French Doppler 1997.

Methods	Multicentre randomised controlled trial. Individual women. Randomisation in blocks of 4.
Participants	Inclusion criteria Women attending routine AN visit between 28‐34 weeks who had a normal US scan (fetal biometry above 10th centile of reference curve).    N = 4187 randomised with 3898 analysed. Exclusion criteria Women who had indications for Doppler at last 2nd trimester appointment, e.g. medical history of hypertension, diabetes, previous fetal death, IUGR, hypertensive disorder of pregnancy, treatment with beta agonists, insulin‐dependent diabetes. Women who at their last second trimester appointment had indications for umbilical Doppler. Women who had undergone an umbilical Doppler before 28 weeks for any reason whatever.
Interventions	Experimental intervention:  umbilical Doppler ultrasound Single assessment at 28‐34 weeks. Umbilical Doppler US on day of randomisation, immediately after ultrasound scan monitoring fetal growth. All further tests performed at clinicians’ request according to standard practices in their AN clinics. N = 2099 with 1950 analysed. Control/comparison intervention: no Doppler ultrasound No Doppler US on day of randomisation. Access to Doppler studies was allowed on obstetrician’s request. N = 2088 with 1948 analysed. Single estimation between 28‐34 weeks.
Outcomes	AN consultations; days of AN hospitalisation; CTG; ultrasound and Doppler tests. Peri‐ and neonatal deaths; FD; Apgars; neonatal resuscitation; neonatal transfer; birthweight; SGA. Disorders occurring after randomisation e.g. PIH; pre‐eclampsia, uterine bleeding, oligohydramnios, suspected IUGR, abnormal CTG patterns.
Notes	France 1997 study in previous version of the review (Bricker 2007). Authors did report umbilical cord pH < 7.20 but only on a subsample of women and the groups were not randomised groups. Findings were: Doppler 188 / 757 (24.8%) and no Doppler 181/761 (23.8%).
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	“...randomly divided...” Central randomisation in blocks of four. Sequence generation not described.
Allocation concealment (selection bias)	Low risk	“The randomisation procedure using sealed envelopes was standard in each centre and was carried out by the ultrasonographer immediately after verification of the inclusion criteria and performance of the standard ultrasound scan. The envelopes were prepared centrally and were sent to each centre consecutively numbered.”  No ‘opaque’ mentioned. “The randomisation sequence was verified in two ways. After the end of the study every centre was asked to send back the enveloped that had not been used. It was also checked that the envelopes were used in ascending order.” Blocks of 4 means that the sequence may have been predicted for at least 1/4 of women. However, with central randomisation and sealed envelopes there is no risk of bias.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Blinding not possible.
Blinding of outcome assessment (detection bias) All outcomes	High risk	Blinding not possible.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Describe any loss of participants to follow‐up at each data collection point: 174 women lost to follow‐up. Doppler; 25 (1.2%) lost just after randomisation and 66 (3.3%) data were not available for analysis. No Doppler: 27 (1.3%) lost just after randomisation and 56 (2.8%) data were not available for analysis.                                                         Describe any exclusion of participants after randomisation: 115 were excluded (58 (2.8%) in Doppler and 57 (2.7%) in No Doppler. All these women were from 3 centres where randomisation was not undertaken properly in that envelopes were not used in ascending order.   Was the analysis ITT? If not has the data been able to be re‐included? Appears to be ITT and losses are only small % and evenly distributed between the 2 groups.
Selective reporting (reporting bias)	Unclear risk	We have not assessed the trial protocol.
Other bias	Low risk	There appear to be no other biases.