Lessons learned for computer assisted interviewing in low-resource settings: Experience from an HIV behavioral surveillance survey in South Sudan

Background: There are many challenges for the use of computer-assisted interviewing methods in rural, low resource settings in sub-Saharan Africa. Lack of computer familiarity, low literacy rates, and the lack of reliable power supply may be barriers to using this reliable survey technology. Sudan People's Liberation Army (SPLA) research staff implemented strategies to overcome these obstacles. The objective of this research was to determine the feasibility of using computer-based data collection and share lessons learned while conducting the first HIV sero-prevalence and behavioral risk survey among members of the SPLA.

Methods: A cross-sectional HIV behavioral surveillance survey of 857 randomly selected soldiers from six SPLA bases was conducted from July to September 2010. Interviews were conducted in English and Juba Arabic using touch-screen Netbooks with either computer assisted personal interview (CAPI) or audio computer assisted self interview (ACASI) methods. Eligibility criteria for ACASI included basic literacy and some computer familiarity. Interviewers gathered feedback on usability from ACASI respondents.

Results: Of the 824 interviews completed (96.1% response rate), 14.4% of participants used ACASI. ACASI participants were more likely English literate than CAPI participants (45.4% vs 7.1%) and semi-urban (Juba) (83.9% vs 66.1%). Overall, ACASI participants expressed a high comfort level and preferred the privacy offered. No data were lost or interviews left incomplete.

Conclusions: As the first HIV surveillance study using CAPI/ACASI in South Sudan, we demonstrated its success for data collection in resource-poor settings with low literacy/low computer-use populations. Strong commitment of SPLA staff and participant screening helped ensure success.