OBJECTIVES: To estimate the number and to determine the clinical and demographic characteristics of non-insulin treated patients with type 2 diabetes (T2DM) who fulfil conditions for insulin treatment (insulin-gap) in a Portuguese primary care database.

METHODS: Population-based study with real world data covering 3.6 million patients attending primary care in a large health region in Portugal. All adult users coded for T2DM or prescribed antidiabetic medication with at least one GP visit in 2014 were initially selected. Exclusion criteria were applied to select eligible patients. Non-insulin treated patients who fulfilled at least one of three predefined criteria were identified as being in insulin-gap. These criteria were based on glycated hemoglobin levels and on the estimated number of antidiabetic medications. We analyzed patients’ characteristics, cardiometabolic status, comorbidities, and resource use related to medical visits and antidiabetic medication.

RESULTS: Of the initial 250,391 patients identified, only 136,977 were eligible for analysis. Around 11% were insulin treated (n=14,408). We estimate that 8,475 patients (6.9% of all non-insulin treated patients analyzed) fulfilled conditions for insulin treatment. These patients are mostly women (51%) and on average 65.8 ±11.0 years old. They show higher mean HbA1c (9.5 ±1.2%), LDL cholesterol (106 ±37 mg/dL) and SBP (140 ±18 mmHg) compared to either insulin-treated (8.0 ±1.5%, 100 ±36 mg/dL and 138 ±18 mmHg, respectively) or other oral antidiabetic medication only treated patients (6.6 ±0.8%, 103 ±34 mg/dL and 137 ±17 mmHg, respectively). The most frequent comorbidities among patients in insulin-gap are hypertension and dyslipidemia (69.4% and 45.5%, respectively). Around 40% of patients are overweight or obese.

CONCLUSIONS: The proportion of patients with insulin-gap was estimated at approximately 7% of non-insulin treated patients in a large population in a primary care setting in Portugal. As expected these patients show a worse cardiometabolic status compared to insulin-treated patients.