Chinese Journal of Evidence -Based Pediatric ›› 2019, Vol. 14 ›› Issue (6): 406-412.

• Original Papers • Previous Articles     Next Articles

Risk factors for 28-day mortality in immunocompromised and immunocompetent pediatric patients admitted to PICU for sepsis: A case-control study

PENG Chun-ying, LIU Ting-yan, YING Jia-yun, YAN Gang-feng, LU Guo-ping   

  1. Children's Hospital of Fudan University, Shanghai 201102, China
  • Received:2019-09-30 Revised:2019-12-05 Online:2019-12-25 Published:2019-12-25
  • Contact: LU Guo-ping

Abstract: ObjectiveTo analyze the risk factors for 28-day mortality in immunocompromised pediatric patients admitted to PICU for sepsis. MethodsWe conducted a case-control study, and retrospectively collected the clinical data of patients who were admitted to the PICU of Children's Hospital of Fudan University for sepsis. Patients were divided into immunocompetent group and immunocompromised group based on immune status and 28-day mortality was the primary outcome. ResultsFrom December 1st, 2015 to December 31st, 2018, 385 consecutive cases were diagnosed as sepsis at discharge, of which 251 fitted the eligibility criteria, and others were excluded due to ICU-acquired sepsis or being discharged within 24 h. We identified the immunocompetent group (n=141) and immunocompromised group (n=110). The 28-day mortality was 69.1% in immunocompromised patients. Compared with the immunocompetent patients, they showed a larger portion of in-patients, less age disparity and higher PRISM Ⅲ score, and they were more dependent on life support therapies (vasoactive agents, ventilation) and less likely to localize infection sites, the RRT treatment survive rate was 17.4% and none of the 5 ECMO receivers survived. Univariate analysis of 28-day mortality within two groups identified several common factors including septic shock, invasive-ventilation, CRRT, PRISM Ⅲ score, PICU length of stay, hospital stay,PICU-free time, therapy limitation within 24 h and overall therapy limitation, with "other comorbidities" unique to the immunocompetent patients and "vasoactive agents" unique to immunocompromised patients. Multivariate Cox regression revealed that PRISM Ⅲ score, invasive-ventilation and serum lactate above 2 mmol·L-1 were shared risk factors in both groups, and septic shock was also a predictor in the immunocompromised group. Conclusion28-day mortality in pediatric patients admitted to PICU for sepsis remained high, with the immunocompromised status more likely to succumb to death. PRISM Ⅲ score, invasive ventilation and serum lactate above 2 mmol·L-1 were strong predictors for short-term mortality, hence early recognition and effective management might improve patients clinical outcome.