Utilizing a Geocoding Methodology to Improve Tribal Usage of Hospital Discharge Data

BACKGROUND:  An estimated 5.5 percent of the Arizona (AZ) population is American Indian/Alaska Native (AI/AN) according to the 2010 United States Census. Of the state’s 73 million acres, almost 27 percent of the acreage is Tribal reservation land, which is home to nearly 46 percent of the AI/AN population. Tribal health departments in these jurisdictions often rely on public health data collected by the state health department for disease surveillance and planning. However, it is often difficult to extract Tribe-specific information from these databases. Tribal health departments often rely on aggregate AI/AN data that is not reflective of their communities. Therefore, a geocoding methodology was applied to the Arizona Department of Health Services Hospital Discharge Data (AZ HDD) in order to improve disease surveillance and public health planning information for Tribal communities. 

METHODS:  Approximately 3 million AZ inpatient and outpatient discharge records are collected annually and recorded in the AZ HDD. Records contain hospital identifiers, patient demographics, diagnoses, external cause codes, procedures, revenue, and physician information. Utilizing geocoding, U.S. Census defined tribal reservation boundaries and patients’ reported race, we developed a methodology to serve as a proxy for defining the Tribe-specific Hospital Discharge Data in order to enhance precision in estimating incidence of health outcomes and assist in public health planning. Depending on the level of data available by record, geocoding to the reservation boundaries was performed using a physical address, zip code, or town.

RESULTS:  Between 2007 and 2013, 156,949 records had non-missing data that was able to be geocoded to a reservation boundary. Of these, approximately 29% were geocoded using a physical address, 35% were geocoded using the town of residence, and 36% were geocoded using the zip code of residence. 

CONCLUSIONS:  This methodology is an important step forward for Tribes using state data sources. Although Tribal members are still not identified exactly, this methodology does not require additional efforts for data collection and can be retrospectively applied. Therefore, this method has been permanently adopted into the AZ HDD. The inclusion of reservation code to indentify Tribe in AZ HDD is an improvement in disease surveillance and can be used for preliminary public health planning for Tribal health departments.