Gardiner 1990.

Study characteristics
Methods	Design: RCT or quasi‐RCT Unit of randomization: baby
Participants	Location/Setting: UK Sample size: 79 Number of withdrawals/dropouts: none stated Sex: intervention = 56 girls, 23 boys; control = 54 girls, 25 boys Mean age: 40 (SD 1.6) weeks Graf: babies: I = 34, IIa = 50, IIc = 19, III=44, IV = 11; controls: I = 82, IIa = 61, IIc = 6, III = 8, IV = 1 Inclusion criteria: diagnosed within 24 hours of birth; clinically dislocatable hips Exclusion criteria: clinically dislocated hips (they were splinted)
Interventions	Intervention (sample size): splinted immediately, type of splint unclear but follow‐up paper suggests Aberdeen Abduction Splint (n = 41) Control 1 (sample size): sonographic surveillance group (n = 38). Ultrasound scan repeated at 10 to 14 days in this group. If remained clinically unstable or no sonographic improvement, then splinting was commenced (n = 11) and if improved, then continued without splint for the full 6 weeks (n = 27) Control group 2 (sample size): matched for sex, first‐born status, fetal presentation, gestational age, and family history of congenital dislocation of the hip in a first‐degree relative (n = 79)
Outcomes	Primary outcome(s): hip normal/abnormal Secondary outcome(s): avascular necrosis; need for operative intervention Timing of outcome assessment: 6‐8 weeks, 6 months, 1 year
Notes	Study start date: 1988 Study end date: 1989 Funding source: financial support from Children Nationwide, Southmead Hospital Research Fund; and the Van Neste Foundation Conflicts of interest: not stated Comment(s): babies in the control group were matched but the analysis does not seem to take this into account. “Static hips scans were measured according to Graf’s system, types Ia, Ib, and IIa being classified as normal, and types IIb, IIc, IIIa, IIIb and IV as abnormal. Hips that showed dynamic instability were regarded as abnormal even if the static sonographic morphology appeared normal”. Classified according to Graf. Alpha angles not mentioned (but are integral to Graf)
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	Judgement comment: babies were randomized by alternate allocation to splinting or surveillance. No information was given on the method used.
Allocation concealment (selection bias)	Unclear risk	Judgement comment: not stated
Blinding of participants and personnel (performance bias) All outcomes	High risk	Judgement comment: caring physician and baby could not be blinded.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Judgement comment: those assessing outcomes were blinded.
Incomplete outcome data (attrition bias) All outcomes	High risk	Judgement comment: no causal analysis performed to account for treatment switching, which may lead to bias
Selective reporting (reporting bias)	Unclear risk	Judgement comment: no protocol/trial registry information available to compare pre‐specified outcomes with reported outcomes. The three outcomes mentioned in the methods section were fully reported.