Learning environment as a predictor for medical students’ empathy: a multicenter cross-sectional study from 12 medical schools in China

Empathy of medical students is crucial, yet it tends to decline as students enter later academic years. Empathy appeared to be affected by the learning environment (LE), which could be a potential contributor. We conducted a cross-sectional study to investigate the association between LE and empathy. The study comprised overall 10,901 medical students from 12 Chinese medical schools. Jefferson Scale of Empathy (JSE) was utilized to assess empathy, while a 5-point question and the Johns Hopkins Learning Environment Scale were employed to evaluate LE. JSE scores were compared across LE levels using Pearson Chi-square, Welch's ANOVA, and univariable linear regression. After adjusting for covariates, multivariable linear and logistic regression were used to assess the independent connection between LE and empathy. Confounders were controlled by stratified subgroup analysis. LE was significantly associated with medical students' empathy, as shown by the Pearson Chi-square test (p < 0.001) and Welch's ANOVA (p < 0.001). In multivariable logistic regression analysis, lower JSE scores were linked to common, bad, and terrible LE (OR = 2.226, 95% CI = [2.011, 2.465], p < 0.001; OR = 2.558, 95% CI = [1.745, 3.751], p < 0.001; OR = 1.889, 95% CI = [1.131, 3.153], p = 0.015), while excellent LE was associated with higher JSE scores (OR = 0.348, 95% CI = [0.312, 0.388], p < 0.001). Linear regression and stratified subgroup analysis confirmed logistic regression's findings. LE was a significant predictor for empathy. Better LE was associated with higher empathy and worse LE was related to lower empathy. This inspired us to improve LE to enhance medical students' empathy.