Reviewer Critique: “An Economic Evaluation of the 2003 ESH-ESC Guidelines for the Management of Mild-to-Moderate Hypertension in Greece”

Stafilas et al compare the cost-effectiveness of five antihypertensive medications, one from each of the major classes, for the Greek health system.¹ The motivation for the analysis is that the 2003 European Society of Hypertension–European Society of Cardiology guidelines recommend all classes equally as initial therapy, whereas the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC-7) recommends diuretics.² The results presented by Stafilas et al support JNC-7: a diuretic, chlorthalidone, is the most cost-effective monotherapy for uncomplicated mild or moderate hypertension.

As a reviewer I focused on two questions: 1) How well does the analysis represent the clinical choices? and 2) Does the analysis, which is based on Greek costs, apply to the United States?

With respect to the first question, the important issues are effectiveness and compliance. The analysis assumes that all agents are equally effective, making the least costly automatically the most cost-effective. Compliance matters because other antihypertensive agents are thought to be better tolerated, so patients are more likely to continue taking them. The authors were careful to test alternative assumptions for effectiveness and compliance. They find that if newer agents are 30% more effective than chlorthalidone, enalapril is most cost-effective. If compliance with chlorthalidone is much lower than for other antihypertensive agents, (Stafilas et al,¹ Table 1) propranolol is most cost-effective. Breakeven analyses to identify exactly how much less effective chlorthalidone could be, or how much lower compliance could be, before it was no more cost-effective than each alternative, would also have been informative.

Is the analysis relevant to the US? That depends on prices in each country and on patterns of care, since part of a regimen's cost comes from its monitoring schedule. In answer to my query, the authors report that the European and JNC-7 guidelines recommend identical monitoring schedules. Unit costs clearly differ between the two countries. As the debate over importing drugs from Canada has shown, patients in other countries can acquire drugs, especially those under patent, at lower prices than in the US. The crucial point, however, is not price levels, but relative prices. If relative prices of drugs, visits, and tests are similar in both countries, the authors’ results hold for the US. If not, the cost-effectiveness ranking could change. I asked the authors to recalculate cost-effectiveness using US prices. They first used prices from a single state. I asked for something more representative; they then used Medicare reimbursement rates, the closest thing to national prices for the US. Again, they found that chlorthalidone was most cost-effective.

The authors calculate costs from the perspective of the Greek social security system. This “comprehensive payer” perspective, which includes all direct medical costs, is the perspective most often used by all authors, regardless of country. Stafilas et al initially explained that because they were not using the societal perspective, they had omitted indirect costs (such as wages lost to sickness and premature death). It is a common misunderstanding that indirect costs are part of the societal perspective. They are not.³ What is part of the societal perspective, which measures costs and health outcomes to all parties affected by an intervention, is the time of patients (and informal caregivers) required to produce it (for example, time spent visiting the doctor). These time costs are converted to dollars using a wage rate; for an example, see Cromwell et al.⁴