Background
Comprehensive Emergency Obstetric Care
(
CEmOC
)
| • Caesarean section • Blood transfusion |
Basic Emergency Obstetric Care
(
BEmOC
)
| • IV/IM antibiotics • IV/IM uterotonic drugs/oxytoxics • IV/IM anticonvulsants for pre-eclampsia and eclampsia (ie. magnesium sulfate) • Manual removal of placenta • Assisted vaginal delivery (episiotomy, instrumental delivery (forceps or vacuum extraction), advanced skills for manual delivery of shoulder dystocia, breech) • Removal of retained products (manual vacuum extraction, dilation and curettage) * Assuming no access to Caesarean section or blood transfusion |
Skilled childbirth care
| Skilled birth attendant defined by WHO, ICM, and FIGO as “an accredited health professional – such as a midwife, doctor or nurse – who has been educated and trained to proficiency in the skills needed to manage normal (uncomplicated) pregnancies, childbirth and the immediate postnatal period, and in the identification, management and referral of complications in women and newborns.” [13] The core intrapartum skills that should be provided include: • Clean delivery care • Monitoring onset and progress of labor with partograph • Monitoring maternal and fetal well-being during labor, identify maternal/fetal distress and taking appropriate action including referral • Manage normal vaginal delivery (including releasing a cord around the neck, delivery of shoulders, assisting a breech delivery) • Active management of third stage of labor • First line management of hemorrhage and hypertension in labor, referral as needed • Pain relief, hydration * For the purposes of this estimate assuming no access to instrumental delivery (forceps or vacuum extraction), Caesarean section or blood transfusion |
Trained Traditional Birth Attendant
| Traditional birth attendant defined by WHO as “a person who assists the mother during childbirth and who initially acquired her skilled by delivering babies herself or through an apprenticeship to other TBAs”[15]. A “trained TBA” is “any TBA who has received a short course of training through the modern health sector to upgrade her skills” [61]. TBAs may range from family members attending only occasional births to women with considerable expertise attending 20+ births/year. TBAs are not usually salaried, and typically not civil servants or employed by Ministry of Health. |
Timing of intervention and effect:
These packages include care provided during labor and birth, but in order to be effective, the care may have been initiated during the antenatal period (e.g., screening for abnormal lie and decision for elective Caesarean section, or screening and management of hypertensive disease of pregnancy/eclampsia). Some interventions are primarily intrapartum in timing such as management of acute intrapartum events including antepartum hemorrhage, cord prolapse and obstructed labor.
Not included in these effect estimates:
The effects on neonatal survival of specific interventions after birth for the baby are not included here as they are treated as single additional interventions in LiST and have been considered in detail in other reviews: - Stimulation and neonatal resuscitation at birth, - Postnatal healthy practices (breastfeeding, hygienic cord and skin care, thermal care). In addition, a few specific obstetric interventions which are in LiST but affect other neonatal causes of death have been considered in detail in other reviews including the following: - Corticosteroids for preterm labor (affects preterm deaths), - Antibiotics for preterm premature rupture of membranes (affects deaths from infections). |
Objective
Methods
Searches for intervention evidence
Inclusion/exclusion criteria
Intervention definitions and those not considered in this review
Comparison group
