Commentary

Based on major trials and the Joint National Committee recommendations,^1,2 diuretics should be the initial treatment for most hypertensive people. One shortcoming of some of the recent trials is a lack of direct comparisons between β blockers and either ACE inhibitors or diuretics.

Psaty et al attempt to add to this literature by using a methodologically complex method, the “network” meta-analysis. The advantage of this technique over a traditional meta-analysis is to combine “direct” comparisons with “indirect” comparisons of drugs (ie, when they are used in 2 different studies with a common comparison agent). This technique is usually frowned upon because of differences in populations and other sources of variability between studies, but this design is said to minimise those issues. Determining the validity of such a technique is difficult, but comparing the findings with other, more direct results would better support its conclusions, and the analytic method remains a second choice to well designed clinical trials.

Despite the fact that the authors provided several alternative analytic designs, the results were consistent with most other direct studies showing that diuretics were unsurpassed in decreasing cardiovascular risk outcomes compared with other treatments. In fact, in 6 of the 30 comparisons seen in the table, diuretics were superior to other treatments. This result led the authors to call for the use of diuretics as the “treatment of first choice” for patients with uncomplicated hypertension.

Unfortunately, most hypertensive patients require >1 drug for control, and because of a lack of consistency in many trials, we have little information about which combination of drugs is most effective. This is an important next step in determining the most appropriate algorithm for the management of hypertension. What is clear at this time is that most, if not all, patients with uncomplicated hypertension should be started on diuretics as initial treatment.