CHEERS is Sufficient for Reporting Cost-Benefit Analysis, but May Require Further Elaboration

HUSEREAU, D.; DRUMMOND M; PETROU, S.; GREENBERG, D.; MAUSKOPF, J.; AUGUSTOVSKI, F.; BRIGGS, A. H.; MOHER, D.; LODER, E.; CARSWELL, C.

PHARMACOECONOMICS

ADIS INT LTD

Año: 2015

1170-7690

Roberts and colleagues make the point that our Consolidated Health Economic Evaluation Reporting Standards (CHEERS) checklist may be insufficient for reporting costbenefit analysis (CBA) in biomedical journals. In particular, they believe CHEERS may need an extension tosupplement potentially missing reporting items required for CBA [1].We were pleased to hear CHEERS was suitable for reporting a cost-utility analysis (CUA) of the same intervention and that the concerns raised are practical, rather than theoretical, and came from using the checklist.CHEERS is still a relatively new instrument that has enjoyed steady uptake, having been cited over five times a month since its publication in 2013 and endorsed in 12 journals (that we are aware of). We would encourage all authors, editors and peer reviewers to provide feedback tothe CHEERS Task Force to help improve it after reading our explanation and elaboration document [2] and putting the CHEERS checklist to practical use. However, we are not entirely convinced by Roberts and colleagues? arguments that CHEERS has ?significant gaps?. While it is true that CHEERS leans strongly toward providing direction to those conducting cost-effectiveness analyses (CEAs) (including CUA), we would argue that the checklist adequately captures the reporting of a CBA. Forexample, Item 12 asks authors to report methods of valuing preference-based outcomes and assumes that authors of a CBA would be concerned with valuation of benefits based on preferences for the consequences of different interventions and using methods compatible with standard welfare economic theory. Item 19 then asks authors to report theirfindings separately for costs and benefits (losses and gains). It is not clear from this letter, then, what additional information included in a CBA would not fit into these and other items in CHEERS.We would acknowledge that CHEERS reflects the current emphasis on CEA approaches in the literature. For example, the recommendation for Item 12 describes methods to elicit preferences although those conductingCBAs may use a revealed preference approach. Similarly, Item 10 asks authors to report their choice of health outcomes, implying those outcomes are incorporated in a CBA and valued in monetary terms. We recognize that, inprinciple, CBAs in the health field could also considerbenefits from interventions that are unrelated to health, although many healthcare decision makers, the main audience for these studies, do not favour this approach. Rather than the checklist having missing items, wewould suggest that further elaboration of items might be needed to ensure that the application of the checklist to the area of CBA is fully specified. Elaborations of other reporting guidelines have been developed where the need for further consistency in reporting has been identified [3, 4].The many potential approaches to valuing outcomes in CBAs, including discrete choice and willingness-to-pay approaches, require reporting of details that will vary by method. As Roberts and colleagues point out, a guide to reporting stated preference discrete choice experiments inCBA has already been developed [5].We must also not lose sight of the small prevalence and impact of currently published CBAs. While publication of CEAs and CUAs in particular continues to rise with their use in decision-making [6], very few CBAs have been published recently, apart from those labelled ?CBAs? but valuing benefits only in terms of cost offsets or productivitygains. In reviewing the literature to update the Methods for the Economic Evaluation of Health Care Programmes textbook [7], one of us (MD) found only a single published CBA valuing benefits using contingentvaluation since the release of the third edition, whereas there is growing literature concentrating only on measuring and valuing benefits, using contingent valuation or conjoint analysis.Despite the current predominance of CEA compared with CBA for evaluating new healthcare interventions, we still contend that CHEERS should be neutral about what methods of economic evaluation are most appropriate and be applicable to all forms of evaluation. To this end, wethink the comments made by Roberts and colleagues are helpful and are likely to inform future revisions or elaborations of the CHEERS checklist. We would certainly welcome and encourage collaboration between the CHEERS Task Force and those interested in producing aCHEERS elaboration document focussing on CBA.Conflict of interest Don Husereau is Chair of the CHEERS Task Force, and has served on the Board of Directors for the study funder (ISPOR) as well as received directed contracts for unrelated work from ISPOR and support for travel to a face-to-face meeting to discuss the contents of the report. All authors have completed the ICMJE uniform disclosure from at www.icmje.org/coi.disclosure.pdf and declare: FA and MD have also served as board members for the study funder; FA, AHB, CC, MD, DG, DH, EL, JM, and SP were provided support for travel to a face-to-face meeting to discuss the contents of the report; FA and MD have received payment from the study sponsor for serving as co-editors for Value in Health; no other relationships or activities that could appear to have influenced the submitted work have occurred. CC is the editor of this journal.