Outcome definitions
Ecologic analysis of variation in neonatal encephalopathy incidence
Delphi Process for establishing expert consensus
Analyses and summary measures
Results
Results of literature review
Emergency obstetric care
Author | Study Years | Setting | Study Design | Intervention definition | Concurrent interventions | Intervention Coverage | Total Births A) Endline B) Baseline | Outcomes | Effect on outcome RR/OR (95% CI) |
---|---|---|---|---|---|---|---|---|---|
Ronsmans 2010[37] | 1987-2005 | Matlab, Bangladesh | Observational cross-sectional | 1987-1996: skilled home birth care w/midwives providing antenatal care, basic obstetric care (labor monitoring), essential newborn care; 1996 onwards facility based birth with BEmOC (partograph, active management 3rd stage, antibiotics, management preeclampsia). Highest level care received (BEmOC, CEmOC, vs no skilled care) | Antepartum care, Essential newborn care, Strengthening of referral and transport systems | CEmOC 0.5% in 1987 to 11.7% in 2005 BEmOC 4.7% in 1987 to 40.9% in 2005 | CEmOC 3084; BEmOC 9954; No skilled Care 40177 | 1) ENMR 2) Stillbirth | 1)CEmOC aOR 2.69 (2.16-3.37) BEmOC aOR 1.47 (1.27-3.37) 2) CEmOC aOR 6.61(5.62-7.79) BEmOC aOR 1.51(1.31-1.73) |
Berglund 2010[44] | 2003-2004 | 3 Maternity Hospitals; Ukraine | Observational before-after | Training all maternity staff (obstetricians, neonataologists, midwives, anesthesiologists) in 2 week WHO "Effective Perinatal Care" program, including use of partogram, emergency obstetric and neonatal care (resuscitation). | Anesthesia; neonatal resuscitation & special care, thermoregulation | All maternity staff in 3 hospitals | A) 1696 B) 2439 | 1) ENMR | No significant effect |
Hounton | 2001-2005 | Rural Ouargaye and Diapaga districts, Burkina Faso | Quasi-experimental | Upgrading of hospital, health centers in intervention area. Mid-level, referral facilities: emergency obstetric care training. First-level centers: training in prevention of complications and early detection -referral for emergencies. Quality improvement infrastructure upgrading, equipment and supplies | National policies and guidelines; Mobilising/educating communities to plan for and use maternal health services | Training in 1 district hospital and 13/19 health centers | 18,658 births intervention district 2004-5; 21,788 births comparison district 2004-5 | 1) PMR | 1) OR 0.75(0.70-0.80) |
Draycott 2006 [41] | 1998-2003 | South Mead Hospital, UK | Before-after | EOC training course: CTG interpretation, course of action, obstetric emergency drills (dystocia, PPH, eclampsia, twins, breech, resuscitation) | Mandatory course for all midwives | A) 11030 B) 8430 | 1) HIE (MacLennan): | 1) RR 0.50(0.26-0.95) | |
Edmond 2002[42] | 1995-1998 | Natal, Northeast Brazil | Observational before-after | Opening of primary maternity facilities at polyclinic to serve low risk deliveries in the community. Pre-booking of deliveries of high risk pregnancies at Maternity hospital with CEmOC capacity. | ANC, community health agents training in community health clinics | Deliveries at maternity clinics increased from 0% to 51% | A) 536 B) 679 | 1) ENMR 2) Stillbirth 3) PMR | 1) RR 0.12 (0.04-0.40) 2) RR 0.66 (0.47-0.94) 3) RR 0.52 (0.37-0.73) |
McCord 2001[43] | 1996-1999 | Rural Maharashtra, India | Cross-sectional | Comparison of perinatal mortality among births occurring at home vs. in hospital, some with CEmOC | 85% home births, 15% in hospital. | Home: 2436 Hospital: 425 | 1) PMR | PMR 27.1 (home births) vs 87 (hospital deliveries) | |
Koblinsky 1999[40] | 1957-1990s | Malaysia | Historical-ecological | 1960 s Training of professional village midwives, linking to regional clinics, referral to district hospitals; 1980's shift to facility births with BEmOC | 3 decades of perinatal care and obstetric care upgrading | 95% of births by midwives (1996); 80% of risk deliveries in hospital (1998) | NS | 1) NMR | NMR from 75.5 (1957) to 14.8 (1991) |
