issue that impacts the community’s health outcomes — Maternal mortality in Kenya
Search snapshot / sources used
I searched recent peer-reviewed papers and national survey analyses on maternal health and facility maternal mortality in Kenya. Key recent sources I used include:
-
Kenya/DHS and national maternal health analyses.
-
Sub-national analyses of facility maternal mortality and modeling studies for Kenya.
-
Studies on system barriers and quality of maternal care in Kenya.
Suggested population sample (for a local study on maternal outcomes / maternal care quality)
If you were to study maternal outcomes (or satisfaction/experience with maternal services) in your community / county the population and sampling design might be:
Population: Women of reproductive age (15–49 years) who delivered in the last 12 months and who used health facilities in the selected county/organization (facility births and recent postpartum women).
Sampling design (recommended for local/sub-national work): Multi-stage, stratified cluster sampling
-
Stage 1 (clusters): Select health facilities in the county stratified by level (primary health center, sub-county hospital, county referral hospital). This accounts for differences in resources and case-mix.
-
Stage 2 (within-facility): From facility delivery registers (or postnatal clinic lists), take a stratified random sample of postpartum women — stratify by delivery mode (vaginal/CS) or by month of delivery to ensure recency.
-
Sample size: For estimating a proportion (e.g., % reporting good quality postnatal care) with ±5% margin at 95% confidence, a typical target is ~350–450 participants after design-effect adjustment (cluster sampling increases variance; apply a design effect of 1.5–2). Use standard sample-size formulae to get precise numbers. (See sampling methods literature for formulas and design-effect guidance.) APA
Rationale: multi-stage + stratified sampling captures facility-level differences and produces estimates that are generalizable across the county while being operationally feasible (you sample selected facilities, then sample patients there). This approach matches methods used in recent Kenyan facility-based maternal mortality and quality studies.