Identifying and estimating effects of sustained interventions under parallel trends assumptions

06/12/2022
by   Audrey Renson, et al.
0

Many research questions in public health and medicine concern sustained interventions in populations defined by substantive priorities. Existing methods to answer such questions require a measured covariate set sufficient to control confounding, which can be questionable in observational studies. Differences-in-differences relies instead on the parallel trends assumption, allowing for some types of time-invariant unmeasured confounding. However, existing difference-in-differences implementations are limited to point treatments in restricted subpopulations. We derive identification results for population-wide effects of sustained treatments under parallel trends assumptions. In particular, in settings where all individuals begin follow-up with exposure status consistent with the treatment plan of interest but may deviate at later times, a version of Robins's g-formula identifies the intervention-specific mean under SUTVA, positivity, and parallel trends. We also develop consistent asymptotically normal estimators based on inverse-probability weighting, outcome regression, and a double robust estimator based on targeted maximum likelihood. Simulation studies approximately confirm theoretical results and support the use of the proposed estimators at realistic sample sizes. As an example, the methods are used to estimate the effect of a hypothetical federal stay-at-home order on all-cause mortality during the COVID-19 pandemic in Spring-Summer 2020 in the United States.

READ FULL TEXT
research
07/26/2022

When Do Outcome Driven Treatments Break Parallel Trends?

Under what circumstances is it a threat to the parallel trends assumptio...
research
05/07/2021

Adjusting for Unmeasured Confounding in Marginal Structural Models with Propensity-Score Fixed Effects

Marginal structural models are a popular tool for investigating the effe...
research
11/04/2021

Estimating SARS-CoV-2 Seroprevalence

During the COVID-19 pandemic, governments and public health authorities ...
research
11/09/2020

Reducing bias in difference-in-differences models using entropy balancing

This paper illustrates the use of entropy balancing in difference-in-dif...
research
11/15/2022

Time-varying Bayesian Network Meta-Analysis

The presence of methicillin-resistant Staphylococus Aureus (MRSA) in com...
research
02/14/2018

Using Longitudinal Targeted Maximum Likelihood Estimation in Complex Settings with Dynamic Interventions

Longitudinal targeted maximum likelihood estimation (LTMLE) has hardly e...
research
05/30/2022

Mediation Analysis with Multiple Mediators under Unmeasured Mediator-Outcome Confounding

It is often of interest in the health and social sciences to investigate...

Please sign up or login with your details

Forgot password? Click here to reset