Critique of an Economic Evaluation Using the Drummond Checklist
Doran, Christopher M., Applied Health Economics and Health Policy
In this issue of Applied Health Economics and Health Policy, Sladkevicius et al. report the results of an economic evaluation of establishing a neonatal screening programme for phenylketonuria (PKU) in Libya. The authors undertook a cost-effectiveness analysis (CEA) and a return on investment analysis. Against a background of limited resources and rational decision making, economic evaluations have been used more frequently over the past 2 decades as pressures to control healthcare spending have increased. The central purpose of an economic evaluation is to compare the relative value of different interventions in creating better health, longer life or a return on investment. The results of such evaluations are typically summarized as a ratio, where the denominator reflects the gain in health from an intervention (measured as units natural to the programme at hand, such as life-years gained, for a CEA or converted to monetary terms for a cost-benefit analysis [CBA]), and the numerator reflects the cost of obtaining that health gain. A variant of CEA, cost-utility analysis (CUA) measures the gain according to health-state preference scores or utility weights, thus providing an opportunity to measure the quality of life-years gained (or disability-adjusted life-years averted), not just the crude number of years.
Given the rise in popularity of the use of economic evaluations, guidelines have been developed to improve the methodological rigour, and hence comparability, of each evaluation. Two of the more commonly cited references are the Washington Consensus Panel on Cost-Effectiveness in Health and Medicine and a textbook by Drummond et al. Both provide recommendations for the conduct of economic evaluations in order to improve their quality. Drummond et al. developed a checklist for assessing economic evaluations, which provides useful guidance for those undertaking their own study. It is in the general context of these guidelines that the contribution by Sladkevicius et al. is critiqued.
1. Was a Well Defined Question Posed in Answerable Form?
The Libyan healthcare system is faced with resource constraints, with only 3.5% of total GDP being allocated to health. Libya's population is characterized by large families, high maternal and paternal age and high rates of consanguineous marriages, which are thought to increase the prevalence of disabling and expensive inherited metabolic diseases. With the evident success of neighbouring countries in initiating neonatal screening, primarily for PKU, which can lead to a patient having a healthy life, the Libyan Government is considering a neonatal screening programme for PKU. Against this rationale, the authors' research question is timely, relevant and clearly stated as an examination of the cost effectiveness of establishing a national neonatal screening programme for PKU in Libya from a societal perspective.
2. Was a Comprehensive Description of the Competing Alternatives Given?
The authors provide an excellent overview of the current management of PKU in Libya and model the proposed screening intervention on a protocol produced by the Medical Advisory Panel of the National Society for Phenylketonuria, following adaptation to reflect the Libyan infrastructure.
3. Was the Effectiveness of the Programme Established?
The authors based their results on a decision model constructed in Microsoft® Excel. The model was populated with data collected from experts, families of patients with PKU and secondary data sources. The primary outcome used in the economic evaluation was the incremental number of life-years gained. The time-horizon of the model comprised the shelf life of the screening equipment and the lifetime of patients with PKU. In a sensitivity analysis, the authors varied, among other things, the potential incidence of PKU and life expectancy of treated PKU patients.
4. Were All the Important and Relevant Costs and Consequences for Each Alternative Identified? …