Despite the dual objectives of many health care systems of improving total health and reducing health inequality, trial designs seem to ignore the assessment of inequality effects. Our study aimed to illustrate an empirical framework for the assessment of inequality effects alongside policy-oriented trials to inform a possible efficiency versus equality trade-off.
We measured inequality in the concentrations of all-cause and disease-related mortality and hospital admissions across ranks of socioeconomic status in a randomized controlled trial that tested the efficacy of general population screening of men for vascular disease. We used alternative definitions of inequality (relative/absolute, in attainment/shortfall, ranked by education/income), and supplemented the classical ‘frequentist’ approach to statistical inference with Bayesian posterior probabilities. Equality contours for health improvement that leave inequality unaffected are illustrated graphically. We used bootstrapping for interpretation.
We estimated the posterior probability of screening increasing inequality to be between 0.21 and 0.93 depending on the inequality definition. Income-ranked inequality appeared to be generally higher than education-ranked inequality but less affected by screening. For the shortfall-relative index based on education-rank, the mean health improvement of a 7% relative reduction in all-cause mortality generated by screening incurred a mean relative increase in inequality of 28%. For the income-based indices there was no evidence of a trade-off.
We illustrated how decision uncertainty can be reduced by explicit assessment of inequality alongside trials and found some evidence of a possible equity–efficiency trade-off in the context of screening, although this depended on the definition of equality.
1Department of Public Health, Aarhus University, 8000 Aarhus C, Denmark
2Department of Clinical Medicine, Aarhus University, 8200 Aarhus N, Denmark
3Department of Cardiothoracic and Vascular Surgery, Odense University Hospital, 5000
4Department of Vascular Surgery, Viborg Hospital, 8800 Viborg, Denmark
Conflicts of Interest: None declared
Source of Funding: This work was supported by The Danish Council for Independent Research [grant number DFF-4004-00274]
Data sharing: Access to the data can be requested by applying for permission from the Danish
Data Protection Agency and Statistics Denmark.
Acknowledgments: The authors would like to thank three anonymous reviewers whose insightful suggestions substantially helped improve the paper.
Corresponding author: Rikke Søgaard, Department of Public Health, Aarhus University, Bartholins Allé 2, 8000 Aarhus C, Denmark, E-mail firstname.lastname@example.org, Phone +45 28991387