How systematic are reviews of health economics?

Article type
Year
Authors
Rigby K, Silagy C, Crockett A
Abstract
Introduction: The lack of a systematic approach to preparing review articles has been demonstrated across a variety of medical disciplines. Review articles which focus on health economic issues, such as cost-effectiveness, rely heavily on synthesising information about resource utilisation as well as effectiveness of the intervention (s) being examined. It is unclear the extent to which these articles are prepared using a systematic approach.

Objective: To examine the extent to which health economic review articles are compiled in a systematic manner.

Methods: All economic review articles published between 1991 and 1993 inclusive were included in the study providing the article was a review of the cost-effectiveness of a health care intervention, targeted at patients, and was based on the result of more than one RCT. In addition, articles had to be published in English and not include any cost of illness studies. Articles were reviewed independently by two reviewers against 17 criteria, derived from those previously used by Mulrow, with additional criteria adapted from Drummond. For each criteria, articles were given a score of between 0 and 2 depending on the amount of information present.

Results: Only 8 articles met all the inclusion criteria. The maximum score obtained by any article was 25 out of a maximum of 34 points. Only three of the articles obtained scores of 2 against four or more criteria. When examined against the economic criteria, at least 4 of the articles provided no information about opportunity costs, marginal analysis, currency conversions, sensitivity analysis, or discounting of costs.

Discussion: This study has identified that attempts to perform economic reviews of randomised controlled trials frequently lack a systematic approach. This is consistent with the findings of previous analyses of review articles in other fields, which have highlighted the failure to apply the same degree of rigour to this type of research synthesis that the scientific community has come to expect from primary research articles.