Table 2

 Characteristics of studies included in review

Study, countryDetailsPopulationCharacteristics of womenNo (%) of caesarean sectionsIndication for caesarean sectionSurgeon allocationOperation typePotential confoundersOutcomes
White 1987,11 Zaire
 Clinical officersNurses trained to carry out caesarean sections, laparotomies, and supracervical hysterectomiesPregnant women needing caesarean section at two rural hospitals or health centres in northwest ZaireNot given310 (88)No separate details given on surgeon and indication for caesarean section; details only given for Karawa hospital (321 caesarean sections): cephalopelvic disproportion and obstructed labour 159 (50%), uterine inertia 70 (22%), previous caesarean section 54 (17%), fetal distress 22 (7%), placenta praevia 16 (4%)One unit was without a doctor for one year; nurses carried out caesarean sections. In one unit doctors did not routinely do the most complicated operations, whereas in the other unit doctors operated on the most critical cases. Doctors carried out caesarean sections when no nurse was available, or when medical students were being trainedNot givenDelay in seeking care, delay in reaching care (transportation), other medical complications. No adjustments madeMaternal death
 DoctorsQualified doctorNot given43 (12)Not given
Pereira 1996,1 Mozambique:
 Clinical officersAssistant medical officers (clinical officers with surgical training)Pregnant women needing caesarean sections at central hospital. Caesarean section rate was 16.5%. This was only hospital carrying out caesarean sectionsAge 25.3 years, parity 2.2, resided in shanty town 353 (37%), antenatal care 948 (99%), twin pregnancy 25 (3%)958 (46)Fetal distress 308 (32.2%), cephalopelvic disproportion 177 (18.5%), previous caesarean section 89 (9.3%), placental abruption or praevia 79 (8.2%), impending uterine rupture 72 (7.5%), eclampsia 31 (3.2%), pre-eclampsia 26 (2.7%), no information 176 (18.4%)No random allocation for ethical reasons. Some selection may have occurred when patients were allotted to clinical officers or doctorsCaesarean section only 682 (71%), caesarean section plus subtotal hysterectomy 8 (1%), caesarean section plus total hysterectomy 3 (0.5%), caesarean section plus uterine repair 4 (0.5%), caesarean section plus tubal ligation 257 (26.5%), no information 4 (0.5%)Age, parity, resided in shanty town, homeowner, antenatal care, twin pregnancy. No adjustments madeMaternal death, duration of postoperative stay, wound rupture or separation, condition of newborn
 DoctorsTrained obstetricians and gynaecologistsAge 25.5 years, parity 2.2, resided in shanty town 404 (36%), antenatal care 1101 (99%), twin pregnancy 21 (2%)1113 (54)Fetal distress 326 (29.3%), cephalopelvic disproportion 212 (19%), previous caesarean section 133 (11.9%), placental abruption or praevia 74 (6.6%), impending uterine rupture 68 (6.1%), eclampsia 40 (3.6%), pre-eclampsia 33 (3%), no information 227 (20.5%)It was noted that emergencies prevented doctors attending complicated caesarean section. Furthermore, doctors carried out all elective caesarean sections (145, 7.0%)Caesarean section only 832 (74.5%), caesarean section plus subtotal hysterectomy 10 (1%), caesarean section plus total hysterectomy 4 (0.5%), caesarean section plus uterine repair 5 (0.5%), caesarean section plus tubal ligation 256 (23%), no information 6 (0.5%)
Fenton 2003,8 Malawi:
 Clinical officersClinical officer surgeonsWomen requiring caesarean section to deliver baby, dead or alive, after 28 weeks, including surgery for ruptured uterus. 23 district and two central hospitalsNo separate details given on surgeons for characteristics: district hospital 5236 (65%), urban hospital 2834 (35%)5256 (65)No separate details given on surgeon and indication for caesarean. Obstructed labour 5110 (63%), fetal distress 885 (11%), antepartum haemorrhage 384 (5%), pre-eclampsia 268 (3%), haemorrhagic shock 610 (8%), ruptured uterus 333 (4%), other 480 (6%)Not givenNot givenPrevious caesarean section, preoperative complications, fever, status and training of anaesthetist and surgeon, type of anaesthesia, blood loss, and need for blood transfusion. Adjusted for confoundersMaternal death, perinatal death
 DoctorsMedically qualified doctors2814 (35)Not given
Chilopora 2007,10 Malawi:
