Analysis of racial differences in hospital stays in the presence of geographic confounding.

Publisher: Elsevier Country of Publication: Netherlands NLM ID: 101516571 Publication Model: Print-Electronic Cited Medium: Internet ISSN: 1877-5853 (Electronic) Linking ISSN: 18775845 NLM ISO Abbreviation: Spat Spatiotemporal Epidemiol Subsets: MEDLINE

Original Publication: Amsterdam : Elsevier, 2009-

Diabetes Mellitus, Type 2*/ethnology ; Diabetes Mellitus, Type 2*/therapy; Black or African American/*statistics & numerical data ; Healthcare Disparities/*statistics & numerical data ; Length of Stay/*statistics & numerical data ; Veterans/*statistics & numerical data ; White People/*statistics & numerical data; Confounding Factors, Epidemiologic ; Female ; Humans ; Male ; Middle Aged ; Propensity Score ; Spatial Analysis ; United States/epidemiology

Using recent methods for spatial propensity score modeling, we examine differences in hospital stays between non-Hispanic black and non-Hispanic white veterans with type 2 diabetes. We augment a traditional patient-level propensity score model with a spatial random effect to create a matched sample based on the estimated propensity score. We then use a spatial negative binomial hurdle model to estimate differences in both hospital admissions and inpatient days. We demonstrate that in the presence of unmeasured geographic confounding, spatial propensity score matching in addition to the spatial negative binomial hurdle outcome model yields improved performance compared to the outcome model alone. In the motivating application, we construct three estimates of racial differences in hospitalizations: the risk difference in admission, the mean difference in number of inpatient days among those hospitalized, and the mean difference in number of inpatient days across all patients (hospitalized and non-hospitalized). Results indicate that non-Hispanic black veterans with type 2 diabetes have a lower risk of hospital admission and a greater number of inpatient days on average. The latter result is especially important considering that we observed much smaller effect sizes in analyses that did not incorporate spatial matching. These results emphasize the need to address geographic confounding in health disparity studies.
(Published by Elsevier Ltd.)

Stat Med. 2016 May 30;35(12):2074-91. (PMID: 26833893)
Stat Med. 2009 Jan 15;28(1):94-112. (PMID: 19012268)
Health Serv Res. 2000 Feb;34(6):1351-62. (PMID: 10654835)
Stat Med. 2013 Aug 30;32(19):3373-87. (PMID: 23526267)
JAMA. 2002 Mar 13;287(10):1288-94. (PMID: 11886320)
BMJ. 2015 Apr 01;350:h1480. (PMID: 25833157)
Biometrics. 2015 Mar;71(1):208-217. (PMID: 25257036)
J Gen Intern Med. 2007 Sep;22(9):1231-8. (PMID: 17594129)
N Engl J Med. 1999 Jan 7;340(1):32-9. (PMID: 9878643)
JAMA. 2015 Sep 8;314(10):1021-9. (PMID: 26348752)
J Gen Intern Med. 2010 Sep;25(9):930-5. (PMID: 20429040)
J Epidemiol Community Health. 2006 Jul;60(7):578-86. (PMID: 16790829)
Annu Rev Public Health. 2013;34:61-75. (PMID: 23297653)
Stat Sci. 2010 Feb 1;25(1):1-21. (PMID: 20871802)
J Health Care Poor Underserved. 2009 Aug;20(3):896-913. (PMID: 19648715)
Health Serv Outcomes Res Methodol. 2016 Dec;16(4):271-292. (PMID: 27990097)
Multivariate Behav Res. 2011 May;46(3):399-424. (PMID: 21818162)
Diabetes Care. 1998 Jan;21(1):42-8. (PMID: 9538969)
Am J Manag Care. 2008 Mar;14(3):168-72. (PMID: 18333709)

I01 HX003577 United States HX HSRD VA; I01 HX002299 United States HX HSRD VA; P30 AR072582 United States AR NIAMS NIH HHS; T32 AR050958 United States AR NIAMS NIH HHS; U54 GM104941 United States GM NIGMS NIH HHS; UL1 TR001450 United States TR NCATS NIH HHS; P60 AR062755 United States AR NIAMS NIH HHS

Keywords: health disparities; propensity score matching; spatial data analysis

Date Created: 20190819 Date Completed: 20200428 Latest Revision: 20240530

20240530

PMC7359673

10.1016/j.sste.2019.100284

31421795