A statistical tool to predict community deprivation in health care access: Based on a model of combined personal and ecologic characteristics

Context:

Past attempts to explain barriers/delays in accessing healthcare have either used personal characteristics for individuals, or ecological measures like poverty level as near-proxies for health access problems in populations. Few have combined both methods.

Objective:

To develop a simple tool to identify levels of healthcare access deprivation or difficulty accessing care for individuals in geographically small areas in the US. To validate our tool using regional Ohio survey data and create US map of access deprivation.

Methods:

We modeled 2001-2005 NHIS data to identify individual-level predictors of deprivation, and used multi-level logistic modeling techniques to pair-wise substitute Census for NHIS variables. We constructed census-tract-level index of access deprivation, using expected values of dependent variable in best-fit model. We validated our tool using primary data from a regional Ohio survey, mapped the index nationally, and overlaid it with health care provider availability statistics etc.

Results:

We found deprivation more likely if one has functional limitation, no insurance, no phone at home, is under 65, in fair/poor health, lives in census tract where most do not own a home, and live far from a community health center. Our raw census-tract deprivation index ranged from 6.3% to 40.9%. The results include derived deprivation index for each of 66,997 U.S. census tracts, a series of US national maps of all census-tracts indicating tracts with possible health access deprivation, overlaid with provider availability, etc.

Conclusions:

This tool could assist local community assessments, health planning, and identifying populations at risk