Concomitant Use of Sulfonylureas and Warfarin and the Risk of Severe Hypoglycemia: Population-Based Cohort Study

Warfarin may interact pharmacokinetically with sulfonylureas, which could lead to increased sulfonylurea levels and toxicity (1). To date, two observational studies have assessed the potential association between concomitant use of sulfonylureas and warfarin and risk of severe hypoglycemia, with findings showing up to 47% increased risk compared with sulfonylurea use alone (2,3). However, both studies had methodological limitations, including information bias (3), selection bias (3), reverse causality (3), important confounding (3), and potentially inappropriate study design (2). Despite their limitations, the results of these studies have been incorporated into electronic drug-drug interaction tools, potentially affecting prescribing practices.

Given the limitations of the literature and the clinical implications of an increased risk of hypoglycemia due to the interaction of two commonly used drugs, further research is needed in the area. To this end, we conducted a retrospective population-based cohort study, using the U.K.’s Clinical Practice Research Datalink (CPRD), to assess whether concomitant use of sulfonylureas and warfarin is associated with increased risk of severe hypoglycemia compared with sulfonylurea use alone among patients with type 2 diabetes.

We assembled a study cohort including all patients aged ≥18 years who received a second-generation sulfonylurea (glibenclamide, glimepiride, gliclazide, glipizide) between April 1998 and June 2020. We excluded patients with <1 year of medical history or previous use of first-generation sulfonylureas, meglitinides, or insulin. The date of the first prescription for a second-generation sulfonylurea during the study period defined cohort entry. Patients were followed from cohort entry until the occurrence of the outcome (i.e., severe hypoglycemia defined as hospitalization with or death due to hypoglycemia [ICD-10 codes E16.0, E16.1, E16.2]), non-hypoglycemia-related death, administrative censoring, or end of study period.

We used a time-varying exposure definition, in which each person-day of follow-up was classified into one of the four mutually exclusive categories (all of which may or may not include treatment with metformin given its very low risk of hypoglycemia [4]): 1) current concomitant use of a second-generation sulfonylurea with warfarin but without other nonmetformin antidiabetes drugs (“sulfonylureas and warfarin”); 2) current use of a second-generation sulfonylurea without warfarin and without other nonmetformin antidiabetes drugs (“sulfonylureas alone”); 3) current use of a second-generation sulfonylurea with nonmetformin antidiabetes drugs, with or without warfarin; and 4) other combinations including no current use of a second-generation sulfonylurea. For all categories, person-time was defined as prescription duration plus a 30-day grace period. Concomitant use was defined as an overlap in the prescriptions of the drugs of interest.

We used time-dependent Cox models to estimate confounder-adjusted hazard ratios (HRs) and 95% CIs of severe hypoglycemia associated with current concomitant use of sulfonylureas and warfarin compared with current sulfonylurea use alone. To account for residual confounding, we conducted two supplementary analyses, each of which had a reference group of current concomitant use of sulfonylureas with an active comparator (current concomitant use of sulfonylureas and either antiplatelet agents or direct oral anticoagulants [DOACs], drug classes that share common indications with warfarin and do not have an inherent risk of hypoglycemia or interact with sulfonylureas). The study protocol was approved by the CPRD’s Independent Scientific Advisory Committee (protocol 20_195RA3) and the Research Ethics Board of the Jewish General Hospital, Montreal, Canada.

The study cohort included 325,549 patients initiating treatment with a second-generation sulfonylurea. Mean (SD) follow-up was 8.4 (5.7) years. During follow-up, there were 23,039 events of severe hypoglycemia in the entire cohort (crude incidence rate 8.42 per 1,000 person-years; 95% CI 8.31–8.53). Concomitant use of sulfonylureas and warfarin was associated with a 25% increased risk of severe hypoglycemia (adjusted HR 1.25; 95% CI 1.16–1.35) compared with sulfonylurea use alone (Table 1). When compared with concomitant use of sulfonylureas and antiplatelet agents, the increase in risk was slightly attenuated (HR 1.20; 95% CI 1.11–1.30). Moreover, when compared with concomitant use of sulfonylureas and DOACs, the increase in risk disappeared entirely (HR 1.01; 95% CI 0.67–1.52). There was some variation in the effect estimates based on age and protein binding degree of sulfonylureas but not based on sex. Sensitivity analyses addressing potential information bias, selection bias, and time-dependent confounding led to results consistent with those of the primary analysis.

Our large population-based cohort study showed a moderately increased risk of severe hypoglycemia associated with concomitant use of sulfonylureas and warfarin compared with sulfonylurea use alone. However, the results were not consistent in analyses addressing residual confounding based on the use of active comparators, where the increase in risk was attenuated or even disappeared entirely. Therefore, the increased risk is possibly a result of bias and does not seem to reflect a true effect.