Korhonen 1994[45] | 1986-1991 | Helsinki, Finland | Cross-sectional | Emergency Caesarean Team in Hospital vs. On call (out of hospital, 10 minute average delay) | NS | 60 in hospital; 41 on call | 1) Fetal Death; 2) HIE | 3 in utero fetal deaths and 1 HIE in control (on-call) group vs 0 hospital | |
Piekkala 1985[1] | 1968-1982 | University Hospital, Turku Finland | Historical | 15 year improvement in obstetric management: Cesearean rate increase from 4-12%; vaginal breech delivery from 4 to 1%; implementation of antepartum CTG (monitoring increase from 0 to 90%) | Corticosteroids, Neonatal intensive care, respiratory therapy, fluid-nutritional therapy | Referral hospital for 10% of population | A) 5,410 B) 5,996 | 1) PMR 2) Intrapartum mortality | 1) RR 0.39 2) RR 0.29 |
Skilled childbirth care
Author | Study Years | Country | Setting | Study Design | Primary Intervention | Concurrent Interventions | Intervention Coverage | Total N A) Intervention B) Comparison | Outcomes Measured | Effect on outcome (95% CI) |
---|---|---|---|---|---|---|---|---|---|---|
Ronsmans 2008[50] | 1975-1999 | Matlab, Bangladesh | Rural, 1987-1996 SBA at home | Quasi-experimental (†use of before-after data in pooled anlaysis) | Posting of midwives in villages to increase skilled home birth (antenatal, basic obstetric, care including labor monitoring, essential newborn care) until 1996. After 1996, facility based strategy with upgrading of health centers in basic obstetric care (partograph use, active management 3rd stage, antibiotics, magnesium) | Strengthening referral systems, Transport to BEMOC or CEmOC | 25% of births attended by SBA during home birth period | A) 19085 (ICDDR,B 1989-1995) B) 22821 (ICDDR,B 1982-1988) | 1) IPR-NMR 2) NMR † 3) ENMR† 4) PMR† | 1) 0.78 (NS) 2) 0.83 (0.76-0.91) 3) 0.89 (0.80-0.97) 4) 0.92 (0.84-0.98) |
Yan 1989[48] | 1983-1986 | Shunyi, China | Rural Shunyi County, 7 of 29 townships | Before-after | Village doctors-midwives identify risk and either manage (external cephalic version, blood pressure monitoring) or refer mothers to county hospital | Improvement of neonatal ward in county hospital | 96% of pregnant women seen by village doctor-midwife | A) 2335 B) 2212 | 1) PMR 2) EMR 3) IP-PMR | 1) 0.66 (0.44-0.98) 2) 0.77 (0.43-1.36) 3) 0.73 (*) |
Ibrahim 1992[49] | 1985-1988 | Khartoum, Sudan | Rural, 91% home delivery | Before-after | Training and upgrading of skills of village midwives (antenatal care, monitoring in labor) | Data collection maternal-perinatal outcomes, referral system to hospital | 91% of births delivered by village midwives | A) 2298 B) 3977 | 1) NMR 2) ENMR 3) SBR | 1) 0.68 (0.48-0.97) 2) 0.78 (0.61-1.01) 3) 0.85 (0.60-1.19) |
Alisjahbana 1995[51] | 1992-1993 | West Java, Indonesia | Rural villages, West Java; Tanjungsari district | Quasi-experimental (use of before-after data in pooled analysis) | Training physicians and village midwives on danger signs, case management in pregnancy, labor, delivery, postpartum; development of birthing homes | Training TBAs in pregnancy detection, complications and referral; communications and transportation | 92% of births with professional provider | A) 1176 B) 1099 | 1) PMR | 0.75 (0.51-1.10) |
Author | Study Years | Country | Setting | Study Design | Primary Intervention | Concurrent Interventions | Intervention Coverage | Total N A) Intervention B) Comparison | Outcome Measured | Effect on outcome RR/OR (95% CI) |
---|---|---|---|---|---|---|---|---|---|---|
