Authors’ reply to Solomon writing on behalf of 18 othersBMJ 2012; 344 doi: http://dx.doi.org/10.1136/bmj.e2588 (Published 10 April 2012) Cite this as: BMJ 2012;344:e2588
- 1Department of Primary and Community Care, Radboud University, Nijmegen Medical Centre, Netherlands
- 2International Centre for Evidence on Disability, London School of Hygiene and Tropical Medicine, London, UK
- 3Department of Health Systems Financing, WHO, Geneva, Switzerland
Solomon and colleagues inappropriately label our analysis of trachoma control erroneous, but their comments are largely related to methodological issues in (standardised WHO-CHOICE) cost effectiveness analysis and present gaps in knowledge.1 2
Firstly, mass treatment with antibiotics of all residents is probably more effective and costly than treatment of children. Whether the first intervention is cost effective according to international thresholds, and more cost effective than trichiasis surgery, is an area for further research.
Secondly, to derive internally consistent results, we retrieved costs and effectiveness from a single study, while verifying that effectiveness estimates were similar to those from other countries.3 Sensitivity analyses showed robustness of study conclusions.2 4
Thirdly, WHO-CHOICE analyses use different coverage rates to capture (dis)economies of scale in intervention costs. Coverage rates do not affect effectiveness estimates.
Fourthly, the natural course of trachoma and benefits of intervention are uncertain. Our model, developed and published with trachoma control experts, provides a first quantification of these benefits and captures the most important ones.4
Fifthly, although we agree that, in principle, cost effectiveness analysis should include any future effect, there is much uncertainty around these effects and their discounting. This may explain why analyses of—for example, measles elimination programmes—use time horizons up to 40 years. WHO-CHOICE uses a time horizon of 100 years.
Lastly, the valuation of donated drugs hinges on assumptions about whether donations will continue. Regrettably, our careful discussion of this and many other disease specific debates cannot adequately be reflected in the summary paper.2 4 5 We previously reported that treatment of children in a zero cost scenario would be cost effective,4 and it would be fair to evaluate current programmes under the zero cost scenario. This would show their ranking compared with other trachoma control interventions and establish the basis for efficient trachoma control.
Cite this as: BMJ 2012;344:e2588
Competing interest: None declared.
This is a shortened version of a rapid response, which appears in full at: www.bmj.com/content/344/bmj.e586/rr/577307.