Child Life Services Linked to Lower Pediatric Procedural Pain
Key Takeaways
- Child life services were associated with lower procedural pain than standard care across the pooled evidence in children.
- Heterogeneity was substantial, and assessor type was the only significant prespecified moderator, with larger effects in child self-report than in observer-rated outcomes.
Investigators conducted a registered systematic review and three-level meta-analysis under PRISMA 2020, searching six databases from inception to 2 June 2025. Eligible studies included randomized controlled trials and quasi-experimental comparisons of child life services against standard care. The synthesis included 12 studies, 1,366 participants, and 24 effect sizes, with seven randomized trials and five quasi-experimental studies published from 2015 to 2025. Most studies were conducted in China, with smaller representation from the United States and Australia. Child life services were described as non-pharmacological support delivered by trained professionals using therapeutic play, medical preparation, distraction, psychological support, and family support.
Heterogeneity was substantial and was driven mainly by differences between studies, with Level 3 I2 = 85.0%. Within-study heterogeneity was lower, with Level 2 I2 = 8.1%, while the overall association still favored child life services. Assessor type was the only significant prespecified moderator, with F(1,22) = 8.68 and p = 0.008. Self-reported outcomes showed a larger pooled effect, at k = 9 and g = -1.06, p < 0.001, whereas observer-rated outcomes were k = 15 and g = -0.32, p = 0.261.
Age group, procedure type, and measurement timing were not significant moderators overall. Within-level analyses nevertheless showed significant associations in children 7 years and younger, one-time injections, and post-procedure assessments. Corresponding estimates in older children, surgical procedures, and during-procedure assessments did not reach significance, which limited firm contrasts between these subgroups. Sensitivity analyses remained significant, including leave-one-out testing and removal of one identified outlier, and still favored child life services. Publication-bias checks did not show statistical evidence of publication bias, with non-significant Egger tests and trim-and-fill analyses that imputed studies without changing the overall direction.
9 of the 12 studies were conducted in China, which narrowed broad geographic generalizability. No included studies systematically monitored adverse effects, leaving safety reporting sparse across the available evidence. Reporting on intervention duration, number of sessions, provider background, and years of experience was inconsistent, limiting how precisely the services could be characterized. Risk-of-bias assessments also raised concerns, with four randomized trials at high risk and all five quasi-experimental studies at serious risk, partly because independent assessors were not used consistently.