Analytic approach

Ten countries in sub-Saharan Africa were selected for this analysis—five from eastern Africa and five from western Africa. All 10 countries implemented a DHS in each of the following time periods: time 1, 1995–2000; time 2, 2001–2005; time 3, 2006–2011. In the 1995–2000 period, all of the countries had low or very low CPRs, ranging from 3% to 26% of married women of reproductive age. The five eastern African countries—Ethiopia, Madagascar, Malawi, Rwanda, and Zambia—had marked improvements in their CPRs over the 10–15 year time period, with gains ranging from 16 percentage points in Malawi to 39 percentage points in Rwanda. This group of five countries is herein referred to as the CPR-improved countries. The CPRs of the five western African countries—Benin, Ghana, Mali, Nigeria, and Senegal—remained low and, for the most part, stagnant with increases of less than five percentage points over the same time period. This latter group is herein referred to as the CPR-stagnant countries.

In our first analysis, we investigated trends and differentials in access to health workers in relation to use of modern methods of contraception among married women for the two groups of countries. The Health Workers Reach Index (HWRI) (Save the Children 2011) was used as the measure of health worker access. The index incorporates indicators on health worker density and use of health workers as measured by skilled birth attendance (SBA) and DPT3 immunization coverage. Index scores range in value from zero to one, with one being the best score for health worker reach. See sidebar below for a more in-depth description of the HWRI and values of the HWRI for all 10 countries included in this study.

Next, we examined levels and trends in the values of two of the HWRI components (SBA and DPT3) in relation to levels and trends in CPR for the 10 countries. For several reasons, the health worker density ratio was excluded from the in-depth analysis. The DHS collects information on skilled birth attendance and DPT3 coverage, thus we had access to high-quality data on these two variables over a period of time. However, data for the health worker density ratio are not collected in DHS surveys. The health worker density ratio is considered to be of poor validity and reliability in that the data required for its estimation are highly variable with respect to timeliness, comprehensiveness, accuracy, and other dimensions of data quality.

The analysis of trends in SBA and DPT3 for the 10 countries over the 10–15 year period produced somewhat surprising results in that the countries did not fall neatly into the same two CPR groupings (e.g., SBA and DPT3-improved/stagnant countries). Consequently, we disaggregated CPR by geography (urban/rural), method type, and source of method (public/private), and then analyzed how trends and differentials in the disaggregated values varied between the two CPR groupings of countries. Potential health workforce explanations for the findings of these analyses were then developed.

The absence of high-quality data on health workforce variables in eastern and western Africa limits our ability to move beyond exploratory, descriptive analysis into explanatory analysis and to provide subsequent programmatic recommendations. However, we hope that the proposals we generate on potential health workforce underpinnings of differences between the two groupings of countries will encourage additional data collection and analysis on the association between health workforce variables and CPRs.

Next >>