Systematic Review and Meta-Analysis of Prevalence and Population-Level Factors Contributing to Posttraumatic Stress Disorder in Pediatric Intensive Care Survivors
Rebecca E. Hay, Katie O’Hearn, David J. Zorko, Laurie A. Lee, Sarah Mooney, Cara McQuaid, Lisa Albrecht, David Henshall, Vanessa Campes Dannenberg, Veronica Flamenghi, Céline Thibault, Wai Kit Lee, Michelle Shi Min Ko, Michele Cree, Julia St. Louis, Julia A. Heneghan, Karen Ka Yan Leung, Andrea Wood, Eliana López, Mohamad‐Hani Temsah, Mohammed Almazyad, Jennifer Retallack, Mounika Reddy, Nedaa Aldairi, Rubén Eduardo Lasso Palomino, Karen Choong, Geneviève Du Pont‐Thibodeau, Laurence Ducharme-Crevier, Anne Tsampalieros, Lamia Hayawi, James Dayre McNally, Gonzalo Garcia Guerra
Abstract
OBJECTIVES: In survivors of illnesses or surgeries requiring PICU admission, there is a risk of posttraumatic stress disorder (PTSD). We aimed to estimate PTSD prevalence and potential contributing factors in survivors of PICU admission. DATA SOURCES: We performed a PROSPERO registered systematic review (CRD42022348997; Registered August 2022) using MEDLINE, Embase, CINAHL, and Cochrane Central Register of Controlled Trials (CENTRAL) databases, 2000 to 2022, with no language restrictions. STUDY SELECTION: Observational or interventional studies evaluating the incidence or prevalence of PTSD in patients' after PICU admission and/or contributing factors to PTSD. We used studies describing patients younger than 18 years old. Since there were a large number of citations, we used an integrated crowdsourcing and machine-learning model for citation screening. Each citation was reviewed independently and in duplicate by two reviewers at each stage of screening and abstraction. DATA EXTRACTION: Data items included study and participant demographics, details of case definition (PTSD screening), and risk factors. DATA SYNTHESIS: We followed the Preferred Reporting items for Systematic Reviews and Meta-analysis guidelines. Random-effects models were used to analyze PTSD prevalence and subgroup differences. In 24 citations meeting final review criteria, 19 had data for meta-analysis. There were 1898 PICU survivors with a median (interquartile range) cohort size of 59 (49-76). PTSD prevalence in the studies ranged from 3% to 37%; PTSD occurred in 529 of 1898 survivors ( I2 = 72%). Factors influencing PTSD variability included timing of assessment ( p < 0.01) with the highest prevalence (29%) at 6 months and the type of assessment instrument ( n = 10; range, 4-27%; p = 0.04). There was lower prevalence of PTSD (8%) in postoperative cardiac patients ( p < 0.01). Last, we failed to find an association between PICU length of stay and PTSD prevalence ( p = 0.62; I2 = 80%). CONCLUSIONS: PICU follow-up studies from 2000 to 2022 indicate that one-in-three of admissions surviving to 6 months have PTSD. However, there are population, study design factors and heterogeneity in PTSD assessment that indicate more standardization in this research is needed.