 Clinical officersLicensed to practise independently and to carry out major emergency and elective surgery. Length of surgical experience: 44% >4 years, 24.3% 2-3 years, 21.4% <1 year, 9.3% 0 years, 0.6% no dataAll women undergoing caesarean section during study period were included in 38 district hospitals; most were emergency caesarean sectionsNo major differences between groups operated on by clinical officers and doctors1875 (88)No separate details given on surgeon and indication for caesarean section: cephalopelvic disproportion or obstructed labour 1230 (57.7%), previous caesarean section 452 (21.2%), fetal distress 264 (12.3%), impending uterine rupture 87 (4%), antepartum haemorrhage 77 (3.6%), cord prolapse 62 (2.9%), failure to progress, 60 (2.8%), breech in primigravida 53 (2.5%), eclampsia 49 (2.3%)Not givenCaesarean section only 1569 (84%), caesarean section plus subtotal hysterectomy 11 (0.5%), caesarean section plus total hysterectomy 7 (0.5%), caesarean section plus uterine repair 59 (2.5%), caesarean section plus tubal ligation 224 (12%), not indicated 5 (0.5%)Duration of surgeon’s practice, region of hospital, type of operation (if hysterectomy needed also). No adjustments madeMaternal death, infection, wound dehiscence, fever, perinatal death
 DoctorsTrained doctors. Length of surgical experience: 59% >4 years, 19.9% 2-3 years, 17.2% <1 year, 3.9% no data256 (12)Caesarean section only 185 (72%), caesarean section plus subtotal hysterectomy 8 (3%), caesarean section plus total hysterectomy 3 (2%), caesarean section plus uterine repair 7 (3%), caesarean section plus tubal ligation 53 (20%), not indicated 0 (0%)
Hounton 2009,3 Burkina Faso:
 Clinical officersClinical officers working in both rural and urban hospital2305 pregnant women needing caesarean section in 22 public sector urban and rural hospitalsAge 25 years, urban hospital 198 (27%), rural hospital 535 (73%)733 (32)Obstructed labour 388 (53%), ruptured uterus 81 (11%), eclampsia 15 (2%), haemorrhage 44 (6%), other 205 (28%)No details given on selection of operator; retrospective study. Doctors were, however, associated with more referred cases, although this was not statistically significantNot givenPlace of hospital, maternal age, reported clinical conditions; obstructive labour, ruptured uterus, eclampsia, haemorrhage, referral from other unit, type of anaesthesia. Adjustments made but adjusted statistics not givenMaternal death, perinatal death, haemorrhage, wound infection, wound dehiscence
 DoctorsTrained doctors based at urban and rural hospitals Age 25 years, urban hospital 877 (56%), rural hospital 143 (9%)1572 (68)Obstructed labour 679 (43%), ruptured uterus 151 (9%), eclampsia 76 (5%), haemorrhage 84 (6%), other 581 (37%)Not given
McCord 2009,12 Tanzania:
 Clinical officersOfficially authorised assistant medical officers to provide clinical services, prescriptions, and minor surgery. Permitted to carry out obstetric care and caesarean section1088 pregnant women in 14 mission and government hospitalsNot given945 (87)Absolute maternal indication (antepartum haemorrhage, postpartum haemorrhage, malpresentation, eclampsia, ectopic pregnancy, ruptured uterus, sepsis, and repair of vesicovaginal fistula) 312 (33.1%), major acute problem (no details given) 63 (6.6%), major chronic problem (severe anaemia, symptomatic AIDS, symptomatic malaria) 172 (18.2%), no information 398 (42.1%)No reason given for choice of operator. Clinical officers were, however, more likely to carry out caesarean section with absolute maternal indicators or clear fetal indicator. Clinical officers had more difficulties than doctors in obtaining blood for transfusionNot givenCondition on admission, type of operation, indication for operation, blood transfusions. No adjustments madeMaternal death, perinatal death
 DoctorsTrained doctorsMedical school graduates with >1 year internshipNot given143 (13)Absolute maternal indication (antepartum haemorrhage, postpartum haemorrhage, malpresentation, eclampsia, ectopic pregnancy, ruptured uterus, sepsis, and repair of vesicovaginal fistula) 48 (33.6%), major acute problem (no details given) 14 (9.8%), major chronic problem (severe anaemia, symptomatic AIDS, symptomatic malaria) 22 (15.3%), no information 59 (41.3%)No reason given for choice of operator. Doctors were, however, less likely to carry out caesarean section for absolute maternal indicators or clear fetal indicator. Doctors had less difficulties than clinical officers in obtaining blood for transfusionNot given