糖皮质激素治疗期免疫性血小板减少性紫癜儿童发生医院感染的危险因素分析及预测指标评价

Risk factors analysis and predictive index evaluation of nosocomial infection in glucocorticoid treatment in children with immune thrombocytopenic purpura

目的探讨免疫性血小板减少性紫癜(ITP)患儿糖皮质激素治疗期发生医院感染的危险因素,评价不同指标对患儿发生医院感染的预测能力。方法回顾性分析2014年1月-2018年6月本院362例ITP初诊患儿的临床资料。根据是否发生医院感染,分为感染组18例和未感染组344例,分析患儿发生医院感染的危险因素、病原学分布,并用ROC曲线预测不同指标对医院感染的预测能力。结果 362例ITP儿童发生医院感染18例,感染率为4.97%;粒细胞缺乏症、激素用量、累积用药时间是患儿发生医院感染的独立危险因素(P<0.05);联合激素用量和累计用药时间2项指标预测发生医院感染的ROC曲线下面积(AUC)为0.863,灵敏度为0.833,特异度为0.798,诊断效率优于单一指标。结论 ITP患儿糖皮质激素治疗期发生医院感染危险因素较多,联合以上指标有助于提高患儿发生医院感染的预测能力。

Objective To investigate the risk factors of nosocomial infection during the glucocorticoid therapy in patients with immune thrombocytopenic purpura(ITP), and to evaluate the predictive power of different indicators for nosocomial infection in glucocorticoid treatment. Methods Clinical data of 362 patients with ITP who received glucocorticoid therapy in our hospital from January 2013 to February 2018 was analyzed retrospectively. According to whether the patients was infected during hospitalization, 18 cases were infected and 344 cases were uninfected. The risk factors and pathogen distribution of nosocomial infection were analyzed, and the ROC curve was used to test the predictive ability of different indicators for nosocomial infection. Results In 362 cases of ITP children, 18 cases of nosocomial infection occurred, with the infection rate of 4.97%. Agranulocytosis, hormonal dose, and cumulative medication time were independent risk factors for nosocomial infection in children with glucocorticoid therapy during ITP(P<0.05). Two indicators of combined hormone dosage and cumulative medication time predicted ROC curve of nosocomial infection. The area under the area(AUC) was 0.863, the sensitivity was 0.833, and the specificity was 0.798. The diagnostic efficiency was better than the single index. Conclusion There are many risk factors for nosocomial infection in the treatment of glucocorticoid in children with ITP. To combine more than 2 indicators may help us to improve the predictive ability of nosocomial infection in ITP children.