In a study of reference-based pricing, Chantal Bourgault and colleagues1 challenge the therapeutic equivalence of 3 ACE inhibitors on the grounds that patients with hypertension whose treatment was started with captopril had more subsequent hospital admissions and ambulatory medical visits than those who were initially prescribed enalapril or lisinopril.
First, it is unclear why they used an intention-to-treat analysis. What they needed to determine was whether the medication used was the cause of these outcomes. In a recent Canadian study2 by 3 months 13% of patients with hypertension who were initially prescribed an ACE inhibitor did not persist, and by 4.5 years only 53% were still taking that medication.
Second, before wading into a database analysis to look for correlations, there should surely have been a more precise statement of a clinically plausible hypothesis than, I wonder if all 3 drugs are associated with the same admission rate? These were patients who were starting treatment for uncomplicated hypertension, a condition that is unlikely of itself to be the cause of hospital admissions.
Was the hypothesis then that captopril was less effective in controlling blood pressure, and that this caused a higher incidence of heart disease and stroke in the subsequent 4 years? There is ample evidence that blood pressure control with these 3 agents is fairly comparable, and even if blood pressure was less well controlled, the resultant increase in heart disease and stroke would be unlikely to appear so rapidly.
Or was the hypothesis that with captopril there would be more side effects or drug interactions of sufficient severity to warrant hospital admission? These hospital discharge diagnoses, which were ascertained but not reported, would surely tell us the reason for hospital admission and would suggest the way in which captopril was (or was not) the cause of increased admission rates.
Numerous possible reasons for this result, other than the absence of therapeutic equivalence, are cited in this paper and in an accompanying editorial.3 The most likely explanation is that the patients initially prescribed captopril were more sick before therapy started than the patients prescribed the other drugs, as evidenced by higher drug use and higher admission rates. Surely one can never say that one has corrected for this, but only that one has tried to correct for this.
I think we should conclude that this is a provocative study that merits clarification, but the conclusion that the result "suggests that ACE inhibitors may not be therapeutically equivalent" is premature.