This was a good cost-effectiveness analysis, however the following points should be noted regarding its validity.
1. The study did not include patients with poorly controlled BP, whose management would alter cost-effectiveness measures (incidentally, this fact limits the generalisability of this analysis to the whole population);
2. Analysis was based on treatment completers only, which may bias effectiveness results due to non-random dropouts;
3. A sensitivity analysis on cost measures would have been useful, to allow for uncertainty in measurement of, for example, administration costs or errors;
4. Indirect costs amount to a substantial proportion of total costs, but they were measured according to a source published in 1987 (no information on whether costs were updated to allow for price changes is provided);
5. No incremental cost-effectiveness ratios, which would have given a good insight into relative cost-effectiveness, were calculated.