Obtaining reliable estimates of the prevalence of induced abortion remains a significant challenge in abortion research. Recently, one indirect, survey-based technique for measuring abortion outcomes, the confidante method, has gained particular attention. The method has been applied in various social and legal contexts; however, its efficacy has not been uniformly established. Increasingly, focus has shifted to assessing the method's key assumptions and quantifying the biases that arise from violations of them. We propose a general statistical framework to conceptualize and quantify the impact of biases on measuring abortion prevalence from such surveys. Specifically, we define the relationship between observed and true abortion prevalence based on misclassification error related to the sensitivity and specificity of the survey instrument. This formulation leads naturally to a Bayesian modeling approach to estimate abortion prevalence, allowing for differing knowledge of and different levels of uncertainty about the misclassification parameters to be incorporated in the modeling process, with that uncertainty being propagated through to the final estimates. We illustrate our framework and modeling approach on data from an application of the confidante method in Uganda in 2018, where we account for systematic differences in confidante abortion reports based on the self-reported abortion experiences of survey respondents.
Pejchinovska et al. (2026) studied this question.