Developing appropriate methods for handling missing data in health economic evaluation.

Policy makers worldwide use health economic evaluation to help decide which health care interventions to provide. An important concern faced by economic evaluation studies is that there may be missing data. However, most published studies ignore this or use inappropriate methods to address the problem. If missing data are not addressed appropriately this can lead to misleading cost-effectiveness results and scarce health care resources being misallocated. Careful analytical methods are required to address missing data across different circumstances in health economic evaluation.

I propose to conduct a comprehensive programme of research to address this gap in knowledge, using both simulation work and data from clinical areas of high policy relevance. The research will develop and test alternative methods for addressing missing data and show the impact that better methods can have in the evaluation of health interventions and health care providers. Methods will be thoroughly disseminated to applied researchers and policy makers to improve current practice. By helping improve the quality of cost-effectiveness studies which are used to inform policy making, this research will help ensure that scarce resources are allocated in the best ways for improving the population's health.

Motivation: Missing data raise outstanding methodological concerns in health economic evaluation. The main issue is that individuals with missing costs or outcomes tend to be systematically different from those with complete data. Multiple imputation (MI) have been proposed for handling missing data in cost-effectiveness analyses (CEA), but this MI approach may be insufficient in settings such as multicentre trials and non-randomised studies (NRS). If missing data are not addressed appropriately this can lead to biased cost-effectiveness results and incorrect estimates of uncertainty.

Aim: To develop appropriate methods for handling missing data in health economic evaluation.

Methodology: Firstly, multilevel MI and full-Bayesian hierarchical approaches will be developed and tested in a simulation study for handling missing data in CEA that use multicentre trials. Case studies in intensive care will be used to investigate whether cost-effectiveness results change according to method. Secondly, the performance of MI and full-Bayesian approaches will be compared with that of DR methods, which may be attractive in NRS where the imputation model may be more difficult to specify. Implications of method choice will be assessed in the evaluation of medical devices, alternative forms of elective surgery and providers' performance. Thirdly, the project will develop a framework for exploring the robustness the cost-effectiveness results to departures from the missing at random assumption. The research will develop approaches such as full-Bayesian selection models and pattern mixture models that explicitly address possible missing not at random mechanisms.

Research output: This project will provide insights about the relative merits of alternative methods for handling missing data across different settings in health economic evaluation. Methods will be widely disseminated so that future studies can provide a sound basis for policy making.

This research is anticipated to benefit a larger audience beyond the academic community, including the industry, public sector, those specifically involved in the policy making, and through these, the wider public.

This project will improve the methods for handling missing data in health economic evaluation so that future studies can provide a sound basis for policy making. By improving the quality of economic evaluation studies, which are typically used by policy makers to inform decisions on resource allocation, this research will help ensure that scarce societal resources are allocated in the best ways to improve population health. To maximise the impact of the methodological advances on improving research practice and health policy decisions, the methods will be disseminated directly to policy makers such as NICE. For example, I will encourage the dissemination of the methods by writing discussion papers for the NICE Decision Support Unit (as dicussed with the Director, Prof Alan Wailoo).

The methodologies developed in this project will also contribute to improving the evaluation of health care providers, a high priority area for policy makers. The research will engage members of the Policy Innovation Research Unit, a Department of Health-funded research unit. The director of PIRU, Prof Nick Mays has indicated that he and his colleagues will help me communicate the findings of the proposed research directly to policy makers and analysts at the Department of Health. A better assessment of the performance of providers will help inform future policy initiatives, for example, encouraging those providers with poor outcomes to improve their performance, which would contribute to better health and quality of life of the population.

The research will also help individuals working in industry and public sector by providing them with a better understanding about the implications that missing data pose to economic evaluation studies, and with the appropriate tools to deal with those issues. By encouraging more rigorous health economic evaluation that are relevant, for example, to the pharmaceutical industry and health policy-makers, this can lead to saved NHS resources and contribute to increasing the productivity in the UK economy.

The activities undertaken to facilitate maximum impact of the research to the beneficiaries identified here are detailed in the attachment "Pathways to Impact".