Objective  To understand the impact of chronic disease comorbidity on healthcare institution-seeking behavior of elderly people in China, and to provide a theoretical basis for optimizing healthcare order and enhancing the management capacity of primary care for chronic diseases.

Methods The data on 4 508 elderly people aged ≥ 60 years in 25 provincial-level administrative divisions across China were collected from the China Family Panel Studies (CFPS) conducted in 2020. The Stata 17.0 statistical software was used to analyze the impact of chronic disease comorbidity on healthcare facility-seeking behavior of elderly people in China using the probit regression model and the propensity score matching (PSM) method.

Results Of the 4 508 elderly people surveyed, 748 (16.59%) reported having chronic disease comorbidity (suffering from two or more chronic diseases); 2 536 (56.26%) and 1 972 (43.74%) reported seeking medical treatment at primary and non-primary healthcare institutions during past six months. After adjusting for predisposing factors (gender, age, education level, marital status), enabling factors (household registration type, monthly retirement pension, satisfaction with medical treatment, participation in medical insurance, emotional relationship with children, financial support from children, caregiving provided by children), and need factors (self-rated health status, changes in health conditions, physical discomfort in the past two weeks, the ability to take care of oneself), the results of probit regression analysis showed that the elderly people with comorbid chronic diseases were more inclined to seek medical treatment at non-primary healthcare institutions (β = – 0.130, 95% confidence interval: – 0.237 – – 0.022). The estimated values of average treatment effect on the treated (ATT) of chronic disease comorbidity on the healthcare institution selection of the elderly people were – 0.052, – 0.048, and – 0.047, respectively, based on matched propensity scores of predisposing, enabling, and need factors with K-nearest neighbor, kernel, and radius matching (all P < 0.05), indicating that chronic disease comorbidity can reduce the probability of elderly people seeking medical treatment at primary healthcare institutions. The heterogeneity test results showed that the impact of chronic disease comorbidity on females (ATT = – 0.078 – – 0.063, all P < 0.05) and the elderly people with a junior high school education or below (ATT = – 0.058 – – 0.052, all P < 0.05) was more significant, but there was no statistically significant difference in the impact on males and elderly people with a high school education or above (all P > 0.05).

Conclusion Chronic disease comorbidity may affect the selection of medical institutions for elderly people in China to some extent, and its influence differs by gender and education level of the elderly.