Randomised controlled trials (RCTs) of at least 24 weeks duration that compared rosiglitazone (2mg/day to 8mg/day) alone or in combination with other anti-diabetic drugs were eligible for inclusion. Studies with any other anti-diabetic medication in adults with haemoglobin A1c (HbA1c) over 7% or fasting blood plasma glucose (FPG) greater than 126mg/dL were eligible. Primary or secondary outcomes in the studies had to be cardiovascular and non-cardiovascular adverse effects or events.
The primary objective of the review was to evaluate effects of rosiglitazone on cardiovascular outcomes in patients with type 2 diabetes mellitus. The secondary objective was to assess effects of rosiglitazone and pioglitazone on lipid profiles. Only studies published between 2000 and 2007 were included for the primary objective. For the secondary objective, RCTs published between 1998 and 2007 that compared effects of either rosiglitazone with pioglitazone or thiazolidinedione monotherapy with placebo on lipid profiles were eligible.
For the primary objective, most studies compared rosiglitazone 8mg/day with another monotherapy (metformin, sulfonylurea or placebo); other studies assessed 4mg/day or 2mg/day. Half of the studies evaluated rosiglitazone in combination with other anti-diabetic agents. Study duration ranged from 24 weeks to four years. Baseline body mass index (BMI) ranged from 28kg/m2 to 34.6kg/m2. Patient age ranged from 52 to 64.3 years. HbA1c ranged from 7.35% to 9.5%. FPG ranged from 151mg/dL to 268mg/dL. Duration of diabetes was from 4.6 to 12.7 years. For the secondary objective, studies compared rosiglitazone (2mg/day to 8mg/day) with pioglitazone (15mg/day to 45mg/day) or placebo and pioglitazone with placebo. Study durations ranged from eight weeks to one year. Baseline body mass index ranged from 24.3kg/m2 to 35.6kg/m2. Patient age was from 53 to 63 years.
The authors did not report how many reviewers performed the study selection.