The 131st Annual Meeting (November 15-19, 2003) of APHA

Abstract #62173

Diane M. Makuc, DrPH¹, Nathaniel Schenker, PhD¹, Trivellore E. Raghunathan, PhD², and Pei-Lu Chiu¹. (1) National Center for Health Statistics, 3311 Toledo Rd., Room 6208, Hyattsville, MD 20782, 301-458-4360, DMakuc@cdc.gov, (2) Biostatistics, University of Michigan, M4218 Sph II, Ann Arbor, MI 48109

The National Health Interview Survey (NHIS), a household survey of the U.S. civilian noninstitutionalized population, is a major data source for studies of health status and health care access and utilization. Although item nonresponse in the NHIS is generally low, annual family income, a key variable in many analyses, has a high level of nonresponse. In 2000, 28% of families in the NHIS did not report family income and about 10% reported family income that was inconsistent with reported family earnings. Starting with 1997 data, multiple imputation of unknown or inconsistent family income and unknown personal earnings is being carried out. Multiple imputation allows the variability due to imputation to be taken into account in analyses. Five independent sets of imputed values have been derived using sequential regression multivariate imputation, a procedure developed by Raghunathan et al. (2001, Survey Methodology). The application of this procedure to the NHIS is discussed in documentation for public use files containing the 1997-2001 imputed income values (Schenker et al., forthcoming). This presentation will describe the missing income data patterns and the multiple imputation of income in the 1997-2001 NHIS. Use of the multiply imputed data will be illustrated and analyses of access to health care that are based on the multiply imputed data will be compared with analyses based on singly imputed data and analyses that delete respondents who did not report income. Differences in point estimates and their standard errors will be presented.

Learning Objectives:

• At the conclusion of the session, the participants in this session will be able to:
• 1. Describe missing income data patterns in the National Health Interview Survey.
• 2. Describe the multiple imputation procedure used for missing income in the 1997-2001 National Health Interview Survey.
• 3. Describe the differences between estimates obtained using multiple imputation, single imputation, and deletion of those who did not report income.

Keywords: Survey, Methodology

Presenting author's disclosure statement:
I do not have any significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.