# A systematic review identifying effective teaching methods and their combinations for increasing empathy in physicians: pairwise and network meta-analysis

## TL;DR

This study identifies teaching methods that effectively increase empathy in physicians and medical students through a systematic review and meta-analysis.

## Contribution

The study provides novel insights into the most effective teaching method combinations for improving empathy in medical training.

## Key findings

- A medium effect of interventions on empathy skills was found compared to control groups.
- Didactic and rehearsal methods were most frequently included in effective teaching combinations.
- Transparent and standardized reporting could reduce heterogeneity in future studies.

## Abstract

Demonstrating empathy is fundamental for providing patient-centered care, however, what components are most effective is not known. Thus, the aim of this study was to identify key teaching methods and intervention characteristics for increasing empathy skills across medical training.

This study was part of a larger systematic review, which was pre-registered: CRD42018100100. We performed a systematic review, pairwise meta-analysis (PMA) and network meta-analysis (NMA). A systematic search was used across PsycINFO, Medline, CINAHL, Social Work Abstracts, ERIC, ABI/INFORM, and the Cochrane Central Register of Controlled Trials from the inception of the respective databases to October 9, 2022. Studies included randomized controlled trials (RCTs) examining behavioural interventions which targeted empathy skills for physicians and medical students. Studies were excluded if reported summary data could not be converted to an effect size; if the author were unable to be contacted; and if the study did not compare substantively different intervention combinations. Risk of bias was assessed using the Cochrane risk-of-bias tool. Data were pooled using random-effects PMA and NMA.

308 full-text studies were found, of which 111 met the inclusion criteria, totalling 11,111 participants. Overall, a medium effect of interventions was found [d = 0.50 (95% CI = 0.40, 0.60)], meaning the empathy skills of participants improved moderately compared to those in control groups. Publication bias was evident and heterogeneity was high (I2 = 79.19, p < .001). Subgroup analyses of the PMA revealed the following moderators were statistically significant: teaching method, intervention formats, control group; measurement type, and number of teaching methods used. The consistency assumption was met [χ2 [ (33)= 39.11, p = .21] for the NMA. The NMA revealed that didactic and rehearsal were most frequently included among the most effective teaching method combinations.

By using PMA and NMA, we provide novel insights on effective intervention components for improving empathy in medicine. To improve aggregation of evidence, transparent and standardized reporting from studies may help reduce heterogeneity. Overall, our results support the notion that interventions need not be expensive nor prolonged to be effective.

The online version contains supplementary material available at 10.1186/s12909-025-07917-x.