Portal de Periódicos da CAPES

Easy money?: Health cost savings resulting from the switch from a branded drug to a low-cost generic drug in the same class

2011; Wiley; Volume: 65; Issue: 3 Linguagem: Inglês

10.1111/j.1742-1241.2011.02642.x

1742-1241

Bertram Pitt, Stevo Julius,

Health Systems, Economic Evaluations, Quality of Life

The review and meta-analysis by Grosso et al. (1) comparing the effectiveness of the angiotensin receptor blocker candesartan with losartan in patients with hypertension and in those with systolic heart failure suggesting that one should substitute generic losartan for candesartan has, as pointed out by the authors, the potential to reduce healthcare costs. In view of the economic pressure to reduce healthcare costs in most parts of the world, it is important that we consider the most cost-effective therapy for our patients. As clinicians, we have our primary obligation to our patients, but we also have an obligation to society and our healthcare providers to be certain that we provide not only the best but also the most cost-effective therapy. The implications of this analysis are therefore important. Many clinicians and patients depending upon their healthcare plans are familiar with this dilemma and have received notices from their healthcare provider suggesting that they substitute less expensive generic drugs for more expensive branded drugs of the same class. The recommendation to substitute generic losartan for candesartan is therefore not surprising and probably has already occurred in many instances. However, before we recommend routine substitution of losartan for candesartan, it might be prudent to evaluate critically the data provided by Grosso et al. (1) and the implications of their recommendation. Their analysis clearly shows that candesartan within its available dose range is more effective than losartan within its available dose range to reduce systolic blood pressure. While it is likely that the two angiotensin receptor blocking agents candesartan and losartan have a similar, although not identical, mechanism of action and safety profile, the approximately 3 mmHg difference in systolic and 2 mmHg in diastolic blood pressure lowering favouring candesartan requires further discussion. While a 2–3 mmHg difference in systolic and diastolic blood pressure may at first glance not appear to be meaningful, numerous epidemiological and prospective randomised clinical trials suggest that there may be important long-term cardiovascular consequences resulting from a 2–3 mmHg reduction in systolic and/or diastolic blood pressure. For example, in a meta-analysis of BP lowering in nine randomised trials in 62,605 patients with high-risk hypertension, Staessen et al. (2) found that 2–3 mmHg lower BP is associated with a large reduction of strokes. In a follow-up report (3), these authors stated that 1–3 mmHg decrease in systolic BP is associated with a 20–30% reduction in the relative risk of strokes. It is fair to say that most clinicians accept that even a small decrease in BP reduces stroke. The importance of this reduction becomes particularly significant when one realises the worldwide frequency of strokes. Thus, the laudable and important Science advisory Committee on Nutrition sodium reduction initiative in the UK expects to induce a relatively small reduction in blood pressure, but rightly expects that this may substantially decrease the incidence of strokes. In view of these facts, it is hard to understand why the model used by Grosso et al. (1) did not predict an excess of strokes in the losartan group. The problem may well be related to the 10-year timeframe of the Markov State Transition model used to evaluate cost effectiveness. In high-risk hypertension, differences in BP have an almost imminent effect on strokes and other cardiovascular events. In the VALUE trial (4), a five-step medication up-titration scheme was applied in the first 6 months of the treatment. In chosen doses, amlodipine lowered the BP faster than valsartan. In the first 3 months of the trial, this resulted in a 3.8 mmHg systolic BP difference and a 94% higher incidence of strokes. In the next 3 months, the BP difference was 2.3 mmHg and the stroke incidence was 50% higher. After 6 months when the target BP was achieved, the subsequent incidence of strokes was similar in the two treatment groups. Stroke is the most debilitating consequence of hypertension and the pain that even one extra stroke causes to a patients and the family cannot be measured in terms of cost effectiveness. However, the financial impact of stroke on the healthcare delivery system can be assessed and is very likely substantial. Grosso et al. (1) should evaluate why their Markov model did not reflect this cost. Furthermore, their model does not appear to take into account the potential for developing heart failure in a patient with hypertension, an important factor