Matthews 2004[59] | 1999-2002 | Ghana | Rural Brong Ahafo district | Before-after | Training midwives in health facilities on use of partograph and emergency obstetric skills | TBA Training in danger signs, Emergency obstetric transport service | NS | A) 768 B) 575 | 1) PMR | NS |
Andersson 2000[55] | 1831-1899 | Sweden | 18 Parishes Northern Sweden | Historical | 1829 Training of midwives in use of forceps, "sharp hooks and perforators" | 1881 antiseptic techniques | 73% of home deliveries attended by midwives at endline (43% baseline) | NS | 1) PMR | 1) 0.71(0.62-0.82) |
Hatt 2009[56] | 1986-2002 | Indonesia | National DHS Data | Historical | Village midwife training program started in 1989, by 1995 50,000 trained. In 1996 competency based training, neonatal resuscitation | 2 decades of national perinatal care and obstetric care upgrading | Proportion of deliveries attended by midwives increased from 12% (1986) to 30% (2002) | NS | 1) ENMR 2) First day mortality | 1) 0.97 (0.95-0.99) per year reduction 2) 0.98(0.95-1.02) per year reduction |
Koblinsky 1999[40] | 1957-1990s | Malaysia | National NMR | Historical-ecological | 1960 s Training of professional village midwives, linking to regional clinics, referral to district hospitals; 1980's shift to facility births | 3 decades of perinatal care and obstetric care upgrading | By 1986, 95% of home births by midwives; by 1995, 88% institutional delivery; 90% of women with high risk, 80% moderate risk delivering in hospitals | NS | 1) NMR | NMR from 75.5 (1957) to 14.8 (1991) |
PATH 2006[58] | 2003-2004 | Cirebon, Indonesia | Rural Cirebon district, west Java, pop 2 mill | Before-After | Training mid-wives in management of labor, birth asphyxia, tube-mask resuscitation, refresher training/supervision | 60% of asphyxia cases managed by midwives. Uncertain coverage | Est 44000 | 1) IPR-NMR 2) NMR 3) SBR | 1) 0.39 (0.31- 0.48) 2) 0.60 (0.53-0.68) 3) 0.39 (0.31-0.48) | |
Shankar 2008[57] | 1989-2003 | Indonesia | National NMR | Historical | Village midwife training program started in 1989, by 1995 50,000 trained. In 1996 competency based training program including neonatal resuscitation | 2 decades of national perinatal care and obstetric care upgrading | In rural areas skilled attendance increased from 22% to 55% | NS | 1) NMR | NMR decreased from 32 to 20/1000 over 14 years |
Study Quality | Summary of Findings | |||||||||
---|---|---|---|---|---|---|---|---|---|---|
Directness | Endline | Baseline | ||||||||
No of studies | Design | Limitations | Consistency | Generalizability to Population of Interest | Generalizability to intervention of interest | Events | Births | Events | Births | Relative Risk (95% CI) |
Neonatal Mortality
(
Intrapartum-related
)
: Low outcome specific quality
| ||||||||||
1 [50] | Quasi-experimental | Several interventions simultaneously and changes also in comparison villages | Community-setting LIC-MIC, South Asia | Yes | NS | 19,085 | NS | 22,413 | 0.78 (0.64-0.95) | |
Neonatal Mortality
(
All Cause
)
: Low outcome specific quality
| ||||||||||
Observational, before-after | Low quality, before-after comparisons | No evidence of heterogeneity (p=0.28) | Community-setting LIC-MIC | Yes | 794 | 21383 | 1186 | 26798 | 0.82 (0.75-0.90)a | |
Early Neonatal Mortality
(
All Cause
)
: Low outcome specific quality
| ||||||||||
Observational, before-after | Low quality, before-after comparisons | No evidence of heterogeneity (p=0.50) | Community-setting LIC-MIC | Yes | 597 | 23718 | 837 | 29010 | 0.87 (0.79-0.97)a | |
Perinatal Mortality
(
All Cause
)
: Low outcome specific quality
| ||||||||||
Observational, before-after | Low quality, before-after comparisons | Evidence of heterogeneity (p=0.12) | Community-setting LIC-MIC | Yes | 670 | 21981 | 909 | 27621 | 0.88 (0.83-.95)b |
Traditional birth attendant training
