Im 2005.

Methods	Randomised trial, allocation by drawing from a box of envelopes Assessor blinding: one physician who had not been involved in the treatment performed an independent evaluation of all trial participants at 24 months Length of follow‐up: 24 months
Participants	Dongguk University Hospital, Goyang, Korea Period of study: July 1998 to June 2001 78 participants, each with a distal metaphyseal fracture of the tibia. Two died and 12 did not complete the study at 2 years follow‐up; these were excluded from the trial analyses and baseline characteristics Sex (of 64): 18 female, 46 male Age: mean 41.1 years, range 17 to 65 Fracture type: AO/OTA 43‐A1: 26, 43‐A2: 19, 43‐A3: 9, 43‐C1: 10 13 open fractures (Gustilo type I) and 51 closed fractures; 35 also had fibula fractures Inclusion criteria 1. Distal metaphyseal fractures of tibia 2. In all of the included cases, the centre of the fracture was in the distal metaphysis of tibia and either entirely extra‐articular (A1, A2, and A3) or with minimally displaced extension into the ankle joint (C1) Exclusion criteria 1. Open fractures of Gustilo‐Anderson type II or III 2. Fractures with displaced intra‐articular fragments 3. Based on the preoperative roentgenograph examination, two distal screw fixations were found to be not feasible (this applied to one case) Assigned: ?/? participants (intramedullary nail/plate) Analysed: 34/30 (24 months)
Interventions	1. Closed reduction and intramedullary nailing (ACE tibial nails (ACE Depuy, El Segundo, California) used in 16 cases; or cannulated tibial nails (Mathys, Bettlach, Switzerland) used in 18 cases). Nails were inserted after reaming. Static locking applied in all cases. 2. Open reduction and fixation with anatomic plates (Waldermar‐Link, Hamburg, Germany) and screws. Eleven of the 35 fibular fractures were fixed with small dynamic compression plates (6 in the nail group versus 5 in the plate group). Other care: A long leg plaster cast was applied. After wound swelling had subsided, ankle joint mobilisation was encouraged under physiotherapist supervision. Timing of weight‐bearing was on an individual basis. A short leg cast was applied for participants with minimally displaced intra‐articular fractures.
Outcomes	Olerud and Molander Ankle Functional Score (OMAS) Re‐operation or surgery for complication Fracture non‐union, tibia malalignment Complications including superficial and deep wound infection, Fracture healing time Operation time, intraoperative fracture comminution Ankle range of motion
Notes	‐
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"The treatment method was chosen at random by drawing from the box containing an equal number of envelopes with either of the methods"
Allocation concealment (selection bias)	Unclear risk	"The treatment method was chosen at random by drawing from the box containing an equal number of envelopes with either of the methods". Adequate safeguards not described.
Blinding of participants and personnel (performance bias) All outcomes	High risk	The difference between the two methods means that blinding is not possible for the surgeon and unlikely for the participants. There was no statement on blinding.
Blinding of outcome assessment (detection bias) All outcomes	High risk	"To limit the bias inherent in all clinical examinations and manual tests, the final follow‐up examination, 2 years after the operation, was performed by an independent physician who had not been involved in the actual treatment of any of the patient." However, this was not blinded and blinding of outcome assessment was not reported for other time points prior to 2 years after the operation.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Treatment allocation and outcome were not provided for 14 participants (18% of 78), 2 of whom had died and 12 who were lost to follow‐up.
Selective reporting (reporting bias)	High risk	No protocol provided and the OMAS results only reported in the abstract and not in a way described in the methods.
Other bias	Unclear risk	Since the baseline characteristics of 14 participants were not provided, we cannot tell if the baseline characteristics were balanced. Although similar numbers in the 2 groups had surgically fixed fibular fractures (6 in nail group and 5 in plate group), we cannot be sure that the decision to operate was comparable in the 2 groups. Experience with both methods was stated; all operations were carried out or supervised by one senior surgeon.