Regional variations in frequency of glycosylated hemoglobin (HbA1c) monitoring in Korea: A multilevel analysis of nationwide data

Purpose Suboptimal frequency of glycosylated hemoglobin (HbA1c) monitoring is associated with poor diabetes control. We aimed to analyze compliance to HbA1c testing guidelines and explore associated individual and area-level determinants, focusing on regional variation. Methods This cross-sectional study between the period of 2012–2013 was conducted by using the Korean National Health Insurance Research Database, and included 45,634 patients diagnosed with diabetes mellitus, who were prescribed any anti-diabetic medications, including insulin. We calculated the proportion of each HbA1c testing frequency (≥1, ≥2, or ≥4 times per year) stratified by 17 administrative regions. Multilevel and multivariate logistic analyses were performed with regional (proportion of farmer population) and individual characteristics (age, sex, income level, duration of diabetes, and most visited medical institution). Results Overall, 67.3% of the patients received ≥ 1 HbA1c test per year; 37.8% and 6.1% received ≥2 and ≥4 tests per year, respectively. Those managed in secondary-level hospitals or clinics and those living in rural areas were less likely to receive HbA1c testing. Even after adjusting for individual and regional level characteristics, significant area level variation was observed (variance participant coefficients were 7.91%, 9.58%, and 14.43% for testing frequencies of ≥1, ≥2, and ≥4 times a year, respectively). Conclusions The frequency of HbA1c monitoring is suboptimal in Korea, especially in rural areas. Moreover, significant regional variation was observed, implying a contextual effect. This suggests the need for developing policy actions to improve HbA1c monitoring. In particular, access to HbA1c testing in rural primary care clinics must be improved.