especially in the elderly. They do, however, point out that if a 2–3 mmHg reduction in blood pressure is considered important, it should be relatively easy to achieve further blood pressure lowering in a patient switched from candesartan to losartan by adding another antihypertensive agent. However, in clinical practice where blood pressure monitoring may not be rigorous, the 2–3 mmHg loss in blood pressure control resulting from substitution of losartan for candesartan may not be recognised and, if recognised, may not be thought important. If, however, the 2–3 mmHg in blood pressure difference is detected and thought important, one would need to add an additional antihypertensive agent to achieve the blood pressure control associated with the prior use of candesartan. Each additional antihypertensive agent while providing further blood pressure control carries with it the potential for further side effects such as hypokalemia, peripheral oedema, increased insulin resistance, fatigue, bradycardia, constipation, etc. Long-term adherence to antihypertensive therapy in asymptomatic patients is difficult at best. The occurrence of unwanted side effects in such an asymptomatic individual could result in the patient discontinuing not only the additional antihypertensive medication added to compensate for the 2–3 mmHg blood pressure difference but all of their antihypertensive drugs with disastrous consequences for both long-term cardiovascular outcomes and healthcare costs. The situation with regard to substituting losartan for candesartan in patients with systolic heart failure is even more problematic. Candesartan using a dosing strategy of up to 32 mg/day has been shown to be effective in reducing the combined end-point of cardiovascular mortality and hospitalisations for heart failure in the Candesartan in Heart Failure assessment of reduction in Mortality and Morbidity (CHARM) trials (5, 6). Losartan 50 mg/day, however, has not been shown to be equivalent to the dose of the angiotensin converting enzyme inhibitor captopril that has been shown to be effective in patients with systolic heart failure, 50 mg tid (7). In the Heaal trial (8), a dose of 150 mg/day of losartan was shown to be more effective than 50 mg/day. However, this dose is not licensed in the UK or approved in the USA and relatively infrequently used in clinical practice. Moreover, it remains uncertain whether 150 mg/day of losartan is the optimal dose for patients with systolic heart failure. Losartan, in contrast to candesartan, has a relatively short half-life. There is information suggesting that the optimal dose of losartan for reducing blood pressure is 100 mg bid (9). The optimal dose in patients with systolic heart failure remains to be determined. Therefore, switching patients with systolic heart failure from candesartan 32 mg/day to 100 mg/day of losartan (the currently approved dose in the USA and the UK) or even 150 mg/day may result in a reduction in target organ protection and an increased incidence of death and or hospitalisations for heart failure. Even if the difference between the effectiveness of losartan and candesartan results in only a few percentage difference in hospitalisations for heart failure, given the cost of a single hospitalisation for heart failure, this could negate all of the potential savings postulated from switching from candesartan to losartan and in fact result in a substantial increase in healthcare costs that could be far greater than the postulated saving proposed by Grosso et al. (1). Thus, while we would welcome the potential cost savings resulting from switching from candesartan to generic losartan, we would not advocate such a strategy without further data from direct comparative well-powered cardiovascular outcome trials assuring us that the two strategies are equivalent or possibly that losartan is non-inferior to candesartan within a boundary for cardiovascular outcomes that does not negate the potential savings resulting from the switch. Although our focus in this article is the proposed switch from candesartan to losartan in patients with hypertension and or heart failure, the caution expressed above and the need for convincing long-term outcomes data apply to several other proposed switches from so called more expensive to less expensive drugs. It would indeed be tragic in this era of limited economic resources if we traded a short-term potential saving in healthcare costs for a real increase in long-term costs and cardiovascular risk. As in many aspects of life, one should be wary of the temptation for ‘Easy Money’. Bertram Pitt, is a consultant for, has received grants from, or held stock options for the following: Pfizer, Merck, Novartis, Takeda, Bayer, AstraZeneca, Boehringer Ingelheim, Bristol-Myers Squibb, Forrest Laboratories, GE-Healthcare, Relypsa, BG-Medicine, Nile Therapeutics, Aurasense, Medtronic. Stevo Julius has no disclosures.