Effectiveness of three‐dimensional printed cardiac models in teaching congenital heart anatomy: A systematic review and meta‐analysis of randomized controlled trials

Abstract Congenital heart disease challenges spatial understanding. 3D printed cardiac models are increasingly used, but randomized quantitative evidence remains limited. This review and meta‐analysis evaluated their effectiveness for teaching congenital heart anatomy. PubMed, Embase, Scopus, and Web of Science were searched through November 12, 2025. RCTs comparing 3D‐printed cardiac models with conventional instruction were included. The primary outcome was objective knowledge performance. SMDs were pooled using random‐effects REML models. Risk of bias was assessed with RoB 2 and the certainty of evidence evaluated with GRADE. Six RCTs met the inclusion criteria; five RCTs ( n = 228 participants) contributed data to the primary meta‐analysis. The use of 3D‐printed cardiac models significantly improved objective post‐intervention knowledge performance compared with conventional instructional methods (SMD = 0.62; 95% CI, 0.21–1.02; p = 0.003; I 2 = 53.2%), corresponding to a moderate effect size according to Cohen's conventional interpretation. Because of the small number of eligible trials, subgroup analyses were not interpreted as confirmatory, and subjective educational outcomes were summarized cautiously rather than used to support the primary conclusion. The certainty of evidence for objective knowledge performance was rated as moderate using the GRADE approach, mainly due to risk‐of‐bias concerns and imprecision. The use of 3D‐printed cardiac models was associated with moderate improvements in objective learning of congenital heart anatomy compared with conventional teaching approaches. However, the small evidence base, methodological limitations, and short‐term nature of most assessments support a cautious interpretation. These models may be considered adjunctive educational tools, but larger randomized trials using standardized outcomes and longitudinal follow‐up are needed. image