index equation was set up.846 continuous cases were collected together. 169 of them did not meet the inclusion criteria; 60 cases were excluded; 128 cases had missing data. The remaining 489 ischemic stroke patients were included in the study. 328(67.1%)patients were men and 161(32.9%)patients were women. The median age of patients was 65 years (range: 22-93; IQR: 56, 71). The follow-up time was 22.98±9.8 months. 55(11.2%)cases were dropped out. The remaining 434 patients were included in the analysis. 244(56.2%)patients had good outcome and 190(43.8%)had poor outcome. Univariate and multivariate logistic regression analyses showed that the poor outcome was associated with old age (X1) (OR=3.51; 95%CI: 2.10～5.85), lower educational level (X2) (OR=0.69; 95% CI: 0.57～0.83), having stroke history (X3) (OR=2.48; 95%CI:1.44～4.27), and higher NIHSS total scores (X4) (OR=2.62; 95%CI:1.58～4.33).The predictive model was as the following: LogitP= 0.962+1.254X1-0.377X2+0.909X3+0.963X4. The area under the curve was 0.809. The cut-off point was 0.6505. During the period of follow-up, 46 cases died, and 38 of them died of ischemic stroke. The median survival time was 43.36 months. The survival rates of one-year, two-year and three-year were 92.7 %, 89.9% and 84.2%, respectively. Monovariant and multivariant Cox's proportional hazard regression model showed that the risk factors associated with mortality were old age (X1)(HR=1.039;95%CI:1.009～1.070),lower Glasgow scores (X2) (HR=0.748;95%CI:0.693～0.807) and having complication (X3) (HR:3.928;95%CI:2.044～7.549). The personal prognosis index was as the following: PI=0.038 X1-0.291 X2+1.368 X3. The smaller the personal prognosis index was, the better the prognosis was.The results suggested that older age, lower educational level, having stroke history, and higher NIHSS total scores were influential factors of functional outcome of ischemic stroke. The risk factors of mortality of ischemic stroke were old age, lower Glasgow scores and having complication. 2 Prospective study of the prognosis and the influential factor of ischemic strokeWe prospectively collected consecutive patients with ischemic stroke admitted to the Neurology Department of Xijing Hospital between November, 2005 and June, 2006. The ascertainment methods and the outcome were the same as the previous study. The follow-up was finished in March, 2007. Baseline difference between the sample of retrospective study and prospective study was examined by means of the chi-square test or the Mann-Whitney U test. Univariate and multivariate logistic regression analyses were performed to explore functional predictors of ischemic stoke. The probability of the outcome of each patient was calculated according to the predictive model of ischemic stroke based on the retrospective study and compared with the cut-off point to evaluate the prognosis. With the comparison between predictive outcome and factual outcome, the sensitivity and the specificity and other indexes of the model were calculated. The PI of each patient was calculated according to the predictive model of death. The predictive outcome was compared with the factual outcome.267 continuous cases were collected together. 103 of them did not meet the inclusion criteria; 26 cases were excluded; 3 cases had missing data. The remaining 138 ischemic stroke patients were included in the study. 93(67.4%)patients were men and 45(32.6%)patients were women. The median age of patients was 63 years (range: 36～85). The follow-up time was 11.26±3.30 months. 8(5.8%)cases were dropped out. The remaining 130 patients were included in the analysis. 81(62.3%)patients have good outcome and 49(37.7%)have poor outcome.Univariate and multivariate logistic regression analyses showed that the poor outcome was associated with old age(OR=4.354;95%CI: 1.169～16.223), having stroke history(OR=4.205; 95%CI:1.027～17.224), and higher NIHSS total score (OR=2.179;95%CI: 1.652～2.876).The sensitivity and the specificity and the accuracy degree of the predictive model were 79.6%, 65.4%, 70.8% respectively. 9 cases(6.9%)died during the period of follow-up. 49 cases completed 12 months'follow-up. The result of the comparison between the predictive outcome and the factual outcome showed that the higher the personal prognosis index was, the lower the surviving rate was. The model of PI can initially predict the probability of fatality in the first year's follow-up. The probability of fatality in the second and third years'follow-up was to be further observed.The study validated the result of retrospective study and confirmed that old age, having stroke history, and higher NIHSS total score were associated with the poor outcome of ischemic stroke. The functional predictive model was easily manipulated. However, its accuracy needed to be improved. More prospective studies with large scale were needed. And more predictive factors should be included to improve the accuracy of the model.
|
PDF Full Text Request