Christodoulou 2005.

Study characteristics
Methods	Quasi‐RCT: allocation alternated between interventions after stratification by fracture type
Participants	Setting: Hippokratio General Hospital, Thessaloniki, Greece Size: 80 participants; 35 in nail group, 37 in plate group (allocation not reported in 8 participants, 5 died and 3 lost to follow‐up) Recruitment period: January 1994 to June 1999 (2000 in abstract) Baseline characteristics: 25 men, 47 women; median age 73.2 years (range 60–88 years); all fractures were closed without substantial soft tissue damage; fracture types in the nail group were AO/ASIF A1 34%, A2 40%, A3 14%, C1 9%, C2 3%; and in the plate group were A1 35%, A2 35%, A3 16%, C1 8%, C2 6%. Age, sex and other variables not reported Inclusion criteria: supracondylar fracture of femur Exclusion criteria: not reported
Interventions	Participants were allocated to either: RIMN via an intercondylar notch approach. Nails were 20–25 cm long and locked with 2 proximal screws and 2 or 3 distal screws; DCS with 95° angle plate. Primary autologous bone grafting of the medial cortex used in 3 participants. A tourniquet and wound drainage used for all participants. Prophylactic perioperative antibiotics and postoperative anticoagulants used in all participants. Continuous passive motion started on the 2nd or 3rd postoperative day and mobilisation with partial weight‐bearing on 4th or 5th day.
Outcomes	Schedule: 6, 12, 24 and 52 weeks and annually thereafter. Mean follow‐up 28 months (range 18–42 months) Outcomes: Schatzker and Lambert criteria (full extension, loss of flexion, valgus/varus/rotational deformity, pain and joint congruency), length of hospital stay, operation time, blood loss, radiological union, clinical union, complications, death
Notes	Funding/sponsor/declarations of interest: not reported Attempt at personal communication unsuccessful
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	Quote: "allocation alternated between each intervention after stratification by AO fracture classification". Comment: quasi‐randomised.
Allocation concealment (selection bias)	High risk	The next intervention for each fracture type was known.
Blinding of participants and personnel (performance bias) All outcomes	High risk	The nature of the interventions meant that the operating surgeons were not blinded. No details relating to participant blinding. Providers of interventions not stated.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	No mention made of blinding of outcome assessment. Additionally, outcome assessment poorly reported.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Only 10% participants were lost to follow‐up: 5 died and 3 changed address. Allocation of these participants not reported and they were not included in final analysis.
Selective reporting (reporting bias)	Unclear risk	No protocol available.
Other bias	Low risk	We identified no other sources of bias.