An integrated model of condom use in Sub-Saharan African youth: A meta-analysis

Abstract

Objective: We tested an integrated social–cognitive model derived from multiple theories of the determinants of young people's condom use in Sub-Saharan Africa. The model comprised seven social–cognitive antecedents of condom use: Attitudes, norms, control, risk perceptions, barriers, intentions, and previous condom use. Method: We conducted a systematic search of studies including effects between at least one model construct and intended or actual condom use in young people from sub-Saharan African countries. Fifty-five studies comprising 72 independent data sets were included and subjected to random-effects meta-analysis. Demographic and methodological variables were coded as moderators. Hypotheses of the integrated model were tested using meta-analytic structural equation modeling. Results: The meta-analysis revealed significant nontrivial sample-weighted correlations among most model constructs. Moderator analyses revealed differences in six correlations for studies that included a formative research component relative to studies that did not. There was little evidence of systematic moderation of relations among model constructs by other candidate moderators. Meta-analytic structural equation models revealed significant direct effects of attitudes, norms, and control on condom use intentions, and of intention, control, and barriers on condom use. Including past condom use increased explained variance in condom use intentions and behavior but did not attenuate model effects. There were also significant indirect effects of attitudes, norms, and control on condom use through intentions. Conclusions: Findings provide preliminary evidence to support the integrated condom use model in sub-Saharan African youth. The model provides guidance on potential targets for improving the effectiveness of condom promotion interventions.