PICU mortality of children with cancer admitted to pediatric intensive care unit : a systematic review and meta-analysis

Background: Outcomes for children diagnosed with cancer have improved dramatically over the past 20 years. However, although 40% of pediatric cancer patients require at least one intensive care admission throughout their disease course, PICU outcomes and resource utilization by this population have not been rigorously studied in this specific group. Methods: Using a systematic strategy, we searched Medline, Embase, and CINAHL databases for articles describing PICU mortality of pediatric cancer patients admitted to PICU. Two investigators independently applied eligibility criteria, assessed data quality, and extracted data. We pooled PICU mortality estimates using random-effects models and examined mortality trends over time using meta-regression models. Results: Out of 1218 identified manuscripts, 31 studies were included covering 16,853 PICU admissions with the majority being retrospective in nature. Overall pooled weighted mortality was 27.8% (95% confidence interval (CI), 23.7-31.9%). Mortality decreased slightly over time when post-operative patients were excluded. The use of mechanical ventilation (odds ratio (OR): 18.49 [95% CI 13.79-24.78], p < 0.001), inotropic support (OR: 14.05 [95% CI 9.16-21.57], p < 0.001), or continuous renal replacement therapy (OR: 3.24 [95% CI 1.31-8.04], p = 0.01) was significantly associated with PICU mortality. Conclusions: PICU mortality rates of pediatric cancer patients are far higher when compared to current mortality rates of the general PICU population. PICU mortality has remained relatively unchanged over the past decades, a slight decrease was only seen when post-operative patients were excluded. This compared infavorably with the improved mortality seen in adults with cancer admitted to ICU, where research-led improvements have led to the paradigm of unlimited, aggressive ICU management without any limitations on resuscitations status, for a time-limited trial.