Skip to main content
. 2017 Mar 30;2017(3):CD000161. doi: 10.1002/14651858.CD000161.pub2

Thubisi 1993.

Study characteristics
Methods Prospective randomised controlled trial. Women individually randomised. 2 treatment arms.
Participants 288 women randomised.
Inclusion criteria: women with 1 previous transverse lower segment caesarean section.
Exclusion criteria:
‐ abnormal lie or presentation;
‐ obstetric complications requiring planned delivery;
‐ maternal disorders contra‐indicating a trial of scar;
‐ multiple pregnancy;
‐ preterm labour;
‐ grossly contracted pelvis on clinical examination;
‐ intrauterine death.
Interventions Intervention group: 144 women allocated to x‐ray pelvimetry group at 36 weeks. A sagittal inlet < 11 cm, sagittal outlet < 10 cm, transverse inlet < 11.5 cm, and transverse outlet (bispinous) < 9 cm was an indication for caesarean section. The remainder of the group awaited spontaneous labour and underwent a 'trial of scar’.
Comparison group: 144 women had no pelvimetry at 36 weeks and awaited spontaneous labour.
Outcomes

  1. Caesarean section

  2. Perinatal mortality

  3. Birthweight

  4. Scar dehiscence

  5. Puerperal pyrexia

  6. Wound sepsis

  7. Blood transfusion
Notes 153 women were randomised to either group. In the study group, 1 withdrew consent, 2 had breech presentations, 2 had twin pregnancies, 2 had hypertension and 2 developed preterm labour. In the control group 3 elected to have an elective caesarean section, 2 had breech presentations, 1 twin gestation, 2 hypertensives and 1 preterm labour. Each group consisted finally of 144 women. Analysis was on the last number and not according to intention to treat. 6 women had scar dehiscences, 2 diagnosed in labour (control group) and 4 on routine digital examination after delivery. None of the women required hysterectomy or had postpartum haemorrhage.
Trial took place at King Edward VIII Hospital, Durban.
Funding source: not stated.
Dates study was conducted: randomisation occurred "during the second half of 1990", primary outcome follow‐up completed February 1991
Declarations of interest of primary researchers: unclear
Risk of bias
Bias Authors' judgement Support for judgement
Random sequence generation (selection bias) High risk "Randomisation and equal distribution were assured because women were allocated alternately to the two teams by admitting clerks who had no medical training and no knowledge of how they would be managed."
Allocation concealment (selection bias) High risk Not mentioned but a different medical team provided the intervention and control care therefore no concealment attempted.
Blinding of participants and personnel (performance bias)
All outcomes High risk Not mentioned. Difficult to blind this type of intervention.
Blinding of outcome assessment (detection bias)
All outcomes High risk Management of care and outcome recording was done by different teams of staff for women in the 2 groups. This means outcomes may not have been measured and recorded in the same way.
Incomplete outcome data (attrition bias)
All outcomes Unclear risk 306 women randomised. 288 followed up ‐ loss was relatively low but loss of 2 women in the pelvimetry group related to outcomes (women opted for caesarean section).
Selective reporting (reporting bias) Unclear risk Outcomes not mentioned in methods text, protocol not available.
Other bias Low risk Baseline characteristics appeared similar. Other bias not apparent.