中文摘要：

目的 比较综合评分与CTA光点征在预测颅内自发性出血后再出血风险的临床准确性及可靠性.方法 选择2014年2月～2014年10月在我院神经外科治疗的自发性脑出血病人119例,纳入病人均行头部CTA检查,且24小时内未行手术治疗者,复查头部CT,依据已制定的评分标准对每个病人进行评分,最后比较综合评分与CTA光点征的评估价值.结果 评分5～7分的病人有37例（31.1％）,CTA光点征阳性的病人有48例（40.8％,一致性检验,κ=0.87）.综合评分的阴性预测值、阳性似然比、灵敏度及Youden指数均高于CTA光点征,其特异度、阴性似然比较CTA光点征没有优势.结论 我院设计的综合评分在灵敏度、阴性预测值、Youden指数等方面均较CTA光点征有优势,该综合评分在预测颅内再出血的风险方面具有更好的准确性及可靠性.

英文摘要：

Objective The prospective clinical trial was conducted to make a comparison between a new developed mixed prediction score system and ‘spot sign’ only on the accuracy and reliability to predict hematoma. Methods From February to the end of October 2014, according to the protocol we had designed in advance, 119 patients were included in this trial. The medical records of these patients were collected and arranged by our trained resident doctors, the presence or not of spot sign and 24h follow-up head CT were judged by two masked neuro-radiologists. The score for each item and total points were made by resident doctors. The agreement of interobserver was evaluated by the multirater κ statistics. Results Totally, 37 （31.1%） patients located in 5 to 7 scores subgroup and 48 （40.3%） patients with spot sign were detected Go：0.87）. The negative predictive value （80%）, the positive likelihood ratio （2.59）, the sensitivity （56.7%） and the Youden index （0.35） for the score-system were superior to the that of spot sign only, but the specificity （78.1%）, the negative likelihood ratio （0.55） took no advantages to the spot sign. Conclusions The newly developed score system by our center takes advantages in the sensitivity, negative predictive value, likelihood ratio and Youden in- dex to spot sign alone. The new system might be a more accurate and reliable predictor.