Maternal Mortality: What To Measure? Deborah Maine, DrPH AMDD Program, Columbia University Maternal Health Symposium Tokyo, June 2003 ---------- MDG Goal: Improve maternal health Target: Reduce the MM Ratio by 3/4 by 2015 Indicators: ● Maternal mortality ratio ● Proportion of births attended by skilled health personnel ---------- Maternal Health Maternal Mortality ---------- Measuring Progress Impact Indicators (MM Ratios and Rates) or Process Indicators ??? ---------- Impact Indicators: Practical Limitations vital registration: - lack of data - misclassification survey: sample size - cost - confidence intervals ---------- Data Sources (% countries) 1995 UN MMR Estimates Europe n=50 Good Vital Registration: 80 Poor Vital Registration: 14 Sisterhood Estimates: 0 RAMOS: 0 No National Data: 6 Americas n=29 Good Vital Registration: 21 Poor Vital Registration: 21 Sisterhood Estimates: 17 RAMOS: 17 No National Data: 24 Asia & Pacific n=28 Good Vital Registration: 7 Poor Vital Registration: 7 Sisterhood Estimates: 11 RAMOS: 29 No National Data: 46 Middle East & N. Africa n=20 Good Vital Registration: 80 Poor Vital Registration: 14 Sisterhood Estimates: 0 RAMOS: 0 No National Data: 6 Sub-Saharan Africa n=44 Good Vital Registration: 2 Poor Vital Registration: 2 Sisterhood Estimates: 41 RAMOS: 2 No National Data: 52 ---------- % of Maternal Deaths Missing from Official Records USA 1987-1990 (1): 50%* * Pregnancy-related deaths (direct + indirect obstetric) Menoufia, Egypt 1981-83 (2): 68% England & Wales 1982-84 (3): 22% Sources: (1) = Berg et al 1996, (2) = Grubb et al, (3) = Turnball et al 1989 ---------- Comparison of Maternal Mortality Survey Methods Study Conventional Survey: Ethiopia People: 9,315 No. Deaths: 45 MM* Ratio: 457 Years Covered: 2 Sisterhood Survey: Gambia People: 2,163 No. Deaths: 91 MM* Ratio: 1105 Years Covered: 20** Maternal deaths per 100,000 live births Sources: Kwast et al, 1985; Graham, Bass and Snow, 1988; Boerma and Mati, 1989. ** Midpoint in study period = 10 years in past ---------- Maternal mortality ratios using the sisterhood method: Actual estimate*: 767-1243 MMR (dearths/100,000 live births), 1,005 (point estimate) 50% decline: 332-674 MMR, 503 25% decline: 547-961 MMR, 754 95% confidence intervals and point estimates *Data from the Gambia (Graham et al., 1989) ---------- Impact Indicators: Limitations in Interpretation: - The "gold standard"? - Tracking changes - Relevance to action ---------- Process Indicators The crucial point is the Strength of causal chain ---------- The UN Process Indicators UNICEF / WHO / UNFPA, 1997 Measure the availability and utilization of emergency obstetric care (EmOC) ---------- Signal Functions of Basic EmOC 1-3. parenteral antibiotics, oxytocic drugs anitconvulsants 4. manual removal of placenta 5. removal of retained products (e.g. MVA) 6. assisted vaginal delivery ---------- Signal Functions of Comprehensive EmOC Basic EmOC (signal functions 1-6) plus 7. blood transfusion 8. surgery (e.g. C-section) ---------- UN Process Indicators Geographical coverage ● Are there enough functioning EmOC? ● Are they well distributed? For every 500,000 population, there should be: At least 4 Basic EmOC facilities. At least 1 Comprehensive EmOC facility. ---------- UN Process Indicators (cont.) Utilization of services: ● Are enough women using these facilities? At least 15% of pregnant women deliver in and EmOC facility ● Are women with complicationos using these facilities? Met Need: 100% of women estimated to develop complications should receive EmOC ---------- UN Process Indicators (cont.) ● Are enough critical services being provided? Cesarean sections should be at least 5% (and not more than 15%) Of all births in the population (not just in the facility) ---------- UN Process Indicators (cont.) ● Is the quality of the services adequate? Case-fatality rate should be less than 1% (deaths in facility among women treated for obstetric complications) ---------- Benefits of Process Indicators Less expensive: ● don't require population surveys; ● use facility records More valid : ● permit rechecking of data ● permit cross-checking ---------- Benefits of Process Indicators Promote action: ● emphasize functioning ● emphasize coverage More useful: ● can change quickly ● illuminate needs, progress ---------- Met Need for EmOC: 1999-2000 Bhutan*: 19% Rajasthan, India: 9% Morocco*: 33% Mozambique: 8% Nicaragua: 31% Peru: 23% Senegal*: 12% Sri Lanka: 75% ---------- Baseline Data: UN Process Indicators (* = national) Country/Area Senegal * Comp. EmOC: - Basic EmOC: 6% Met Need: 12% Cesar. Sect.: 1.1% Rajasthan Comp. EmOC: 31% Basic EmOC: 37% Met Need: 9% Cesar. Sect.: 1.2% Bhutan * Comp. EmOC: - Basic EmOC: 80% Met Need: 19% Cesar. Sect.: 1.3% Peru Comp. EmOC: - Basic EmOC: 0% Met Need: 23% Cesar. Sect.: 4.7% ---------- Rajasthan, India (UNFPA Project): Availability of EmOC Facilities 2000 CEOC: 17 BEOC: 26 2001 CEOC: 18 BEOC: 36 2002 CEOC: 23 BEOC: 53 ---------- EmOC Facilities in Bhutan 3/00, 9/02 Comprehensive EmOC centers 4 Basic EmOC centers -4 Distribution of Emergency Obstetric Care Centers in the country as of September 2002 ---------- MDG Goal: Improve maternal health Target: Reduce the MM Ratio by 3/4 by 2015 Indicators: ● Maternal mortality ratio ● Proportion of births attended by skilled health personnel ---------- Skilled Attendant at Delivery and MMR, all countries Maternal Mortality Ratio (per 100,000 live births) % Skilled Attendant at Delivery Source: Safe Motherhood Initiative website and Maternal Mortality in 1995: Estimates developed by WHO, UNICEF, UNFPA (2001) ---------- Skilled Attendant at Delivery and MMR, Countries with MMR>400 Maternal Mortality Ratio (per 100,000 live births) % Skilled Attendant at Delivery Source: Safe Motherhood Initiative website and Maternal Mortality in 1995: Estimates developed by WHO, UNICEF, UNFPA (2001) ---------- Matching Indicators with Goals Goal: "Skilled Care" ---------------------------------------------------- MDG Indicator: Skilled Attendants -------------------- UN Process Indicator #1 Coverage of EmOC -------------------------------- ---------------------------------------------------- Normal First Aid Basic EmOC Comprehensive EmOC ---------- Matching Indicators with Goals Goal: "Skilled Care" ---------------------------------------------------- MDG Indicator: Skilled Attendants -------------------------- -------------------------------- UN Process Indicator #1 Coverage of EmOC ---------------------------------------------------- Normal First Aid Basic EmOC Comprehensive EmOC AMDD, Columbia University