Incorporating health spillovers into economic evaluation: Theory, estimation, and practice
Henry, Edward
Henry, Edward
Repository DOI
Publication Date
2024-08-27
Type
interactive resource
Downloads
Citation
Abstract
The health status of an individual is increasingly recognised as having the potential to affect the health and well-being of a cohort of ‘significant others’, primarily encompassing their caregiving and non-caregiving family members. Healthcare interventions that improve the lives of patients may, therefore, also offer benefits, beyond those intended, to this wider network. When these ‘health spillovers’, which may result from the health condition, intervention, or both, are accounted for in health economic evaluation, they have often been shown to be consequential to the resulting funding recommendations. However, health spillover inclusion in cost-effectiveness analysis remains uncommon. This omission appears to relate to a range of calls for caution in incorporating spillovers, owing to: their non-trivial quantification, the challenge they present to the allocative justice of healthcare decision making, and the current lack of methodological guidance to inform incorporation efforts.
Accordingly, this thesis seeks to address each of these concerns directly, and in doing so, assess the extent to which an individual's serious illness impacts upon the health of their family members, and establish how such estimates might be incorporated consistently and equitably into health economic evaluation. Four specific goals, aligned with this overarching objective, form the basis for the four separate essays contained within the thesis.
Given that health spillovers are thought to exert a psychic or emotional toll among both caregiving and non-caregiving family members, Chapter 2 considers the relationship between serious family illness and the mental health of family members, independent of their caregiving experience. The analysis makes use of Irish EQ-5D-5L population normative data to examine the independent association between experience of serious family illness and various dimensions of health-related quality of life, including anxiety and depression. To do so, it combines inverse probability weighting and multivariate ordered probit regression in a doubly robust estimation of this relationship. Experience of serious family illness is observed to be strongly associated with mental health decrements that are independent of caring responsibilities and various other personal, household, and socioeconomic characteristics. The estimated decrements are also shown to be consistent with serious family illness leading to an average reduction of 1.3% in health utility for non-caregiving family members, implying deteriorations in population health beyond those more overtly attributable to serious illness.
There is, however, little in the way of methodological guidance informing how such spillovers are to be accounted for in the economic evaluation of healthcare interventions. To address this gap, Chapter 3 describes the proceedings and consensus recommendations of the Spillovers in Health Economic Evaluation and Research task force. These recommendations, developed via modified nominal group technique, seek to inform the health economics and outcomes research community on emerging good practice in incorporating family and caregiver health spillovers into cost-effectiveness and cost-utility analysis. The chapter delineates 11 consensus recommendations for emerging good practice currently, and in view of the still-nascent evidence base, 12 avenues for future research.
One of the described priorities for future research relates to the distributional impacts of spillover inclusion in economic evaluation, which may challenge the equity of resource allocation decisions. Thus, Chapter 4 presents a novel approach to spillover inclusion in health economic evaluation which, by employing a ‘prioritarian transformation’ of health gains, offers one means of moderation of these distributional impacts. The chapter introduces a normative, theoretical perspective to a predominantly empirical extant literature, describing the axiomatic basis for the method, and its application via an illustrative example.
In light of the importance of the parent-child dyad amongst family relationships, and a requirement for spillover quantification across the spectrum of disease, the final essay focusses on the estimation of maternal mental health spillovers associated with various forms of child illness/disability, this time employing longitudinal data and a system Generalized Method of Moments approach. Evidence of substantial mental health spillovers is observed among mothers of children experiencing chronic mental and behavioural disorders, nervous system disease, or endocrine system disease. The results also indicate that, in general, spillover magnitude is strongly related to the severity of the illness/disability. An extension to the main analysis considers how spillovers might vary depending on the level of household deprivation, and suggests that spillovers likely manifest to a greater extent in families that are financially deprived.
Publisher
University of Galway
Publisher DOI
Rights
Attribution-NonCommercial-NoDerivatives 4.0 International