Author | Study years | Setting | Study Design | Intervention definition | Concurrent interventions | Intervention Coverage | Total N (A=intervention/endline; B=control/baseline) | Outcomes | Effect on outcome RR/OR (95% CI) |
---|---|---|---|---|---|---|---|---|---|
O’Rourke[66] | 1991 | Rural Guatemala | Before-after comparison | 3-month hospital-based training program for TBAs - identification of obstetric emergency and referral; encouragement to attend hospital deliveries; strengthening relationships between TBAs and hospital staff | Studied only those patients who were sucessfully referred | A) 465; B) 39 | 1) PMR among referred infants* | RR 0.73 | |
Greenwood et al. [86] | 1983 | Rural Gambia | Before-after comparison | TBA training in intervention villages within a comprehensive primary care program; 10 week training courseantenatal-postnatal care, referral signs; distribute clean birth kit and malaria prophylaxis | Introduction of comprehensive primary health care program, transport improvements | 65% | A) 1159 B) 659 | 1) NMR; 2) PMR | 1) RR 0.66; 2) RR 0.92 |
Janowitz et al. [74] | 1984-85 | Rural NE Brazil | Cross-sectional | TBA training especially in recognition of childbirth complications and referral. Non-randomized comparison of trained TBAs with high case load (>29 births per year) versus unattended home births | Establishment of “mini- maternities” with telephones for TBA births. | 55% | A) 906; B) 118 | 1) NMR | RR 0.60 |
Jokhio et al. [65] | 1998 | Rural Pakistan, Larkana, | Cluster RCT | TBA training in antepartum, intrapartum, postpartum, and neonatal care; distribution of clean delivery kits; referral for emergency obstetrical care. | Lady health workers also trained to support TBA and link community-health center services. | 74% | A) 10114; B) 9443 | 1) PMR; 2) NMR; 3) SBR | 1) aOR 0.71 (0.62-0.83); 2) aOR 0.70 (0.59-0.82); 3) aOR 0.69 (0.57-0.83) |
Excluded from present review --Primary intervention was neonatal resuscitation
| |||||||||
Carlo et al[68]. | 2005-2007 | Argentina, DR Congo, Guatamala, India, Pakistan, Zambia | Before-after study | training of community birth attendants (TBAs, nurses) in WHO Essential Newborn Care , including basic resuscitation with bag-mask in 6 countries | Clean delivery, thermal protection, breastfeeding, kangaroo care | 78% of births (post) | A) 22,626; B) 35,017 | 1) PMR; 2) SBR; 3) ENMR | 1) RR 0.85 (0.70-1.02); 2) RR 0.69 (0.54-0.88); 3) RR 0.99 (0.81-1.22) |
Kumar et al[63] | ns | Rural India | Quasi-experimental | TBAs trained in "advanced" resuscitation with suction and bag-mask vs. usual mouth-mouth resuscitation | TBAs delivered 92% of babies at home | A) 964; B) 884 | 1) "asphyxia" mortality; 2) PMR | 1) RR 0.30 (0.11-0.81); 2) RR 0.82 (0.56-1.19) | |
Daga et al[87] | 1988 | Rural India | Before-after | TBA training in basic mouth-to -mouth breathing | Management of low birth weight, hypothermia; transport and referral of high risk babies to hospital | 90% | A) 321; B) 660 | 1) PMR; 2) NMR; 3) SBR | 1) RR 0.59 (0.32-1.09); 2) RR 0.39 (0.21-0.69); 3) RR 0.49 (0.16, 1.50) |
Gill et al[67] | 2006 | Rural Zambia | Cluster RCT | Training of TBAs in a modified neonatal resuscitation program (NRP) w/resuscitator facemask | prevention of hypothermia, antibiotic treatment and facilitated referral for presumptive neonatal sepsis | uncertain | A) 2007 B) 1552 | 1) NMR; 2) “asphyxia” mortality | 1) aRR 0.55 (0.33-0.90); 2) aRR 0.37(0.17-0.81) |
Azad et al [88] | 2004 | Rural Bangladesh | Cluster RCT, factorial design | Intervention arm: Training of TBAs in neonatal resuscitation with bag-valve mask, with subsequent retraining; Control arm: Training of TBAs in mouth-to-mouth resuscitation | Intervention and control: Clean delivery, danger signs, emergency preparedness, facility referral. Women’s participatory groups in half of clusters | ~20% of home deliveries in both study arms | A) 13195; B) 12519 | ENMR | 1) RR 0.95, (0.75 - 1.21) |
