Objective: Aim of the study was to evaluate the protective effect of antidiabetic drugs in a large cohort of unselected elderly diabetic patients differing for their clinical status and life expectancy. The evaluation focused, in particular, on a group of patients with a very low survival rate because frail patients are usually excluded from intervention trials, limiting evidence on treatment effectiveness. The protective effect of drug therapy, quantified by the reduction in all-cause mortality with increased adherence to antidiabetic agent therapy, was studied by stratifying patients according to their life expectancy. Design and method: The Lombardy (Italy) residents, aged > 65 years, who received > 2 consecutive prescriptions of antidiabetic agents during 2012 were identified and the date of the third prescription was defined as the index date. A case-control study was nested into the cohort of antidiabetic drug users. Death from any cause was the outcome of interest, and cases were cohort members who died during follow-up (up to 2018). For each case, a control was selected and matched for age, gender, and clinical profile. Conditional logistic regression was used to model the risk of outcome associated with four categories of adherence to antidiabetic drugs. Adherence to drug therapy was measured by considering the proportion of days of the follow-up covered by the drugs. The analysis was stratified according to four categories of the clinical profile (good, intermediate, poor, and very poor) differing for life expectancies, as evaluated by a multisource comorbidity score able to accurately predict the risk of death. Results: Among the 276,336 patients in treatment with antidiabetic agents during 2012, 188,983 met the inclusion criteria and generated 49,219 deaths during follow-up, 49,201 of whom were matched to a control. The 6-year survival decreased from 85% to 52% from the group of patients with good to the group of patients with a very poor clinical status. Adherence to treatment was associated with a progressive decrease in the risk of mortality in all categories of clinical status. The reduction from lowest to highest adherence level was 36% (95% CI, 25–46%), 50% (44–56%), 38% (33–42%) and 26% (17–34%) from good to very poor clinical status (Figure). Conclusions: In a real-life context, adherence to antidiabetic drugs is associated with a reduction in the risk of mortality regardless of the clinical status of the patients. However, in more frail patients, the benefit of treatment is less than in patients in good clinical condition.
Read full abstract