A causal framework for evaluating the total effect of strategies aiming to expand screening and to improve outcomes

TL;DR

This work proposes a novel extension of Two-Stage targeted minimum loss-based estimation (TMLE), and uses Counterfactual Strata Effects: causal estimands where the outcome is only relevant for a group whose membership is subject to missingness and/ or impacted by the exposure of interest.

Abstract

For many health conditions, there are highly efficacious treatment and prevention products. Maximizing their impact requires strategies that improve the reach of health screening in order to establish who could benefit. For example, HIV prevention strategies aim to expand risk screening and to improve uptake of pre-exposure prophylaxis (PrEP) among those experiencing risk. Often, these strategies induce changes at the group-level (e.g., health clinics or communities) and are evaluated through cluster randomized trials. This scenario creates a complex, multilevel-mediation-missing data problem for the following reasons. First, the strategy is delivered at the cluster-level, while health screening and outcomes are at the individual-level. Second, the strategy improves health outcomes directly and indirectly through improved health screening. Third, everyone has an underlying status, which is only observed among those screened. To formally define the total effect in such settings, we use Counterfactual Strata Effects: causal estimands where the outcome is only relevant for a group whose membership is subject to missingness and/ or impacted by the exposure of interest. To identify and estimate the corresponding statistical estimand, we propose a novel extension of Two-Stage targeted minimum loss-based estimation (TMLE). Simulations demonstrate the practical performance of our approach as well as the limitations of existing approaches.

Figures (3)

• Figure 1: Simplified causal graph for the OPAL trial to increase HIV risk screening and PrEP use among persons with HIV risk: individuals are nested within clusters; HIV risk screening mediates the intervention effect, and HIV risk status is missing (unobserved) if an individual does not screen. New methods are needed to evaluate overall effectiveness, while accounting for the multilevel-mediation-missing data problem.
• Figure 2: A simplified hierarchical causal graph to illustrate the data generating process in a CRT where the intervention strategy expands health screening and improves health outcomes. Since the causal model is defined at the cluster-level, we drop the subscript $j$ throughout. For simplicity, we have omitted the unmeasured factors: $\bm{U}=(U_{E^c}, U_{\bm{W}}, U_{A^c}, U_{\bm{Y1}^*}, U_{\bm{\Delta}}, U_{\bm{Y2}})$.
• Figure 3: A simplified hierarchical causal graph to illustrate the data generating process where the cluster-level intervention $A^c$ impacts time-varying covariates $\bm{L}$ as well as underlying status $\bm{Y1}^*$. The time-varying covariates $\bm{L}$ also impact measurement $\bm{\Delta}$, underlying status $\bm{Y1}^*$, and the outcome $\bm{Y2}$. For simplicity, we have again omitted the unmeasured factors: $\bm{U}=(U_{E^c}, U_{\bm{W}}, U_{A^c},U_{\bm{L}}, U_{\bm{Y1}^*}, U_{\bm{\Delta}}, U_{\bm{Y2}})$.