Study Quality | Summary of Findings | |||||||||
---|---|---|---|---|---|---|---|---|---|---|
Directness | Endline/Intervention | Baseline/Control | ||||||||
No of studies | Design | Limitations | Consistency | Generalizability to Population of Interest | Generalizability of intervention of interest | Events | Births | Events | Births | Relative Risk (95% CI) |
Neonatal Mortality
(
All Cause
)
: Low outcome specific quality
| ||||||||||
1 [65] | Cluster RCT | Direct, rural LIC | Yes | 340 | 9710 | 439 | 8989 | aOR 0.70 (0.59-0.82) | ||
1[74] | Cross-sectional | Low quality | Direct, rural LIC | Yes | 23 | 909 | 34 | 119 | RR 0.60 (NS) | |
1 [86] | Before-after | Low quality before-after, improved surveillance post | Direct, rural LIC | 15 | 445 | 23 | 383 | RR 0.66 (NS) | ||
Perinatal Mortality
(
All Cause
)
: Low outcome specific quality
| ||||||||||
1 [65] | Cluster RCT | Direct, rural LIC | Yes | 823 | 9710 | 1077 | 8989 | aOR 0.71 (0.62-0.83) | ||
1 [86] | Before-after | Low quality before-after, improved surveillance post | Direct, rural LIC | Yes | 99 | 1220 | 29 | 398 | RR 0.92 (NS) |
Overall level of evidence
Results of Delphi process
Discussion
Conclusion
Effect of Comprehensive Emergency Obstetric Care
Cause specific mortality to act on:
Intrapartum related neonatal deaths
Quality of input evidence:
Very Low – effect estimates derived from Delphi panel consensus Low quality supporting evidence (8 observational, 1 quasi-experimental)
GRADE recommendation
Strong, based on clear biological mechanism
Cause specific effect and range:
Reduction in intrapartum related neonatal deaths: 85%; IQR 67.5-87.5%
Limitations of the evidence:
Evidence without cause-specific mortality effect, and with varying content of packages and varying contexts for evaluation. Only one quasi experimental design study identified |
---|
Effect of Basic Emergency Obstetric Care
Cause specific mortality to act on:
Intrapartum related neonatal deaths
Quality of input evidence:
Very Low – effect estimates derived from Delphi panel consensus No studies identified specifically of BEmOC with perinatal health outcomes reported
GRADE recommendation
Strong based on clear biological mechanism
Cause specific effect and range:
Reduction in intrapartum related neonatal deaths: 40%; IQR 40-52.5%
Limitations of the evidence:
No evidence available regarding effect of this specific package, even from observational designs. |
Effect of Skilled Childbirth Care
Cause specific mortality to act on:
Intrapartum related neonatal deaths
Quality of input evidence:
Very low – effect estimates derived from Delphi panel consensus Low quality supporting evidence (2 Quasi-experimental, 8 observational)
GRADE recommendation
Strong
Cause specific effect and range:
Reduction in intrapartum related neonatal deaths: 25%; IQR 15-30%
Limitations of the evidence:
Single study with cause-specific mortality effect. For the studies identified the content of the packages tested and the contexts for evaluation and evaluation designs were variable |
Effect of Trained Traditional Birth Attendants
Quality of input evidence:
Low quality supporting evidence (3 cRCT, 1 quasi-experimental, 5 observational)
GRADE recommendation
Conditional, dependent on local context and health system
Cause specific effect and range:
Not estimated for LiST since GRADE recommendation is conditional
Limitations of the evidence:
Supporting evidence without cause-specific mortality effect, and with varying content of packages and varying contexts for evaluation. 5 studies primarily of TBA training in neonatal resuscitation that is NOT included as part of the estimate for childbirth care package |