Retrospective or prospective studies of patients who initiated exenatide in clinical practice were eligible for inclusion. Studies had to include 100 patients or more per treatment group and report relevant outcomes. Studies such as randomised trials, reviews, case studies and articles that did not include at least one end point of interest were excluded.
The outcomes of interest were the effects of exenatide on haemoglobin A1C (HbA1C), fasting glucose, weight, systolic blood pressure, medication use, hospitalisation and cardiovascular disease outcomes.
Most studies were conducted in the United States. The authors stated that, compared with clinical trial populations, these exenatide-treated patients typically had a higher level of disease severity and comorbidities (such as, hypertension, hyperlipidaemia, obesity, history of cardiovascular disease). In some studies, patients were treated with exenatide in combination with the oral glucose lowering therapies metformin, sulphonylurea or thiazolidinedione with or without insulin. Comparators were other glucose lowering therapies, insulin, insulin glargine, NPH insulin (Neutral Protamine Hagedorn insulin) and sitagliptin. Some studies did not have comparator group.
The authors did not state how many reviewers were involved in study selection.