Unmet need for contraception (UNC) quantifies the proportion of exposed, fecund women of reproductive age (15–49 years) who want to limit or postpone childbearing but are not using contraception. UNC is central to family planning policy, program evaluation, and population-health discussion because it signals potential unfulfilled demand for contraception and correlates with fertility outcomes and women’s autonomy.
The currently accepted standard for estimating UNC in DHS-derived data is the Revised algorithm, which classifies women into mutually exclusive categories: met need, unmet need, no unmet need, and infecund/menopausal. Despite improvements in comparability and simplification of prior algorithms, measurement challenges remain, including behavioral inconsistencies, exposure misclassification, and conceptual complexity.
Existing literature and methodological critiques raise the possibility that some women labeled as having unmet need for contraception may, in fact, have no unmet need because the risk of conception is reduced by partner characteristics. The authors argue that a subset of exposed women may consciously decline contraception because their husbands experience diminished fecundity or virility, leading these women to perceive a low or negligible pregnancy risk.
Incorporating husband-related fecundity constraints into the measurement model aims to reduce potential overestimation of unmet need and produce estimates that better align with true conception risk.
The study defines the husbands’ effect as diminished male fecundity or loss of virility arising from multiple domains:
These male factors can meaningfully reduce conception probability despite ongoing sexual activity, and when reported by wives as reasons for nonuse, they may justify treating those women as having no unmet need.
To capture the husbands’ effect, the authors generated a modified Revised algorithm by adding a constraint to the standard Revised classification. This new constraint reclassifies exposed, fecund women who report husband-related diminished virility as having no unmet need rather than unmet need.
The intention is to identify women who legitimately face low pregnancy risk due to partner factors and therefore should not be counted in the pool of women with unmet contraceptive needs.
The modified algorithm was applied to the 2014 Kenya Demographic and Health Survey (KDHS 2014). From the KDHS data the authors selected 8,710 reproductive-age women who met the fertility-related inclusion criteria for UNC estimation. Analysis was performed using Stata.
Data provenance: The KDHS data are third-party DHS Program data available on request; the authors state that other researchers can access the same dataset following DHS access procedures.
Applying the modified Revised algorithm that incorporates husbands’ virility produced an estimated unmet need for contraception of 16.4 percent in the analytic sample. This estimate is 1.1 percentage points lower than the national KDHS estimate, which the authors interpret as approximately a 6.6 percent overestimation in the standard estimate of unmet need.
The modification therefore reclassified a subset of women previously counted as having unmet need into the no unmet need category based on partner-related fecundity considerations.
The modified estimates showed subgroup variation:
These subgroup patterns mirror known sociodemographic gradients in contraceptive need and access, while indicating that partner-related reclassification affects groups differently.
The study demonstrates that including an explicit husbands’ virility constraint in the Revised algorithm yields lower estimates of unmet need and reclassifies some women from unmet need to no unmet need. The authors argue that recognizing partner-related diminished fecundity can improve UNC as an estimator of total fertility and help prioritize resource allocation for family planning.
By identifying cases where nonuse is driven by perceived low conception risk owing to male factors, program planners may refine unmet need metrics and avoid overestimating service gaps.
The analysis used KDHS 2014 data obtained from the DHS Program; interested researchers may request access through the DHS data access process. The authors report no specific funding for this work and declare no competing interests.
Note: The original article provides methodological context and historical evolution of unmet need measurement and describes the rationale, implementation, and results of the modified Revised algorithm. Details on exact algorithmic criteria, statistical model specifications, and variable operationalization beyond what is summarized here were reported in the source article.