Identifying Ontarians with Type 2 Diabetes Mellitus in Administrative Data: A Comparison of Two Case Definitions

Introduction: This study compared two previously validated sensitive and specific diabetes case definitions to explore the impact of different classification methods in Ontario ICES administrative data. Methods: This study included patients captured by the Ontario Diabetes Database with type 2 diabetes using either the sensitive cohort definition (>= 2 physician visits for diabetes within 1 year or >= 1 drug claim for diabetes or >= 1 hospitalization with diabetes), or the specific cohort definition (>= 3 physician visits for diabetes within 1 year), between October 1, 2013 to September 30, 2015. Each cohort's demographic and clinical features were described using descriptive analysis. Results: Using sensitive and specific definitions, 1,093,812 and 783,228 patients with type 2 diabetes were identified, respectively. Overall, the demographic and clinical characteristics were similar between cohorts. Patients in the sensitive cohort had mean age of 64.1 years and were 52.4% male, compared to 64.8 years and 53.6% male in the specific cohort. In the sensitive and specific cohorts respectively, 64.4% and 55.7% of patients reported one-year mean HbA1c of < 7% (53 mmol/mol) and 25.3% and 31.5% reported levels between 7.0-8.5% (53-69 mmol/mol). Conclusions: Although sample sizes were different between sensitive and specific cohorts, demographic and clinical characteristics were similar.