论文部分内容阅读
目的验证急性一氧化碳中毒后迟发性脑病(DEACMP)预测概率方程的可靠程度,评价其对该病的临床预测价值。方法以2001年1月至2014年12月豫北地区3座城市12家医院符合急性一氧化碳中毒(ACMP)诊断标准的752例住院患者为研究对象,采集患者基本信息并随访至清醒后90 d。应用秦洁等1987年10月至1990年3月从中国人民解放军海军总医院收集的被诊断为ACMP的223例患者随访结果分析中所得出的预测发生DEACMP的概率logistic回归方程,计算ACMP患者发生DEACMP的概率。预测结果与实际DEACMP发病率相比较,分析该方程的符合率及临床预测价值。结果 752例ACMP患者在随访期内实际有127例发生DEACMP,发生率16.9%,预测概率≥50%时,实际发生DEACMP者19例,明显低于理论推测发病数49例(χ~2=20.27,P<0.05);预测概率分别≥60.0%、≥80.0%、≥90.0%时,实际发生DEACMP者分别为13、3、2例,明显低于理论推测发病数37、21、8例(χ~2=19.30、25.07、8.10,P<0.05)。随着预测概率的增加,实际发生DEACMP的比率逐渐降低。预测概率越大,患者发生DEACMP假阴性率越大,灵敏度越低,假阳性率越小,特异度越高,诊断符合率越小。结论应用DEACMP预测概率方程预测本研究752例ACMP患者发生DEACMP的概率与实际发病情况符合率很低。
Objective To verify the reliability of predicting the probability equation of delayed encephalopathy after acute carbon monoxide poisoning and evaluate its clinical predictive value. Methods From January 2001 to December 2014, 752 hospitalized patients who met the diagnostic criteria of acute carbon monoxide poisoning (ACMP) in 12 hospitals in 3 cities in the north of Henan Province were selected as research objects. Basic information of patients was collected and followed up to 90 days after awake. Application of Qin Jie and other from October 1987 to March 1990 collected from the People’s Liberation Army Navy General Hospital was diagnosed as ACMP 223 cases were followed up in the analysis of the results obtained predicts the probability of DEACMP logistic regression equation to calculate the occurrence of ACMP patients DEACMP probability. The predicted results compared with the actual incidence of DEACMP analysis of the equation coincidence rate and clinical prediction value. Results In the 752 cases of ACMP, there were actually 127 cases of DEACMP in the follow-up period, the incidence rate was 16.9%. When the predictive probability was 50% or more, 19 cases of DEACMP actually occurred, which was significantly lower than 49 cases (χ ~ 2 = 20.27 , P <0.05). When the predictive probability was respectively ≥60.0%, ≥80.0%, ≥90.0%, the actual occurrence of DEACMP were 13, 3, and 2 cases, respectively, which were significantly lower than the predicted incidence of 37,21 and 8 (χ ~ 2 = 19.30, 25.07, 8.10, P <0.05). As the predicted probability increases, the actual rate of DEACMP decreases. The larger the probability of prediction, the greater the false negative rate of DEACMP, the lower the sensitivity, the smaller the false positive rate and the higher the specificity, the smaller the diagnostic coincidence rate. Conclusion The DEACMP predictive probability equation predicts that the probability of occurrence of DEACMP in 752 ACMP patients in this study is very low with the actual incidence.