Page 1 of 1

Table IV.1. Selected Firm Experience with Geocoding and Surname Analysis During Part of the Collaborative

Evaluation of a Learning Collaborative's Process and Effectiveness to Reduce Health Care Disparities Among Minority Populations

FirmaTarget PopulationUse of DataValue of Exercise
Firm AMembers served by providers in their system that do not report race and ethnic data.The geocoded data rate of African Americans was far below that in reported data. The results were therefore disappointing and not used.The limited usefulness of geocoding reinforced the firm's previous commitment to begin collecting primary race/ethnic data for all members.
Firm BManaged care members with diabetes and later other conditions (about 25 percent of membership).Developed HEDIS indicators for 2003 and 2004. Efforts will be continued using 2005 HEDIS data with firm's purchased software.Confirmed that existing approaches were similar to those of other firms. Results helped to identify and target specific areas of disparities.
Firm CDiabetes and AMI members for a subset of health plans.Results were analyzed and presented to firm leadership.Staff say findings reinforced the value of work in this area among some key staff.
Firm D

Adult commercial members in the CAHPS sample frame for their largest state (n=450,000).

Selected Medicare and Medicaid members.

Data used for analysis of a diverse range of HEDIS indicators over time. Firm has 2004 and 2005 data and plans to analyze 2006 HEDIS indicators.Results led to an intervention aimed at improving eye exams for Hispanic diabetic members, focusing on practices identified in geocoding and surname analysis.
Firm ESample of members with chronic disease across regions (n=335,000).Examined HEDIS indicators on diabetes management for four racial/ethnic groups. Focused especially on differences in outcomes in the region where firm is developing an intervention. Firm is continuing to pursue geocoding and may, in one region, expand it to other providers.Results were to be used in identifying disparities until more self-identified race/ethnic data are captured via the electronic health record. Results also were to provide a baseline for pilot interventions.
Firm F16,000 diabetic members of disease management programs.Analyzed results to distinguish socioeconomic from racial influences on process and outcome measures related to diabetes and presented these to management.Results focused the firm and put disparities on its radar screen. Led to a number of interventions, as well as collection of member reported data on race and ethnicity.
Firm GAll members known to be eligible for disease management programs (2 million); more intensive analysis in a single geographical area.Results were not sufficiently targeted to support the intended member-specific outreach. Data are being reanalyzed to identify "hot spots" for geographically based pilots.Staff formed an ad hoc working group to coordinate analyses and use them in structuring interventions.

 a. Firms are presented in random order.

Return to Document

Page last reviewed December 2007
Internet Citation: Table IV.1. Selected Firm Experience with Geocoding and Surname Analysis During Part of the Collaborative: Evaluation of a Learning Collaborative's Process and Effectiveness to Reduce Health Care Disparities Among Minority Populations. December 2007. Agency for Healthcare Research and Quality, Rockville, MD. http://www.ahrq.gov/research/findings/final-reports/learning/tab4-1.html