Monitoring AIDS and other rare population events: a network approach

Area of Science:

• Epidemiology
• Sociology
• Public Health

Background:

• Tracking the distribution of Acquired Immunodeficiency Syndrome (AIDS) is crucial for public health interventions.
• The General Social Survey (GSS) offers a unique dataset by asking respondents if they know a person with AIDS (PWA).
• Previous attempts using GSS data to track AIDS demographics showed potential biases.

Purpose of the Study:

• To replicate and extend earlier findings on using GSS data to track AIDS distribution.
• To compare the demographic composition of PWAs reported in the GSS with official Centers for Disease Control (CDC) data.
• To assess the accuracy of the GSS network technique by comparing it with data on known homicide victims.

Main Methods:

• Analysis of GSS data from 1988-1991, comparing reported PWA demographics (gender, race, age, region) with CDC data.
• A parallel analysis using GSS respondents' knowledge of homicide victims to validate the network method.
• Utilizing National Health Interview Survey time series data to corroborate GSS findings on knowing at least one PWA.

Main Results:

• The GSS data consistently identified proportionately more White and Midwestern cases of AIDS compared to official CDC reports.
• The validation analysis using homicide victim data corroborated the earlier findings regarding GSS biases.
• Time series data from the National Health Interview Survey supported the GSS findings.

Conclusions:

• The GSS network technique, while valuable, exhibits biases in representing the demographic distribution of AIDS cases.
• Discrepancies suggest potential underrepresentation of minority and non-Midwestern populations in GSS-identified PWA networks.
• Suggestions are provided for refining the GSS network approach to improve its accuracy and utility in epidemiological surveillance.