Female respondent acceptance of computer-assisted personal interviewing (CAPI) for maternal, newborn and child health coverage surveys in rural Uganda

Abstract

IntroductionHigh maternal and child mortality continues in low- and middle-income countries (LMIC). Measurement of maternal, newborn and child health (MNCH) coverage indicators often involves an expensive, complex, and lengthy household data collection process that is especially difficult in less-resourced settings. Computer-assisted personal interviewing (CAPI) has been proposed as a cost-effective and efficient alternative to traditional paper-and-pencil interviewing (PAPI). However, the literature on respondent-level acceptance of CAPI in LMIC has reported mixed outcomes. This is the first study to prospectively examine female respondent acceptance of CAPI and its influencing factors for MNCH data collection in rural Southwest Uganda. MethodsEighteen women aged 15–49 years were randomly selected from 3 rural villages to participate. Each respondent was administered a Women’s Questionnaire with half of the survey questions asked using PAPI techniques and the other half using CAPI. Following this PAPI/CAPI exposure, semi-structured focus group discussions (FGDs) assessed respondent attitudes towards PAPI versus CAPI. FGD data analysis involved an immersion/crystallization method (thematic narrative analysis). ResultsThe sixteen FGD respondents had a median age of 27 (interquartile range: 24.8, 32.3) years old. The majority (62.5%) had only primary level education. Most respondents (68.8%) owned or regularly used a mobile phone or computer. Few respondents (31.3%) had previously seen but not used a tablet computer. Overall, FGDs revealed CAPI acceptance and the factors influencing CAPI acceptability were ‘familiarity’, ‘data confidentiality and security’, ‘data accuracy’, and ‘modernization and development’. DiscussionFemale survey respondents in our rural Southwest Ugandan setting found CAPI to be acceptable. Global health planners and implementers considering CAPI for health coverage survey data collection should accommodate influencing factors during survey planning in order to maximize and facilitate acceptance and support by local stakeholders and community participants. Further research is needed to generate best practices for CAPI implementation and LMIC; higher quality, timely, streamlined and budget-friendly collection of MNCH indicators could help direct and improve programming to save lives of mothers and children.