Davies 1992.

Methods	Randomised controlled trial. Individual women. Queen Charlotte's and Chelsea Hospital, London, UK.
Participants	Inclusion criteria Pregnant women, with singleton pregnancies, at 19‐22 weeks pregnant. Unselected. Low‐ and high‐risk pregnancies: high risk 189 in Doppler group and 192 in control group. 2600 women ‐ 79% of eligible population.  Exclusion criteria Multiple pregnancies.
Interventions	Experimental intervention: Doppler ultrasound of umbilical‐artery and uterine‐artery Multiple assessments at 20 and 32 weeks. Women with low‐risk pregnancies had Doppler at booking and 32 weeks. Low risk for SGA or other compromised infant. Women at high risk had ultrasound every month. High risk = women identified before entry into trial by: pre‐existing medical condition, e.g. hypertension, diabetes, previous SGA baby, previous stillbirth pr neonatal death, hypertension (BP > 140/90 mmHg) in previous pregnancy or at booking, smoking > 10 a day. Any women in low‐risk group who had abnormal Doppler was managed subsequently as high risk. If subsequent examination was normal the woman transferred back to low‐risk group. Clinician could have Doppler at other times as requested. N = 1246. Control/Comparison intervention:  no Doppler ultrasound Intended that women should not have Doppler US at anytime in pregnancy. Normal AN care with no Doppler. N = 1229. Multiple estimations were at 20 and 32 weeks' gestation. Sample size calculation "to have an 80% chance at the 5% level of significance of demonstrating a 20% reduction in antenatal admissions during pregnancy in the doppler group."
Outcomes	Number of days of antenatal admission; number of CTG recordings and US scans; gestational age at birth; mode of birth; birthweight; Apgar scores; need for resuscitation (intermittent positive pressure ventilation either via a mask or endotracheal tube); admission to NICU; fetal and neonatal outcomes. The study was not designed to test the ability of Doppler ultrasound to reduce PNM, so the fact that there were more preventable deaths in the Doppler group is likely to be due to chance. However, the authors do theorise that it is possible that a woman's knowledge of a normal result may have resulted in her taking less notice of symptoms that might otherwise have resulted in a review of fetal well‐being.
Notes	London (UK) 1992 study in previous version of the review (Bricker 2007).
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Sequence generation not described; randomisation conducted in blocks of 500 and 200.
Allocation concealment (selection bias)	Low risk	Allocation "by cards in sealed opaque envelopes."
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.                                    Describe any exclusion of participants after randomisation: 125 women (4.8%) were excluded because: 106 gave birth elsewhere; 8 were randomised then found to have missed abortion; 2 multiple pregnancies; or because randomisation care (7), Doppler data (1) or hospital notes (1) went missing. Demographics similar to rest of study population . Was the analysis ITT? If not, have the data been able to be re‐included? Loss was small and unlikely to impact on outcomes.
Selective reporting (reporting bias)	Unclear risk	There are discrepancies in the numbers of neonatal deaths between reports of this trial. We have used the numbers reported in the Lancet 1992 article, including early neonatal deaths added together with neonatal deaths.
Other bias	Low risk	If the study was stopped early, explain the reasons: Not stopped earlier.  Describe any baseline imbalance: “Women in the Doppler and the control groups did not differ in their demographic details (Table 1)”. So assessed by age, weight at booking, ethnic origins, nulliparous, smoking, shared antenatal care, high risk. 15 (1.2%) of the 1229 women in the control group had Doppler.  Describe any differential diagnosis: Seems fine.