BACKGROUND
The continuity of care in chronic diseases, especially in diabetes, was emphasized from many studies. But large scale studies with long-term observation which confirm the impact of continuity of care on health outcomes are rare. This study tried national level 3 year observation to find differences in hospitalization, mortality and medical costs among patient groups with different utilization pattern. METHODS: The 1,088,564 patients with diabetes diagnosis and diabetes drug prescription in 2002, from 20 to 79 years old, and survived until the end of 2004 were included. Annual drug prescription days, number of visited clinics and quarterly continuity of care were measured. Gender, age group, living area, health insurance premium level (as a proxy of the income level), years of first DM diagnosis, five co-morbidities (hypertension, heart disease, stroke, renal disease, admission with DM), hospitalization experience and the type of main attending clinic were adjusted. Hospitalization, mortality and high costs group (top quintile) in 2005 were predicted by multiple logistic regression model. RESULTS: Patients who failed in continuity of care in 2003 and 2004 showed higher hospitalization (OR =1.29), higher mortality (OR =1.75) and they are more likely to be high costs group (OR =1.34) in 2005 than who fulfilled the continuity of care. Patients who have single attending clinic also showed lower hospitalization, lower mortality and lower cost. Completeness in diabetic drug prescription were correlated with lower hospitalization, lower mortality but with higher cost. Possible cost saving from continual care with single attending clinic was estimated at Won 417 billion (Dollar 1 = Won 943.7). Possible expenditure from complete drug prescription was Won 228 billion. So, net saving was Won 139 billion in our study population. CONCLUSION: Continual care and single attending clinic saves patient's life and national costs. Fragmented primary care system in Korea should be reformed for more effective care of chronic diseases. National Health Insurance Database in Korea enables nationwide long-term observation study which overcomes the many limitations found in hospital-based studies and cross-sectional surveys.