Baseline HbA1c strongly predicts metformin failure

Baseline glycated haemoglobin (HbA1c) is the most robust risk factor for metformin failure, reveals a study, noting that routinely available clinical data may be used to identify high-risk patients who may benefit from closer monitoring and earlier treatment intensification.

To determine the demographic and clinical factors available in the electronic health record (EHR) that predict metformin failure, the authors analysed patients with at least one abnormal diabetes screening test that initiated metformin at three sites in the US, namely Arizona, Mississippi, and Minnesota.

A total of 22,047 metformin initiators (mean age 57 years, 48 percent female) were identified. Of these, 2,141 were African Americans, 1,539 Hispanics, 440 Asians, 16,764 non-Hispanic White people, and 962 multiracial individuals.

Metformin failure was defined as either the lack of a target HbA1c (<7 percent) within 18 months of index or the start of dual therapy. The authors used tree-based extreme gradient boosting (XGBoost) models to evaluate the overall risk prediction performance and relative contribution of each factor when using HER data for risk of metformin failure.

In this population, the overall rate of metformin failure was 43 percent. The XGBoost model that included baseline HbA1c, age, sex, and race/ethnicity corresponded to high discrimination performance (C-index, 0.731, 95 percent confidence interval [CI], 0.722‒0.740) for risk of metformin failure.

Of the factors included, baseline HbA1c resembled the greatest feature performance with higher levels correlating with metformin failure. Notably, adding the other clinical factors improved the performance of the model (C-index, 0.745, 95 percent CI, 0.737‒0.754; p<0.0001).

“Metformin is the first-line drug for treating diabetes but has a high failure rate,” the authors said.