Effects of the type and quality of usual source of care on medical expenditures in adults with diabetes before and during the COVID‑19 pandemic: a panel data analysis using the Korea Health Panel (2019–2022)

Background: South Korea experiences the highest diabetes-related hospitalization rates among OECD countries. Although integrated primary care could potentially lower preventable hospitalizations and healthcare costs, the limited uptake of a usual source of care (USC) and underdeveloped primary care services in Korea have impeded progress, and the role and functions of primary care remain insufficiently defined. Methods: This study investigated how having a USC affects medical expenditures in adults with diabetes by using Korea Health Panel data from 2019 to 2022, spanning the pre-pandemic (2019) and pandemic (2020–2022) periods. The analysis included 6,144 individuals. The main independent variable was the type and quality of USC, categorized into three groups: no USC, place-only USC, and physician-based USC. Quality was assessed by patient-reported comprehensiveness and coordination, combined into an integrated quality index. We applied panel regression analysis using the Hausman–Taylor estimation technique to address both time-varying and time-invariant covariates while mitigating endogeneity concerns. Results: Between 2019 and 2022, the proportion of individuals with a physician at a regular location serving as their USC increased from 58.5% to 66.1%, while the proportion without a physician as their USC decreased from 15.1% to 10.9%, with larger variation among vulnerable populations. The physician-based USC group consistently incurred higher absolute expenditures but showed the smallest relative increase during the pandemic, compared with sharp rises in the no-USC and place-only USC groups. Adults with diabetes whose USC was a high-quality primary care physician—defined by favorable patient ratings of comprehensiveness and coordination—were associated with approximately 13.1% lower healthcare expenditures compared to those lacking a physician as their USC (β = − 0.140, P = 0.006). Moreover, community clinic–based physicians as USC were associated with lower expenditures, while hospital-based USC was associated with higher costs, though not statistically significant. Conclusion: These findings underscore the importance of increasing access to high-quality primary care physicians as USC to optimize chronic disease management and maintain sustainable healthcare systems.