Measurement of Fertility Intentions to Improve Person-Centered Reproductive Health Counseling among Kenyan Women

Abstract

Despite growing recognition that fertility intentions exist along a continuum and encompass ambivalence, most measurement approaches rely on binary categorizations that classify pregnancies as "intended" or "unintended." This framework obscures reproductive decision-making complexity and may contribute to high rates of misclassification of fertility intentions. Accurate measurement of fertility intentions is essential for delivering person-centered reproductive health counseling aligned with women's actual desires and needs. This dissertation addresses measurement gaps through three integrated studies examining fertility intention measurement among Kenyan women, including women living with HIV (WLWH).First, I conducted a systematic review, which identified 26 measures of fertility intentions and pregnancy ambivalence. Less than one-third of measures originated from low- and middle-income countries (LMICs), and domains assessed in those measures varied systematically by geography: LMIC measures emphasized cognitive orientations while high-income country measures emphasized emotional responses and pregnancy planning behaviors. Pregnancy ambivalence was operationalized inconsistently across measures, limiting understanding of its prevalence and meaning. These findings challenge assumptions about universal applicability of fertility intention measures and underscore the need for context-specific measure development and validation. Next, I conducted an evaluation of the Desire to Avoid Pregnancy (DAP) scale among 2,504 Kenyan WLWH, providing novel evidence about this measure's performance in sub-Saharan Africa. Unlike the original unidimensional US validation, the scale exhibited a two-factor structure distinguishing “wanting” pregnancy from “worrying” about consequences, with moderate inter-factor correlation suggesting these dimensions can coexist. Both subscales demonstrated excellent reliability. This bidimensional structure can capture pregnancy ambivalence among women who can simultaneously desire pregnancy and worry about its consequences (e.g. HIV transmission), with implications for how clinicians interpret and respond to DAP scores. In the last study, I developed and validated the HerChoice scale by evaluating the scale among 496 Kenyan women. The final scale resulted in a 5-item measure of fertility intentions including three domains: cognitive orientation, emotional responses, and life course consequences. The scale demonstrated excellent reliability and construct validity evidenced by associations with contraceptive use and sexual activity in the last month. Measurement invariance was observed across age, education, parity, and relationship status, indicating the scale functions equivalently across diverse Kenyan women. Together, these studies demonstrate that more nuanced measurement of fertility intentions is both possible and necessary in LMIC settings. Findings provide evidence about measurement gaps and tools that could address them, including a locally derived brief scale (HerChoice) and cross-cultural validation evidence for an existing multi-domain measure (DAP). However, better measurement tools alone are insufficient without implementation. Tools that accurately capture the complexity of fertility intentions can enable person-centered reproductive health counseling, but only when health systems invest in provider training to interpret and respond to measured intentions, establish counseling protocols that move beyond binary contraceptive decisions, and create service delivery models where women's expressed intentions guide care.