![[ Visit Client Website ]](images/banner.png)
BACKGROUND: The Institute of Epidemiology, Disease Control and Research (IEDCR) is the national institute for disease surveillance and outbreak investigation in Bangladesh. Cell phone coverage has increased significantly in Bangladesh in recent years, which provides a unique opportunity to conduct surveys for collection of population-based health data. A consultation project was sponsored by CSTE to assist the IEDCR with designing a cell phone based BRFSS survey.
METHODS: The first trip was made to Bangladesh in late 2011 for in-person meetings with IEDCR staff and local cell phone companies. A survey protocol with a survey questionnaire was designed for a pilot BRFSS survey in Dhaka City. The English questionnaire was translated to Bengali. The Dhaka survey was implemented by IEDCR in-house interviewers in 2012. The survey randomly selected adult Dhaka residents, and collected data on major local health issues (diarrhea, injury, diabetes, and hypertension) and related risk behaviors. A second trip was made to Bangladesh in 2013 to review the results of the 2012 survey, and to plan for a nationwide BRFSS survey in 2014. The survey methodology and questionnaire were updated based on the 2012 survey results. A national survey sample was designed based on the needs of national health programs and resources available.
RESULTS: The 2012 pilot survey had 42 questions on health status, health care access, physical activity, diet, tobacco use, individual demographics, family information, and children’s health conditions. The survey had 3,388 completed and 1,366 partially completed interviews of adults who had lived in Dhaka City for one year or longer. The American Association for Public Opinion Research (AAPOR) response rate was 31.5%, and the cooperation rate was 87.8%. The 2014 national survey was designed with 15 strata in Dhaka City and urban and rural areas in seven health districts nationwide. New question modules were developed, samples selected from six major cell phone providers in the country, and a random selection of one respondent within each household was added to the 2014 survey protocol.
CONCLUSIONS: This study demonstrates: 1) increased cell phone coverage makes BRFSS surveys a feasible and cost-effective means to collect population-based health data in developing countries; 2) the survey protocol and questionnaire need to be tailored to match specific local conditions and to meet local health care needs; 3) a pilot study to test the methodology is needed for identifying and correcting unexpected problems before implementing large scale surveys in developing counties.