Back to Annual Meeting

Back to Annual Meeting

APHA Scientific Session and Event Listing

Michael W. Link, PhD and Indu Ahluwalia, PhD, MPH. NCCDPHP, Centers for Disease Control and Prevention, 4770 Buford Highway, NE, MS: K-66, Atlanta, GA 30341, 770-488-5444, MLink@cdc.gov

To address declining participation rates and eroding coverage of random-digit dialed (RDD) surveys, a pilot study using address-based sampling (based on a U.S. Postal Service database) and mail survey was conducted as part of the 2005 Behavioral Risk Factor Surveillance System (BRFSS). We explored how this alternative sample frame/survey mode design affected responses to selected health indicators: diabetes, high blood pressure, high BMI, alcohol use, HIV testing, and HIV risk behaviors. Mail pilot data (n=3,010) were compared to telephone survey data (n=18,610) in 6 states. Findings show no significant differences in prevalence estimates by survey design for most indicators once adjustments are made for other potential covariates (sex, race, age, education). However, significant differences were observed across the two designs on reporting of high BMI (>30), binge drinking, and engaging in HIV-related risk behaviors. Participants were less likely to report a high BMI (>30) by mail as compared to phone (AOR=0.83; 95% CI=0.71, 0.97). Those who responded by mail were more likely to engage in binge drinking (AOR=1.77; 95%CI=1.43, 2.9) and HIV risk behaviors (AOR=1.74; 95% CI=1.18, 2.56). Participants in cell phone-only households compared to those with traditional land lines were more likely to report binge drinking (AOR=1.90; 95% CI=1.10, 3.28) and HIV risk behaviors (AOR=3.95; 95%CI=1.64, 9.50). Differences across modes did not, however, appear linked to specific population subgroups, but rather varied by item. Additional investigation is required before address-based survey designs are considered effective alternatives to RDD surveys for conducting health surveys of the general public.

Learning Objectives:

The participants will be able to (
• identify and discuss several methods of collecting surveillance data and (
• understand the potential impact of these different survey designs on specific health conditions and risk behaviors.

Keywords: Surveillance, Risk Behavior

Presenting author's disclosure statement:

Any relevant financial relationships? No