Downe 2015.

Methods	"Multi‐site, pragmatic, non‐blinded RCT based on intention to treat."
Participants	"680 women were randomised." "Participants were 27‐32 weeks' gestation at the time of randomisation, could read and understand English, were not on medication for hypertension or psychological illness, and were not planning an elective caesarean section. Birth partners were eligible to take part if they returned a consent form."
Interventions	Intervention: "The intervention group received self‐hypnosis training in addition to usual care. Two 90‐minute group sessions were offered, 3 weeks apart, at around 32 and 35 weeks gestation". "Participants were invited to attend group sessions at their local Trust, with or without their intended birth companion. They were also advised to listen to a 26‐minute self‐hypnosis CD daily (recorded by VM) until the birth of the baby, and to complete logs of this practice, and of other antenatal educational activities. Fifteen midwives were trained in hypnosis techniques by the same trainers (though at different times). All hypnosis midwives were visited by a member of the research team at least once during a self‐hypnosis session to ensure fidelity to the intervention protocol.” Total number randomised: n = 340. Control/Comparison intervention: “Those randomised to this group continued with usual care only”. Total number randomised: n = 337.
Outcomes	"Primary outcome measure Use of epidural analgesia for labour pain relief. Secondary outcome measures The secondary outcome measures assessed covered four domains. 1. Measures of the hypothesised mechanism of effect of the intervention · Anxiety and fear about labour (and changes in these measures over time) 2. Impact of the mechanism of effect (fear and anxiety) · Mode of labour onset · Length of labour · Use of any pharmacological pain relief 3. Outcomes associated with the use of pharmacological pain relief in labour · Method of infant feeding (breastfeeding at 6 weeks postnatal) · Mode of birth · Postnatal recollection of pain during labour 4. Measures of morbidity and of well‐being in the postnatal period · Clinical morbidity · Psychological morbidity · Satisfaction with childbirth · Preferences relating to hypnosis Secondary outcomes The main outcome of the economic analysis was the Incremental Cost Effectiveness Ratio (ICER), based on resource use per quality‐adjusted life year (QALY), measured using the EQ‐5D instrument. This assesses health status on five dimensions (mobility, self‐care, ability to undertake usual activities, pain, anxiety/depression). Three specific phases of resource utilisation were assessed: the activities undertaken during the antenatal period, an inventory of the resources required during labour, and services required as part of post‐partum admissions. Full details of the cost effectiveness analysis conducted alongside this clinical trial will be reported separately." 7‐point scale used for measure of satisfaction with pain relief at 2 weeks postpartum.
Notes	The author was contacted with a request for additional data in March 2015 and these were provided.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	“We used a computer‐generated sequence on a 1:1 basis, without stratification or blocking.”
Allocation concealment (selection bias)	Low risk	“The allocation was uploaded automatically to the participant management database, accessible by password to the research team, to allow for session allocation for the intervention group.”
Blinding of participants (objective outcomes)	Low risk	“Blinding to randomisation was not possible for participants or the hypnosis trainers." Lack of blinding unlikely to effect objective measures.
Blinding of participants (subjective outcomes)	High risk	“Blinding to randomisation was not possible for participants or the hypnosis trainers."
Blinding of personnel (performance bias)	High risk	Author advised that personnel caring for the women in labour were not blinded to group allocation.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	“Outcomes data were collected by staff that did not know group allocation, and returned separately to the CTU, for data linkage."
Incomplete outcome data (attrition bias) All outcomes	Low risk	Minimal loss of women, "680 women were randomised, three in error (who were withdrawn). Two women requested withdrawal and three were lost to follow up. Data are therefore available for 672 women (337 intervention and 335 control)". None were excluded from analysis. Losses were balanced across groups and not likely to be related to the intervention.
Selective reporting (reporting bias)	Low risk	All outcomes reported in published paper or provided as unpublished data. Further publications planned.
Other bias	Low risk	Baseline characteristics appear similar between the groups, apart for education – (% GCSE or below) – slightly more women had achieved GCSEs in the hypnosis group 21% versus 16.3% in the control group. See table 2 – baseline characteristics.