Adams 2001.

Study characteristics
Methods	RCT; parallel design Review comparison group: Gamma nail versus SHS
Participants	Total number of randomised participants: 400 Inclusion criteria: diagnosis of intertrochanteric fractured femur Exclusion criteria: inability to give informed consent, too frail for any operative intervention, and residence outside the region of the hospital because of the difficulty of follow‐up Setting: single centre; orthopaedic hospital, UK Intervention group 1 (Gamma nail)  Age, mean (range): 81.2 (48 to 99) years Gender, M/F: 39/164 Mobility assessment, independent/1 stick/2 sticks/walking frame/wheelchair and transfer, n: 88/53/2/32/28 Place of residence, own home/part IV or relative or home for elderly/acute hospital/nursing home or long stay, n: 104/27/5/67 Preoperative waiting time, mean (range): 1.7 (1.5 to 1.9) days Fracture classification, AO/OTA A1.1/A1.2/A1.3/A2.1/A2.2/A2.3/A3.1/A3.2/A3.3/B2.1, n: 41/38/0/27/40/33/2/2/2/18 Intervention group 2 (DHS) Age, mean (range): 80.7 (32 to 102) years Gender, M/F: 49/148 Mobility assessment, independent/1 stick/2 sticks/walking frame/ wheelchair and transfer, n: 88/48/6/29/26 Place of residence, own home/part IV or relative or home for elderly/acute hospital/nursing home or long stay, n: 115/27/12/43 Preoperative waiting time, mean (range): 1.8 (1.6 to 2.1) days Fracture classification, AO/OTA A1.1/A1.2/A1.3/A2.1/A2.2/A2.3/A3.1/A3.2/A3.3/B2.1, n:  43/29/1/22/50/27/1/4/5/15 Note:  study authors do not report baseline characteristics for: smoking history, medications, BMI, comorbidities, cognitive status/dementia or ASA status
Interventions	General details: study authors report that surgeons were experienced with both implants; both groups received standard 3‐dose IV cefuroxime and routine antithrombotic prophylaxis; clinical follow‐up for 1 year or until death (3 months, 6 months, 12 months) Intervention group 1 Gamma intramedullary nail (Stryker‐Howmedica Ltd, London, UK); short type. Study authors did not report if the lag screw was static or locked; most common implant was the 130‐degree by 11 mm nail; distal locking screws were used at the preference of the surgeon Number randomised = 203; losses = unknown (study authors report one loss but do not report from which study group so we have used all participants in analysis for both groups); analysed = 203  Intervention group 2 CHS (Smith & Nephew, UK). The most common implant was the 135‐degree, 3‐hole plate Number randomised = 197; losses = unknown (study authors report one loss but do not report from which study group so we have used all participants in analysis for both groups), analysed = 197
Outcomes	Outcomes measured/reported by study authors: length of surgery; operative blood loss; postoperative haemoglobin; tip‐apex distance; number of participants transfused; operative fracture of the femur; later fracture of the femur; cut‐out of implant; detachment of the plate from the femur; re‐operation; deep wound infection; superficial wound infection; DVT; mortality; use of walking aids; place of residence at follow‐up; HHS (available at 3, 6, 12 months) Outcomes relevant to the review: mortality (12 months); unplanned return to theatre (12 months)
Notes	Funding/sponsor/declarations of interest: quote: "Although none of the authors has received or will receive benefits for personal or professional use from a commercial party related directly or indirectly to the subject of this article, benefits have been or will be received but are directed solely to a research fund, the Scottish Orthopaedic Research Trust into Trauma, a non‐profit organisation with which one or more of the authors is associated." Study dates: February 1994 to June 1995
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	No details
Allocation concealment (selection bias)	Low risk	Quotes: "At admission, patients were randomized by a closed, opaque envelope method and were assigned to receive either..." Confirmed by Adams in 2001 that "the opaque envelopes were sequentially numbered" ‐ and that there was concealment of allocation.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	It is not possible to blind surgeons to the type of intervention. We did not, however, expect that this would influence surgeon performance.
Other performance bias: surgeon experience of both implants	Low risk	Quote (from draft report): "The surgeons were experienced in the insertion of both implants".
Blinding of outcome assessment: mortality (detection bias)	Low risk	We did not expect that lack of blinding of outcome assessors would influence objective outcome data.
Blinding of outcome assessment: unplanned return to theatre (detection bias)	High risk	It is not possible to blind surgeons to treatment groups. We judged that knowledge of the type of implant could influence judgments made by surgeons when assessing subjective outcomes.
Blinding of outcome assessment: HRQoL (detection bias)	Low risk	Quote: "Observed‐blinded functional assessments were carried out by the unit research physiotherapist, by use of the Harris hip score."
Incomplete outcome data (attrition bias) All outcomes	Low risk	Most losses were explained by death, which is expected in this population, and losses were reasonably balanced between groups.
Selective reporting (reporting bias)	Unclear risk	Study authors do not report details of pre‐published protocol or clinical trial registration. It is not feasible to effectively assess risk of reporting bias without these documents.
Other bias	Low risk	We identified no other sources of bias.