Factors affecting diabetic screening behavior of korean adults: A multilevel analysis

Purpose: We investigated the role of individual and community level factors on diabetes screening test behavior. Methods: We used individual-level data from 170,193 adults aged 30 years or older who were not diagnosed with diabetes and participated in the 2009 community health survey. Community-level data includes 253 communities and were collected from various national statistics. Multilevel logistic regression analysis was conducted. Results: The rate of diabetes screening within the year prior to this study was 53.2%. Community variance of Model I, Model II and Model III was 0.236, 0.252 and 0.238, respectively. The proportional change in variance of Model II and Model III was e6.8% and e1.2%. The odds ratio for participation of diabetic screening of areas with bottom financial independence compared to areas with top was 0.84 (95% confidence interval, 0.74e0.96); the odds ratio of areas with top internist compared to areas with bottom was 1.15 (95% confidence interval, 1.01e1.31). Conclusion: This study identified a contextual effect influencing the participation of Korean adults in diabetes screening. It is necessary to develop specific policies that consider not only individual factors, but also community factors relating to individual behaviors to increase the likelihood of diabetes screening.