Inequities in effective coverage of family planning services in low- and middle-income countries: linking household and facility surveys.

Background: Despite decades of family planning (FP) programme successes in low- and middle-income countries (LMICs), women still face an unmet need for contraceptives, as well as inequalities in coverage. Including elements of service readiness in FP intervention coverage measures will better inform population-level programme performance.

Methods: We identified five LMICs that had health facility and household surveys conducted <5 years apart within the past ten years: Bangladesh, Haiti, Malawi, Nepal, and Tanzania; only Nepal had time-trend data available. We developed quality readiness-adjusted FP coverage measures by linking health facility assessments (i.e. readiness) and household surveys (i.e. intervention coverage) according to ecological linking methods. We defined linking units by facility type, managing authority, and geographic location using women's reported source of contraceptives. We defined intervention coverage as the percentage of women aged 15-49 years who needed contraceptive services and were using a modern method, and calculated readiness-adjusted intervention coverage with the average FP readiness score in each linking unit. We used a coverage cascade model to understand gaps in health service readiness and access, and performed a health equity analysis for wealth, locality, and age.

Results: Large gaps in FP intervention coverage and readiness were present in all settings. Facility readiness scores ranged from 0.58 to 0.66, with gaps in coverage and readiness-adjusted coverage ranging from 49 percentage points (pp) in Bangladesh to 21 pp in Haiti. Urban, wealthier, and adolescent women had lower readiness-adjusted coverage because they also obtained their contraceptives outside of health facilities. Coverage cascades changed little for Nepal between 2015 and 2021.

Conclusions: By calculating readiness-adjusted FP coverage using a cascade model in five geographically diverse countries, including time trends in one country, we found large gaps in intervention and readiness-adjusted coverage with related inequalities.