对广东省重症严重急性呼吸综合征应用免疫调节剂的多因素回归分析  

Evaluation of the efficacy and safety of corticosteroid in the treatment of severe SARS in Guangdong province with multi-factor regression analysis

作　　者：徐远达[1] 江梅[1] 陈荣昌[1] 方积乾[2] 

机构地区：[1]广州呼吸疾病研究所,510120 [2]中山大学公共卫生学院

出　　处：《中国危重病急救医学》2008年第2期84-87,共4页Chinese Critical Care Medicine

基　　金：广东省广州市科技攻关计划重点项目（2003Z1-E0118）

摘　　要：目的对2002年12月-2003年12月广东省收治的严重急性呼吸综合征（SARS）临床数据库进行分析，评价免疫调节剂对重症SARS患者的作用。方法数据库中共收集了1278例临床诊断SARS和疑似SARS患者的详细资料。首先选取确诊患者402例，再按卫生部重症诊断标准筛选出358例，按激素应用剂量（统一换算为甲泼尼龙）分为小剂量组（〈80mg／d）、中剂量组（80～320mg／d）、大剂量组（≥320mg／d）3组，对是否应用激素做Logistic和COX回归分析。结果①以死亡为因变量的Logistic回归，提示小剂量激素组虽有保护作用，但差异无统计学意义；②以死亡为因变量的COX回归，提示激素对瞬时死亡概率无统计学意义；③以所有生存者住院时间为因变量的COX回归，提示使用中剂量激素者的住院时间是未用激素者的0．619倍（1／1．616，Х^2=7．262，P=0．007），应用免疫调节剂（包括丙种球蛋白、胸腺肽、干扰素等）者的住院时间是未用免疫调节剂者的0．671倍（1／1．491，Х^2=10．252，P=0．001）；④以是否继发肺部或肺外感染为因变量的Logistic回归，提示应用激素的患者感染发生概率是未用激素者的3．095倍（Х^2=4．289，P=0．038）。结论按照卫生部诊断标准确诊的重症SARS患者，使用激素对死亡概率和瞬时的死亡强度无统计学意义，但能显著缩短生存患者的住院时间；同时应用丙种球蛋白、胸腺肽、干扰素等免疫调节剂具有协同作用，但要注意继发的肺内或肺外感染。Objective To investigate the clinical database of severe acute respiratory syndrome （SARS） in Guangdong province and evaluate the efficacy and safety of corticosteroid in the treatment of severe SARS from December 2002 to December 2003. Methods The detail data of 1 278 SARS patients and borderline cases were collected. Four hundred and two confirmed SARS cases were recruited in our study. Out of them, 358 cases were assigned to the severe SARS group based on the criteria issued by Ministry of Health. Subjects who received steroid （which was converted into methylprednisolohe） treatment were further divided into three groups ： small dose group （〈80 mg/day） ;moderate dose group （80 - 320 mg/day） ; high dose group （〈320 mg/day）. A Logistic regression model was applied to investigate the outcome viarables： death, complications, subsequent lung infection and other infections and COX regression was made. Results （1）Small dose of steroid seemed to have protective effect, but it did not reach significant level. （2）COX regression revealed that steroid was not related to instant mortality rate. （3）Length of stay in hospital of patients steroid usage in medium dosage seemed to be 0. 619 time less risky than in patients without steroid usage （Х^2 = 7. 262, P = 0. 007）, and that in patients with immunomodulator （including gamma globulin, thymic peptide and interferon） was 0. 671 time less risky than in patients without immunomodulator （Х^2= 10. 252, P= 0. 001）. （4）Incidence of infections in patients with steroid was 3. 095 times higher than in patients without steroid （Х^2=4. 289, P=0. 038）. Conclusion There is no significant difference in mortality, instant death incidence between steroid treatment and non-steroid treatment group of SARS patients diagnosed with the diagnostic criteria issued by Ministry of Health. However, steroid seem to shorten the length of hospital stay. But attention should be paid that infection rate could be increased in such cases.

关 键 词：严重急性呼吸综合征 激素 LOGISTIC回归 COX回归 

分 类 号：R511.9[医药卫生—内科学]