Addressing geographic confounding through spatial propensity scores: a study of racial disparities in diabetes

Abstract

Motivated by a study exploring differences in glycemic control between non-Hispanic black and non-Hispanic white veterans with type 2 diabetes, we aim to address a type of confounding that arises in spatially referenced observational studies. Specifically, we develop a spatial doubly robust propensity score estimator to reduce bias associated with geographic confounding, which occurs when measured or unmeasured confounding factors vary by geographic location, leading to imbalanced group comparisons. We augment the doubly robust estimator with spatial random effects, which are assigned conditionally autoregressive priors to improve inferences by borrowing information across neighboring geographic regions. Through a series of simulations, we show that ignoring spatial variation results in increased absolute bias and mean squared error, while the spatial doubly robust estimator performs well under various levels of spatial heterogeneity and moderate sample sizes. In the motivating application, we construct three global estimates of the risk difference between race groups: an unadjusted estimate, a doubly robust estimate that adjusts only for patient-level information, and a hierarchical spatial doubly robust estimate. Results indicate a gradual reduction in the risk difference at each stage, with the inclusion of spatial random effects providing a 20% reduction compared to an estimate that ignores spatial heterogeneity. Smoothed maps indicate poor glycemic control across Alabama and southern Georgia, areas comprising the so-called "stroke belt." These results suggest the need for community-specific interventions to target diabetes in geographic areas of greatest need.

Propensity score matching for multilevel spatial data: accounting for geographic confounding in health disparity studies

“…Moreover, while this approach works well for studies of atmospheric data and environmental exposures that have broad impacts over a large spatial domain, patients in our study are clustered into discrete areal units (e.g., counties) that may share community resources that impact health more locally. We therefore extend previous work in propensity score weighting [ 27 ] and an areal-level spatial propensity score matching framework [ 28 ] that encourages matching among individuals in adjacent spatial units. While recent work suggests benefits to within-cluster matching [ 29 ], this recommendation is not easily extended to the spatial setting.…”

“…At the same time, evidence shows that racial minorities have a higher prevalence of diabetes [ 9 ], poorer diabetes outcomes [ 10 , 27 ], and higher mortality rates compared to non-Hispanic whites [ 11 ]. These disparities are explained in part by individual demographics, such as age, sex and marital status [ 12 , 13 ] However, patient demographics may explain only one piece of the puzzle.…”

A Review of Spatial Causal Inference Methods for Environmental and Epidemiological Applications

“…Eliminating this feedback can be done by fitting the model in two stages, that is, first fitting the model for the treatment indicators in Equation ( 10) to obtain an estimate of v and then fitting Equation ( 9) with v fixed at its first-stage estimate. Other possible remedies include 'cutting feedback' in the steps of the MCMC algorithm (Lunn et al, 2009;McCandless et al, 2010) or post-hoc reweighting of the posterior distribution (Saarela et al, 2015;Davis et al, 2019). These methods are discussed below.…”

“…Schnell & Papadogeorgou, 2020) and propensity-score methods (e.g. Davis et al, 2019). We review these methods and conduct a simulation study to compare their precision for estimating a causal treatment effect in the presence of a missing spatial confounding variable.…”

Selecting a Scale for Spatial Confounding Adjustment

“…In a causal inference framework, confounding due to geography has been recently addressed using spatial propensity scores (Papadogeorgou et al, 2019;Davis et al, 2019), however in those settings the exposure of interest was not explicitly spatial.…”