Mapping heterogeneity in family planning indicators in Burkina Faso, Kenya, and Nigeria, 2000-2020

Abstract

Background: Family planning is fundamental to women's reproductive health and is a basic human right. Global targets such as Sustainable Development Goal 3 (specifically, Target 3.7) have been established to promote universal access to sexual and reproductive healthcare services. Country-level estimates of contraceptive use and other family planning indicators are already available and are used for tracking progress towards these goals. However, there is likely heterogeneity in these indicators within countries, and more local estimates can provide crucial additional information about progress towards these goals in specific populations. In this analysis, we develop estimates of six family indicators at a local scale, and use these estimates to describe heterogeneity and spatial-temporal patterns in these indicators in Burkina Faso, Kenya, and Nigeria.

Methods: We used a Bayesian geostatistical modelling framework to analyse geo-located data on contraceptive use and family planning from 61 household surveys in Burkina Faso, Kenya, and Nigeria in order to generate subnational estimates of prevalence and associated uncertainty for six indicators from 2000 to 2020: contraceptive prevalence rate (CPR), modern contraceptive prevalence rate (mCPR), traditional contraceptive prevalence rate (tCPR), unmet need for modern methods of contraception, met need for family planning with modern methods, and intention to use contraception. For each country and indicator, we generated estimates at an approximately 5 × 5-km resolution and at the first and second administrative levels (regions and provinces in Burkina Faso; counties and sub-counties in Kenya; and states and local government areas in Nigeria).

Results: We found substantial variation among locations in Burkina Faso, Kenya, and Nigeria for each of the family planning indicators estimated. For example, estimated CPR in 2020 ranged from 13.2% (95% Uncertainty Interval, 8.0-20.0%) in Oudalan to 38.9% (30.1-48.6%) in Kadiogo among provinces in Burkina Faso; from 0.4% (0.0-1.9%) in Banissa to 76.3% (58.1-89.6%) in Makueni among sub-counties in Kenya; and from 0.9% (0.3-2.0%) in Yunusari to 31.8% (19.9-46.9%) in Somolu among local government areas in Nigeria. There were also considerable differences among locations in each country in the magnitude of change over time for any given indicator; however, in most cases, there was more consistency in the direction of that change: for example, CPR, mCPR, and met need for family planning with modern methods increased nationally in all three countries between 2000 and 2020, and similarly increased in all provinces of Burkina Faso, and in large majorities of sub-counties in Kenya and local government areas in Nigeria.

Conclusions: Despite substantial increases in contraceptive use, too many women still have an unmet need for modern methods of contraception. Moreover, country-level estimates of family planning indicators obscure important differences among locations within the same country. The modelling approach described here enables estimating family planning indicators at a subnational level and could be readily adapted to estimate subnational trends in family planning indicators in other countries. These estimates provide a tool for better understanding local needs and informing continued efforts to ensure universal access to sexual and reproductive healthcare services.

Keywords: Bayesian statistics; Burkina Faso; Contraceptive use; Family planning; Geostatistics; Intention to use contraception; Kenya; Nigeria; Subnational; Unmet need.

Fig. 1

Estimated family planning indicators at the country, region, province, and 5 × 5-km grid-cell level in Burkina Faso, 2020. Bodies of water are coloured in light grey

Fig. 2

Estimated change in family planning indicators at the country, region, province, and 5 × 5-km grid-cell level in Burkina Faso, 2000–2020. Bodies of water are coloured in light grey

Fig. 3

Estimated family planning indicators at the country, county, sub-county, and 5 × 5-km grid-cell level in Kenya, 2020. Bodies of water are coloured in light grey

Fig. 4

Estimated change in family planning indicators at the country, county, sub-county, and 5 × 5-km grid-cell level in Kenya, 2000–2020. Bodies of water are coloured in light grey

Fig. 5

Estimated family planning indicators at the country, state, LGA, and 5 × 5-km grid-cell level in Nigeria, 2020. Bodies of water are coloured in light grey

Fig. 6

Estimated change in family planning indicators at the country, state, LGA, and 5 × 5-km grid-cell level in Nigeria, 2000–2020. Bodies of water are coloured in light grey