Comments on systematic review of clinical- and cost-effectiveness of candesartan and losartan in hypertension and heart failure

Linked Comment: Grosso et al. Int J Clin Pract 2011; 65: 912. To the Editor: We read with interest the article by Grosso et al. (1) and the associated commentaries (2, 3). We are reassured that the author’s meta-analysis is entirely consistent with a more wide ranging and comparable analysis published over a year ago (4). However, certain aspects of the analyses and the interpretation give us cause for concern. The potential clinical importance of a 1.96 mmHg difference in diastolic blood pressure (DBP) is based upon cost-utility analysis utilising a 10-year Markov model. This ignores epidemiological evidence and findings from outcome trials (4) and fails to acknowledge the 2006 Health Survey for England which suggests that a reduction in 2 mmHg of DBP would save 14,000 lives per annum in the UK (5). Although there are no outcome trials comparing different ARBs, there is no acknowledgement of a study that utilised data from electronic primary care journals and mandatory Swedish national registers for death and hospitalisation, which suggested that, in the treatment of hypertension, there is a BP independent risk reduction in cardiovascular disease with candesartan compared with losartan (6). Further analyses indicated that prescribing candesartan resulted in lower long-term health-care costs and comparable costs when proprietary candesartan was compared with generic losartan (7). It is also interesting to note from a recent publication based upon heart failure registry data that the use of candesartan compared with losartan was associated with a lower mortality risk (8). Any cost-utility analysis is dependent upon the assumptions made in formulating the model. The conclusions drawn from the analysis of Grosso et al. (1) are at odds with an earlier study, which suggested that cost savings could be achieved by switching patients from losartan to candesartan (9). The authors of this article concede that their conclusions were dependent upon the costs incorporated in their analysis (9). The primary analysis of Grosso et al. (1) is based upon a model with generic losartan retailing at £6.47 (100 mg; 28-day pack price) compared with the list price for candesartan 32 mg (£16.13; 28-day pack price). This fails to acknowledge that in the treatment of hypertension, most patients are treated at doses of 16 mg or less and that by April 2012 candesartan will lose its patent with consequent significant cost implications. The authors also suggest that the cost of generic losartan may drop to around £0.88 per pack; this seems somewhat unrealistic based upon the costs of longer established generic formulations of ramipril, perindopril and lisinopril (10). Consideration should also be given to the fixed costs associated with switching such as the need to call the patient back, issuing new prescriptions, additional laboratory tests after changing, checking BP a month later, and possibly adding further medication if BP is no longer controlled. These costs need to be offset against the gains. In conclusion, although we welcome the potential cost savings that may be achieved by switching between alternative equivalent therapies, this approach must not be at the expense of poorer outcome, and caution must be exercised in the appropriate application of cost-utility analyses. PAM produced the first draft of the letter. SCJ modified and added to the draft. PAM produced a final version to conform with the appropriate word count. PAM has received research grants, honoraria for consultancy, advisory board attendance and speaker fees from a number of proprietary and generic pharmaceutical companies, including AstraZeneca, Bayer Schering Pharma, Boehringer Ingelheim, GSK, MSD, Novartis, Pfizer and Takeda. SCJ has received honoraria for consultancy, advisory board attendance and speaker fees from proprietary and generic pharmaceutical companies, including AstraZeneca, Boehringer Ingelheim, MSD, Lilly and Takeda.