ObjectiveWe present lessons learned from over a decade of HIV bio-behavioral risk study implementation and capacity-building inAfrican militaries.IntroductionCircumstances within the military environment may place militarypersonnel at increased risk of contracting sexually transmittedinfections (STI) including HIV. HIV bio-behavioral risk studiesprovide a critical source of data to estimate HIV/STI prevalenceand identify risk factors, allowing programs to maximize impact byfocusing on the drivers of the epidemic.MethodsSince 2005, RTI has provided technical assistance (TA) to supportHIV/STI Seroprevalence and Behavioral Epidemiology Risk Surveys(SABERS) in 14 countries across Sub-Saharan Africa and Asia.SABERS are cross-sectional studies consisting of a survey to assessknowledge, attitudes and behaviors related to HIV, coupled with rapidtesting for HIV and other STIs. RTI tailored each survey instrument tobe culturally appropriate in content and methodology, trained militarypersonal to serve as data collection staff, and provided logisticalsupport for study implementation.ResultsKey lessons learned are summarized below:Data collection mode varied from paper-based to computer-assisted surveys, depending on country preference, in-country staffcapabilities, and the country’s technological capacity. Computer-assisted data collection systems were preferable because theyimproved data quality through the use of programmed skip patterns,range, and consistency checks. By eliminating the need for data entry,computer-assisted systems also saved program resources and enabledfaster access to the data for analysis.Survey administration method varied from self-administeredto interviewer-administered surveys. Literacy rates, technologicalfamiliarity, and confidentiality concerns were key drivers indetermining the best data collection method. Self-administeredsurveys such as computer-assisted self-interview (CASI) werepreferable due to the high-level of confidentiality they provide,but required a high-level of literacy and computer familiarity.If confidentiality was a big concern in low-literacy settings, audiocomputer-assisted self-interview (ACASI) was used if the populationhad some computer familiarity. Interviewer-administered surveyssuch as computer-assisted personal interview (CAPI) were used inmost low-literacy settings.Tailoring the survey instrument and administration for culturalappropriateness was vital to the acquisition of sound, viable data.Sexual behaviors and the definition of “regular sexual partner”and other terms varied according to local custom. The sensitivenature of the survey questions also impacted survey administrationoperationally. The preference for same-sex or opposite sexinterviewers varied by country and military setting. It was imperativeto pre-test the survey.A skilled workforce and staff retention are essential to providehigh quality data. Literacy levels, technological familiarity, HIVknowledge, and time commitments must all be considered whenselecting data collection staff. Retention of staff throughout theduration of data collection activities can be a major issue especiallyamong military personnel who were often called away from studyactivities to perform military duties.Host military ownership was integral to the success of the SABERSprogram. By engaging military leadership early and involving themin all decision making processes we ensured the partner military wasinvested in the study and its success and found value in the resultingdata and findings. Host militaries were actively involved in SABERSby providing staff for data collection, leading sensitization activities,and monitoring data collection activities in the field.Inclusion of capacity building elements during studyimplementation led to increased host military buy-in. Capacitybuilding included staff trainings and practical experience in surveymethodology, use of electronic data collection instruments, studylogistics and data monitoring.Confidentiality of survey data and HIV test results was of increasedconcern given that these studies were conducted in a work placeenvironment. For this reason, it was imperative to assure participantsthat disclosures of drug or alcohol use and positive HIV/STI testresults would remain confidential and would not affect their militaryemployment.ConclusionsBased on our experience, the following are required for thesuccessful implementation of an HIV Bio-behavioral Risk Study inresource-poor military settings: (1) selection of a data collection modeand survey administration method that is context-appropriate, (2)utilization of local wording and customs, (3) a skilled workforce, (4)local buy-in/partnership, (5) inclusion of capacity building elements,and (6) assurance of confidentiality.
Going Digital – Computer-Assisted Telephone Interviewing (CATI): Lessons learned from a pilot study
