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

Abstract #67163

Kushang V. Patel, MPH, Dept. of Preventive Medicine and Community Health, University of Texas Medical Branch, 301 University Boulevard, Galveston, TX 77555-1153, 409/772-8658, kvpatel@utmb.edu, Karl Eschbach, PhD, Center for Immigration Research, University of Houston, 468 PGH 3012, Houston, TX 77204-3012, Laura A. Ray, MPA, Division of Sociomedical Sciences, University of Texas Medical Branch, 301 University Blvd., Galveston, TX 77555-1153, and Kyriakos S. Markides, PhD, Dept. of Preventive Medicine and Community Health/ Division of Sociomedical Sciences, University of Texas Medical Branch, 700 Harborside Dr, Ewing Hall 1.128, Galveston, TX 77555-1153.

Background: Hispanics have been reported to have lower mortality compared to non-Hispanic Whites despite their socioeconomic disadvantage. A chief source of information about Hispanic mortality has been population surveys that are linked to the National Death Index (NDI). Purpose: To evaluate the quality of NDI matches to survey data for an elderly Hispanic sample through comparison to an external source of mortality information. We 1) compare mortality from an NDI search for a cohort of 3,050 older Mexican Americans (1993-2000 Hispanic EPESE) to vital status reports by proxy informants; and 2) apply correction factors to age 65+ mortality estimates from 1986-97 National Health Interview Survey (NHIS)-NDI data (N=66,667). Methods: Survival analyses by age, sex, and nativity were performed on both NDI-linked datasets. Correction factors based on the ratio of proxy to NDI-identified Hispanic EPESE deaths were applied to NHIS-NDI data. Results: Age distributions of NDI deaths were similar in the Hispanic EPESE and NHIS cohorts (Mantel-Haenszel c²=0.05; p=0.832). NDI and proxies were discordant on 10.4% of vital status classifications in the Hispanic EPESE (k = 0.73). Mortality underestimation by NDI was largest for women and the foreign-born. NHIS-NDI age-adjusted mortality estimates in elderly women over a 6.5-year period were 18.9% in Mexican Americans and 21.1% in non-Hispanic Whites; mortality was 29.0% in Mexican Americans once correction factors for NDI ascertainment biases were applied. Similar patterns were observed in men. Conclusions: Mortality estimates based on NDI information are likely understated for older Mexican Americans, particularly among women and the foreign born.

Learning Objectives:

At the end of the presentation participants will be able to:
• describe mortality patterns for older Mexican Americans;
• identify special concerns for vital statistics in Hispanics; and
• Evaluate implications of under-ascertainment of deaths for older Mexican Americans when comparing Hispanic to non-Hispanic White mortality rates.

Keywords: Hispanic, Mortality

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.