Surgical fixation compared with cast immobilisation for adults with a bicortical fracture of the scaphoid waist: the SWIFFT RCT

Background: Scaphoid fractures account for 90% of carpal fractures and occur predominantly in young men. Immediate surgical fixation of this fracture has increased. Objective: To compare the clinical effectiveness and cost-effectiveness of surgical fixation with cast treatment and early fixation in adults with scaphoid waist fractures that fail to unite. Design: Multicentre, pragmatic, open-label, parallel two-arm randomised controlled trial with an economic evaluation and a nested qualitative study. Setting: Orthopaedic departments of 31 hospitals in England and Wales recruited from July 2013, with final follow-up in September 2017. Participants: Adults (aged = 16 years) presenting within 2 weeks of injury with a clear, bicortical fracture of the scaphoid waist on plain radiographs. Interventions: Early surgical fixation using Conformite Europeenne-marked headless compression screws. Below-elbow cast immobilisation for 6-10 weeks and urgent fixation of confirmed non-union. Main outcome measures: The primary outcome and end point was the Patient-Rated Wrist Evaluation total score at 52 weeks, with a clinically relevant difference of 6 points. Secondary outcomes included Patient-Rated Wrist Evaluation pain and function subscales, Short Form questionnaire 12-items, bone union, range of movement, grip strength, complications and return to work. Results: The mean age of 439 participants was 33 years; 363 participants were male (83%) and 269 participants had an undisplaced fracture (61%). The primary analysis was on 408 participants with valid Patient-Rated Wrist Evaluation outcome data for at least one post-randomisation time point (surgery, n = 203 of 219; cast, n = 205 of 220). There was no clinically relevant difference in the Patient-Rated Wrist Evaluation total score at 52 weeks: the mean score in the cast group was 14.0 (95% confidence interval 11.3 to 16.6) and in the surgery group was 11.9 (95% confidence interval 9.2 to 14.5), with an adjusted mean difference of -2.1 in favour of surgery (95% confidence interval -5.8 to 1.6; p = 0.27). The non-union rate was low (surgery group, n = 1; cast group, n = 4). Eight participants in the surgery group had a total of 11 reoperations and one participant in the cast group required a reoperation for non-union. The base-case economic analysis at 52 weeks found that surgery cost 1295 per patient more (95% confidence interval 1084 to 1504) than cast treatment. The base-case analysis of a lifetime-extrapolated model confirmed that the cast treatment pathway was more cost-effective. The nested qualitative study identified patients' desire to have a 'sense of recovering', which surgeons should address at the outset. Limitation: There were 17 participants who had initial cast treatment and surgery for confirmed non-union, which in 14 cases was within 6 months from randomisation and in three cases was after 6 months. Three of the four participants in the cast group who had a non-union at 52 weeks were not offered surgery. Conclusions: Adult patients with an undisplaced or minimally displaced scaphoid waist fracture should have cast immobilisation and suspected non-unions immediately confirmed and urgently fixed. Patients should be followed up at 5 years to investigate the effect of partial union, degenerative arthritis, malunion and screw problems on their quality of life.