The Relationship Between Different Classification Of Cerebral Infarction And Short-term Prognosis

Objective Explore the acute cerebral infarction TOAST classification and OCSP classification and MRI, CISS points and short-term prognosis; And further analysis of the different classification prediction effect after the combination, to provide theoretical basis for the prevention and treatment of acute cerebral infarction.Methods Select confirmed by brain MRI of 435 patients with acute cerebral infarction were 48 hours in hospital expert TOAST classification and OCSP classification and MRI,CISS points, collected at the same time the patient’s gender, age, smoking history,drinking history, history of hypertension, diabetes history, history of coronary heart disease(CHD), stroke history, NIHSS score, m RS score data, follow-up after 1 month,NIHSS score and m RS score, observe each parting the correlation of different subtypes and prognosis; And press 1 month when the modified Rankin scale(m RS) were divided into group of 249 patients with good prognosis and bad prognosis group 186 examples,analyze the correlation between single classification and prognosis of cerebral infarction,then four kinds of classification of a variety of combination, from the Angle of the authenticity of the diagnostic test and prediction effect comparing multiple classification or combination, to find the best classification to predict prognosis of cerebral infarction or parting combination.Result 1 Application of single factor Logistic correlation analysis, TOAST classification was associated with a significant 1 month when m RS score(Wald statistic= 57.104, P < 0.001), and its prognosis of various subtypes from good to bad order as:SAO > SOE > LAA > SUE > CE; OCSP classification was associated with a significant1 month when m RS score(Wald statistic = 71.643, P < 0.001), and its prognosis of various subtypes from good to bad order as: LACI > POCI > PACI > TACI; CISS parting was associated with a significant 1 month when m RS score(Wald statistic = 47.438, P <0.001), and its prognosis of various subtypes from good to bad order as: PAD > OE >UE > LAA > CS; MRI classification and 1 month when m RS score also significantly correlated(Wald statistic = 113.253, P < 0.001), and its prognosis of various subtypes from good to bad order as: lacunar infarction group > small infarction in infarction group > > large infarction group. 2 TOAST classification and OCSP classification combination after multiariable Logistic correlation analysis showed that TOAST classification and 1 month when m RS score correlation decreases(Wald statistic =14.563, P = 0.002), the OCSP classification and correlation of 1 month when m RS score also reduce(Wald statistic = 48.678, P < 0.001), after the combination forecast correct percentage is 75.2%; TOAST classification combined with MRI classification, TOAST classification and disappear a month when m RS score correlation(Wald statistic = 8.908,P = 0.063), and MRI classification and 1 month when m RS score still has significant correlation(Wald statistic = 93.646, P < 0.001), after the combination forecast correct percentage is 83.4%; CISS after parting and OCSP classification combination, CISS parting with 1 month when m RS score correlation disappear(Wald statistic = 12.784, P =0.012), while the OCSP classification and correlation of 1 month when m RS score also reduce(Wald statistic = 50.885, P < 0.001), after the combination forecast correct percentage is 75.6%; CISS classification combined with MRI classification, CISS parting with 1 month when m RS score correlation significantly disappear(Wald statistic = 9.397,P= 0.052), and MRI classification and 1 month when m RS score is still significant correlation(Wald statistic = 95.726, P < 0.001), after the combination forecast correct percentage is 84.0%;OCSP classification combined with MRI classification after the OCSP classification and disappear a month when m RS score correlation(Wald statistic =6.453, P = 0.092), and MRI classification and 1 month when m RS score still has significant correlation(Wald statistic =59.597, P < 0.001), after the combination forecast correct percentage is 83.2%.3 By comparing the nine points type or combination of authenticity and the prediction effect, the results showed that the sensitivity of MRI classification(0.900),(0.753),specific degrees about an index(0.653), prediction correct percentage(0.837), Kappa value(0.663), positive predictive value(0.848), negative predictive value(0.830); The sensitivity of MRI classification combined with CISS classification(0.912),(0.753),specific degrees about an index(0.662), the prediction correct percentage(0.844), Kappa value(0.676), positive predictive value(0.864), negative predictive value(0.832), better than other type or combination forecast.Conclusion 1 TOAST classification, OCSP classification, CISS parting and MRI classification can predict the short-term prognosis of cerebral infarction;2 MRI classification than other type or combination can better predict the prognosis of cerebral infarction.