Cost of illness: guidelines for best practice methodology

Abstract

Cost of illness (COI) studies look at the value of resources expended or forgone as the result of an illness. 'Illness' may refer to specific diseases (e.g. cancer), injuries (e.g. motor vehicle accidents), risk factors (e.g. smoking) or groups of health conditions (e.g. sleep disorders).

COI studies have used inconsistent methodologies and findings can be difficult to compare historically and internationally, with many articles in the literature noting the need for more consistent typology. Costs may be calculated with different perspectives, purposes, timeframes or approaches and there is relatively little guidance to assist readers in evaluating methods, leading to some concerns about reliance on the findings. The overarching research question that the dissertation seeks to address is thus: Given the heterogeneity in COI methods historically, what best practice guidelines and tools can be used in future that could assist interested parties in achieving greater consistency in approach, accuracy of findings and comparability across study results?

The thesis presents methods and findings from a literature review of published articles addressing COI methodology, undertaken to inform the issues encountered and their typical resolution. From this review, 33 propositions were derived and tested as a starting point for best practice guidelines, using a four-round Delphi panel process, with panel participants recruited globally from authors of previous COI studies and eminent health economists.

Using an anonymous survey, the Delphi process tested and refined the methodological guidelines, which cover aspects including purpose and perspective, cost classification and estimation techniques, incidence and prevalence approaches, timing, discounting, measuring production, efficiency and wellbeing losses, and dealing with uncertainty. The best practice guidelines were supplemented with tools comprising a cost matrix, taxonomy and checklist, with the purpose of this suite of instruments being to provide guidance to those wishing to undertake new COI research, as well as those wishing to evaluate or monitor the quality of past studies.

The guidelines and checklist were then applied to a case study to illustrate their application to a migraine COI for Australia in 2018, to estimate prevalence, health system costs, productivity losses for migraneurs and their carers, other tangible items (OTIs) and wellbeing losses. Results and sensitivity analysis from the COI were discussed in the context of comparator studies and the guidelines, noting the implications of differences between the studies, including the direction of bias as well as low, medium and high consequence assessments of departing from the guidelines. Conclusions were thus able to be drawn about the subsequent use of the studies in economic evaluation, decision making and policy settings.

The highest consequence areas of departing from guidelines were found to relate to: failure to adopt a comprehensive (societal) perspective; failure to adopt a consistent cost item taxonomy or matrix of costs (type by bearer of cost categories); attribution error due to failure to correctly adjust for sample bias or confounding factors; and failure to correctly calculate and include production and wellbeing losses. These issues are likely to render some COIs unfit for purpose.

In contrast, it is argued that studies that adhere to the guidelines and tools are more suitable for use in health economic evaluations and decision making, engendering more optimal allocation of resources when weighing the benefits of averting cases of illness against the costs of preventive or treatment interventions. It is hoped that, following future publication and awareness raising, the guidelines and tools will be broadly adopted to better inform the evidence base and thus help prioritise health research and service delivery activities. Show more