Eleven RCTs were included (n=1,832). Trial size ranged from 40 to 720 patients. Four trials had a Jadad score of 4, three scored 3, one scored 2 and three scored 1. Allocation concealment and intention to treat analysis were the main shortcomings in most studies.
The funnel plot for clinic diastolic blood pressure reduction indicated publication bias but the funnel plot for the therapeutic response of diastolic blood pressure reduction did not.
Telmisartan was associated with a significant reduction in clinic diastolic blood pressure compared with losartan (10 RCTs) WMD 2.62 (95% CI: 0.96, 4.27, p=0.002), but significant heterogeneity was detected I2 = 80.8% (which became non-significant with the exclusion of one study). Similar reductions were found for systolic blood pressure with telmisartan compared with losartan (10 RCTs) WMD 2.77 (95% CI: 1.90, 3.63, p<0.00001); significant heterogeneity was not indicated. Ambulatory diastolic blood pressure and systolic blood pressure were also significantly reduced for the previous six hours, 24 hours, daytime and night time in patients treated with telmisartan compared with losartan. There was significant heterogeneity in the analysis for diastolic blood pressure for 24 hours, daytime and night time, which again became non-significant when a study was excluded.
There was a significant increase in the therapeutic response with telmisartan compared with losartan in diastolic blood pressure (seven trials) RR 1.14 (95% CI: 1.04, 1.23, p=0.003) and systolic blood pressure (four trials) RR 1.10 (95% CI: 1.01, 1.20, p=0.03). No significant heterogeneity was detected.
No significant differences in withdrawals and adverse events were found with telmisartan compared with losartan.
Results of sensitivity analyses and subgroup analyses were reported in the paper.