Advancing Health Equity in Economic Evaluations, With The Distributional Generalized Risk-Adjusted Cost-Effectiveness (DGRACE) Model
Author(s)
Emmanuel F. Drabo, PhD1, Michael DiStefano, PhD2, Jacqualine Woo, BS3, Shanshan Lin, MPH3, Ravi Gupta, MD, PhD4, Renee Wilson, MS4, Halima Amjad, MD, MPH, PhD5, Emily Mao, BS3, Jodi Segal, MD5;
1Johns Hopkins University, Assistant Professor, Baltimore, MD, USA, 2University of Colorado Anschutz Skaggs School of Pharmacy and Pharmaceutical Sciences, Department of Clinical Pharmacy, Aurora, CO, USA, 3Johns Hopkins University, Baltimore, MD, USA, 4Johns Hopkins University, Health Policy and Management, Baltimore, MD, USA, 5Johns Hopkins University School of Medicine, Baltimore, MD, USA
1Johns Hopkins University, Assistant Professor, Baltimore, MD, USA, 2University of Colorado Anschutz Skaggs School of Pharmacy and Pharmaceutical Sciences, Department of Clinical Pharmacy, Aurora, CO, USA, 3Johns Hopkins University, Baltimore, MD, USA, 4Johns Hopkins University, Health Policy and Management, Baltimore, MD, USA, 5Johns Hopkins University School of Medicine, Baltimore, MD, USA
OBJECTIVES: Standard cost-effectiveness analysis (CEA) assumes both constant returns to health and constant willingness-to-pay (WTP) threshold, without attention to the populations’ untreated illness severity and/or pre-existing disabilities or other distributional considerations. The generalized risk-adjusted cost-effectiveness (GRACE) model allows diminishing returns to health and variable WTP threshold by disease severity. However, GRACE does not address social inequities. Here, we demonstrate an extension to GRACE that incorporates these considerations.
METHODS: We developed the distributional GRACE (DGRACE) model, which extends GRACE to explicitly incorporate the social distribution of intervention effects in the decision-maker’s utility function, while retaining GRACE’s assumptions. Drawing on the inequality literature, we revised the expressions of the GRACE quality-adjusted life year (QALY) and GRACE-WTP threshold to integrate these equity considerations. We conducted simulations over a range of specifications of the utility function, inequality indices (e.g., Atkinson, Kolm, Gini), and values of the inequality aversion parameter, to illustrate the implications of our approach for decision-making, vis-à-vis GRACE, standard CEA, and distributional CEA.
RESULTS: Our model predicts that when the decision-maker has no aversion to inequality in consumption or health, or when the intervention’s effects are equally distributed across groups, the DGRACE-QALY and DGRACE-WTP revert to those of GRACE. Along with the desirable features of GRACE, the DGRACE approach values utility improvements in diseases with more unequal burden across subpopulations (e.g., Alzheimer’s disease and related dementia, HIV) more than utility improvements in diseases with less unequal burden.
CONCLUSIONS: Our DGRACE model offers a novel approach to assess the value of healthcare interventions and technologies and incentivize their innovation and adoption to advance health equity. However, further research is necessary to identify the specific social inequalities most pertinent to address, as well as the inequality aversion parameters and precise functional forms of the utility function that should be used in these valuations.
METHODS: We developed the distributional GRACE (DGRACE) model, which extends GRACE to explicitly incorporate the social distribution of intervention effects in the decision-maker’s utility function, while retaining GRACE’s assumptions. Drawing on the inequality literature, we revised the expressions of the GRACE quality-adjusted life year (QALY) and GRACE-WTP threshold to integrate these equity considerations. We conducted simulations over a range of specifications of the utility function, inequality indices (e.g., Atkinson, Kolm, Gini), and values of the inequality aversion parameter, to illustrate the implications of our approach for decision-making, vis-à-vis GRACE, standard CEA, and distributional CEA.
RESULTS: Our model predicts that when the decision-maker has no aversion to inequality in consumption or health, or when the intervention’s effects are equally distributed across groups, the DGRACE-QALY and DGRACE-WTP revert to those of GRACE. Along with the desirable features of GRACE, the DGRACE approach values utility improvements in diseases with more unequal burden across subpopulations (e.g., Alzheimer’s disease and related dementia, HIV) more than utility improvements in diseases with less unequal burden.
CONCLUSIONS: Our DGRACE model offers a novel approach to assess the value of healthcare interventions and technologies and incentivize their innovation and adoption to advance health equity. However, further research is necessary to identify the specific social inequalities most pertinent to address, as well as the inequality aversion parameters and precise functional forms of the utility function that should be used in these valuations.
Conference/Value in Health Info
2025-05, ISPOR 2025, Montréal, Quebec, CA
Value in Health, Volume 28, Issue S1
Code
EE29
Topic
Economic Evaluation
Topic Subcategory
Novel & Social Elements of Value
Disease
SDC: Geriatrics, SDC: Neurological Disorders