External validity of type 2 diabetes clinical trials on cardiovascular outcomes for a multimorbid population

• Published in: Diabetes, obesity & metabolism obesity & metabolism, 2021-04, Vol.23 (4), p.971-979
• Main Authors: Lazar Neto, Felippe, Mendes, Thiago Bosco, Matos, Paulo Marcelo Pontes Gomes, Oliveira, Julio César, Favarato, Maria Helena Sampaio, Lin, Chin An, Martins, Milton Arruda
• Format: Article
• Language: English
• Subjects: Aged; antidiabetic drug; Antidiabetics; canagliflozin; Cardiovascular disease; Cardiovascular diseases; Cardiovascular Diseases - epidemiology; Cardiovascular System; Clinical outcomes; Clinical trials; Comorbidity; Diabetes; Diabetes mellitus (non-insulin dependent); Diabetes Mellitus, Type 2 - complications; Diabetes Mellitus, Type 2 - drug therapy; Diabetes Mellitus, Type 2 - epidemiology; DPP‐4 inhibitor; GLP‐1; Humans; Hyperglycemia; Hypoglycemic Agents - therapeutic use; Insulin; insulin analogues; Medical diagnosis; Middle Aged
• ISSN: 1462-8902; 1463-1326
• DOI: 10.1111/dom.14303

Aim To investigate the external validity of recent antihyperglycaemic trials evaluating cardiovascular outcomes in a multimorbid population. Materials and Methods Selection criteria of 15 randomized controlled trials from the 2020 American Diabetes Association Standard of Care statement were applied in a stepwise manner to tertiary care patients with type 2 diabetes. Primary outcomes were the number of patients eligible per individual trial and for the aggregate of trials. Secondary outcomes included patient predictors of trial eligibility. Results Of 1059 patients, the mean (SD) age was 66 (10.74) years, the median (IQR) Charlson index was 2 (2, 3) and 458 (43%) had documented cardiovascular disease. The median (IQR) number of patients included in individual trials was 263 (174.25–308.75) and 795 (75.1%) of them were eligible for at least one trial. Among those 264 ineligible, 127 (48.1%) had an HbA1c level of 7% or less and no cardiovascular disease; 53.5% and 34.4% of the patients were eligible for two and three different classes of drugs, respectively. The strongest predictor of trial eligibility was cardiovascular disease (risk ratio 2.17, 95% CI 2.01–2.35). Conclusions A considerable proportion of multimorbid patients would be eligible for recent antihyperglycaemic trials. This positive finding can be attributed to development guidance in diabetes trials and the different approach we took, in which we evaluated inclusion by trials as an aggregate.