Measuring non-events: infertility estimation using cross-sectional, population-based data from four countries in...

Area of Science:

• Reproductive health research in sub-Saharan Africa.
• Epidemiology of infertility and its determinants.
• Population-based survey methodology.

Background:

• Infertility is a neglected research area in sub-Saharan Africa, with existing estimates varying widely.
• Previous studies have not adequately examined the impact of different study populations and model assumptions on infertility prevalence.
• Accurate estimation of infertility is crucial for public health interventions and service provision.

Purpose of the Study:

• To compare the prevalence of 12-month infertility between women actively trying to conceive and a broader population exposed to unprotected sex in Burkina Faso, Côte d'Ivoire, Kenya, and Uganda.
• To assess how model assumptions and adjustments influence infertility prevalence estimates.
• To provide more accurate data for monitoring infertility trends and informing public health strategies in the region.

Main Methods:

• Utilized cross-sectional, nationally representative survey data from Burkina Faso, Côte d'Ivoire, Kenya, and Uganda.
• Employed a multi-stage cluster random sampling design for women aged 15-49.
• Created two analytic samples: 'tryers' and 'women exposed to unprotected sex', applying the current duration (CD) approach and examining various model assumptions.

Main Results:

• 12-month infertility prevalence among 'tryers' ranged from 7.9% in Burkina Faso to 29.6% in Côte d'Ivoire.
• Infertility prevalence was substantially higher in the broader sample of women exposed to unprotected intercourse, ranging from 22.4% in Uganda to 63.7% in Côte d'Ivoire.
• Sensitivity analyses indicated that estimates were most affected by adjustments for pregnancy recognition timing and sexual activity.

Conclusions:

• Infertility estimates are sensitive to analytic approaches and population definitions, with higher prevalence observed in broader populations.
• Data quality issues, including digit preference and underreported reproductive events, can impact the accuracy of the current duration (CD) approach.
• Improved understanding of methodological inconsistencies is vital for accurate infertility monitoring, prevention programs, and treatment access.