Related Experiment Video
Updated: Jan 13, 2026

Inverse Probability of Treatment Weighting Propensity Score using the Military Health System Data Repository and National Death Index
Published on: January 8, 2020
Estimating effects of targeted public health interventions using the interventional disparity indirect effect among
Amalie Lykkemark Møller1,2,3, Kathrine Kold Sørensen4, Nerissa Nance5,6
1Cancer and Medicine, Danish Cancer Institute, Copenhagen, Denmark amlm@sund.ku.dk.
Abstract:
Disadvantaged groups are often defined by characteristics such as income or ethnicity. Reducing health disparities by directly manipulating such exposures may be infeasible. Instead, interventions can target mediators between these exposures and health outcomes. Indirect effects estimated using mediation analysis, interventional effects or the interventional disparity measure can quantify the expected impact of such disparity-reducing interventions. They capture the impact of changing the mediator distribution evaluated among the total population. This means keeping individuals in their exposure group but hypothetically assigning them the mediator distribution of another group. However, when indirect effects are intended to inform about disparity-reducing interventions implemented among disadvantaged groups, estimating effects in the total population does not quantify the effect among those targeted. Instead, we propose evaluating the interventional disparity indirect effect directly among the disadvantaged individuals. We introduce the estimand and illustrate it using a register-based study examining a potential intervention improving medication initiation in low-income heart failure patients. We compare the expected change in 1-year mortality in a hypothetical world where low-income patients were as likely to initiate medication as high-income patients. We included 1700 patients and assessed intervention effects in low-income patients and the total population, respectively. Under the intervention, the 1-year mortality declined from 10.3% to 9.3% (95% CI 8.6% to 10.1%) among low-income patients but 6.6% to 6.2% (95% CI 6.0% to 6.5%) in the total population. In disparity research, evaluating intervention effects in the total population, rather than among disadvantaged groups, may impact the effect size. Therefore, when guiding future disparity-targeted interventions, measuring effects within disadvantaged groups is important.
More Related Videos
05:59Visualization of Intensity Levels to Reduce the Gap Between Self-Reported and Directly Measured Physical Activity
Published on: March 7, 2019
08:36Author Spotlight: Evaluating the Adjuvant Efficacy and Safety of Angong Niuhuang Pill in Viral Encephalitis Treatment
Published on: April 19, 2024
Related Concept Videos
Strategies for Assessing and Addressing Confounding
Confounding can be addressed at both the design phase of a study and through analytical methods after data...
Confounding in Epidemiological Studies
What is an Experiment?
Bias in Epidemiological Studies
Types of Biopharmaceutical Studies: Controlled and Non-Controlled Approaches
Non-controlled studies, commonly employed for initial exploration, lack a control group, rendering them susceptible to biases and external influences. In contrast,...
Statistical Methods for Analyzing Epidemiological Data