![[ Visit Client Website ]](images/banner.jpg)
BACKGROUND: The Behavioral Risk Factor Surveillance System (BRFSS) is a nationwide telephone-based health survey. Although BRFSS is effective for assessing risk factors and other health-related indicators at the state level, it has historically underrepresented the on-reservation, rural American Indian population – a group with few landlines and limited cell tower service. The resulting small sample size and potential selection bias limit the generalizability and validity of BRFSS results in tribal populations. Survey response rates depend on the mode of administration; therefore we investigated which mode improves response rates for rural American Indian populations.
METHODS: This study involved a comparative analysis of telephone versus in-person BRFSS survey administration in three rural tribal communities in New Mexico. Participants 18 years and older were randomly selected to participate in the survey. Telephone survey data were collected via standard Computer Assisted Telephone Interviewing software and protocols used by the New Mexico Department of Health. The in-person survey was administered by trained community interviewers. A total of 704 adults participated in the survey (telephone = 166; in-person = 538). Significance testing was performed to assess differences in response rates, participant demographics, and behavioral risk factor prevalence estimates by survey administration type. A cost analysis also was performed to assess differences in expenditures by survey administration mode.
RESULTS: Several notable differences between the two groups were observed. Specifically, in-person survey administration yielded a higher response rate (68.5%) than the telephone survey (35.7%). Likewise, the demographic profile varied by age, income, and education according to survey administration type; participants in the telephone sample were older, wealthier, and more educated than those in the in-person sample. Several risk factor prevalence estimates also differed by type of survey administration. Participants in the telephone survey sample were significantly more likely to report current smoking (28.4% telephone vs. 15.1% in-person, p-value < 0.01), having diabetes (32.9% telephone vs. 21.3% in-person, p-value < 0.01), and obtaining preventive cancer screenings (e.g., mammography rate = 65.2% telephone vs. 46.4% in-person, p-value < 0.01) than those in the in-person group. In-person survey administration cost slightly less than telephone administration ($191.95 per completed survey vs. $210.84 per completed survey, respectively) due to the low response rate of the telephone survey.
CONCLUSIONS: The findings from this study have important implications for public health surveillance in rural tribal communities, where in-person health survey administration is likely to yield greater coverage and reach a more diverse population than telephone survey administration.