SGEM#510: Take this Broken Radius and just Cast It.

Reference:  Perry DC, et al. Non-surgical casting versus surgical reduction for children with severely displaced distal radial fractures (the CRAFFT Study): a multicentre, randomised, controlled non-inferiority trial and economic evaluation. Lancet April 2026. Date: May 8, 2026 Dr. Andrew Tagg Guest Skeptic:  Dr. Andrew (Andy) Tagg is an Emergency Physician with a special interest in education and lifelong learning. He is the co-founder and website lead of Don’t Forget the Bubbles. Case: A healthy 7-year-old boy presents to the emergency department (ED) with obvious deformity of the wrist after a fall from playground equipment. X-rays show a severely displaced distal radius fracture, with an associated ulnar fracture. The child is neurovascularly intact. But the wrist looks dramatic. It’s quite bent. The child gazes at his arm, a mix of fear and intrigue. You consult the friendly orthopedics specialist who greets the family and recommends reduction under sedation because “it looks too crooked to leave alone.” You recall that in younger children, some fractures can remodel quite well on their own. The child’s father asks you whether you think the boy really needs a procedure to re-align the bones, or if he can just be placed in a cast.  Background: Distal radius fractures are among the most common fractures in childhood, and severely displaced injuries create one of those classic tensions between what looks bad on an X-ray and what matters to patients over time. Traditional teaching has favored reduction, often under sedation or general anesthesia, to restore anatomy and avoid concerns about deformity, loss of motion, or unhappy families. But pediatric bone is not adult bone. Younger children have substantial remodeling potential, especially near active growth plates, and prior observational studies suggested that even very displaced distal radial fractures can straighten out over time with good function. Many clinicians still feel uneasy leaving these fractures unreduced. The visual deformity can be alarming. Families may equate straight bones with proper healing. Procedural reduction also comes with costs and potential harms: anesthesia, sedation, procedural pain, wound complications, etc. Clinical Question: In children aged 4 to 10 years with severely displaced distal radial fractures, is non-surgical casting non-inferior to surgical reduction for functional recovery? Reference:  Perry DC, et al. Non-surgical casting versus surgical reduction for children with severely displaced distal radial fractures (the CRAFFT Study): a multicentre, randomised, controlled non-inferiority trial and economic evaluation. Lancet April 2026. Population: Children aged 4 to 10 years from 49 UK hospitals with severely displaced distal radial fractures, either metaphyseal or Salter-Harris II, with or without an associated ulnar fracture. Exclusion: Injury >7 days, complex wrist fractures that were open or extending into the joint, additional fractured bones elsewhere, inability to adhere to trial procedures or follow up. Intervention: Non-surgical casting without purposeful manipulation, without sedation or general anesthesia. Comparison: Surgical reduction under general anesthesia or conscious sedation, with fixation permitted at the surgeon's discretion. Outcome: Primary Outcome: Patient Report Outcomes Measurement System (PROMIS) Upper Extremity Score for Children at 3 months. Secondary Outcomes: Pain, health-related quality of life, cosmesis, complications, refracture, unplanned surgery, school absence, parental satisfaction, and cost-effectiveness. Trial: Pragmatic, multicenter, randomized, controlled non-inferiority trial with economic evaluation Authors’ Conclusions: “The CRAFFT trial did not demonstrate non-inferiority of non-surgical casting at 3 months against a conservative margin; however, the observed difference in favour of surgical reduction was small, below thresholds that families considered meaningful, and did not persist beyond early recovery. Surgical reduction was associated with higher costs, early procedural complications, and only a modest improvement in cosmetic appearance, supporting consideration of a cast-first strategy for most children.” Quality Checklist for Randomized Clinical Trials: The study population included or focused on those in the emergency department. Yes The patients were adequately randomized. Yes The randomization process was concealed. Yes The patients were analyzed in the groups to which they were randomized. Yes The study patients were recruited consecutively (i.e. no selection bias). Unsure The patients in both groups were similar with respect to prognostic factors. Yes All participants (patients, clinicians, outcome assessors) were unaware of group allocation. No All groups were treated equally except for the intervention. Unsure Follow-up was complete (i.e. at least 80% for both groups). Yes All patient-important outcomes were considered. Yes The treatment effect was large enough and precise enough to be clinically significant. Yes, Who funded the study? Funded by the National Institute for Health and Care Research Health Technology Assessment programme and supported by the NIHR Oxford Biomedical Research Centre. Financial conflicts of interest. One author declared paid lectures for Smith & Nephew and Arthrex; all other authors declared no competing interests. Results: The trial enrolled 750 children, with 375 randomized to non-surgical casting and 375 to surgical reduction. The median age was 7.9 (IQR 6.5-9.5) years, with 456 (61%) being male. 44% had completely off-ended or displaced fractures. Key Results: In children aged 4 to 10 with severely displaced distal radius fractures, surgery provided a small early functional advantage, but the difference was not clinically meaningful and disappeared in later months. Non-surgical casting was less expensive and had fewer early complications. Primary Outcome: PROMIS Upper Extremity score was: 44.9 in the non-surgical group vs 46.6 in the surgical group Adjusted mean difference was minus 1.64 points [95% CI -2.84 to -0.44], favouring surgery. This confidence interval crossed the prespecified conservative non-inferiority margin of 2.5. Non-surgical casting did not meet formal non-inferiority for the full cohort. Secondary Outcomes: There are several secondary outcomes from this study (check out the paper for more detail). Secondary outcomes mostly told a small early advantage, little long-term difference story. First, in the pre-specified subgroup with completely off-ended fractures, non-surgical casting was non-inferiority. Keep in mind this threshold was a bit wider with a margin of minus 5 points. When they looked at PROMIS scores over time, there was really no functional difference between groups by 6 and 12 months. Surgical patients had more early complications, including pressure injury, wound infection, scarring, and nerve irritation. Cosmesis slightly favored surgery early on, but the gap narrowed over time Non-surgical casting saved about £1665 per patient and had a 100% probability of being cost-effective at standard UK willingness-to-pay thresholds. Statistical vs Clinical Significance This was a non-inferiority trial, but whether something is ‘non-inferior’ depends entirely on where you draw the line. The margin was –2.5 PROMIS points, and the result was minus 1.64 with the 95% confidence interval barely crossing the prespecified margin at -2.84. So technically, it failed non-inferiority. It’s important to keep in mind that, because they had incorporated patient and public involvement across many aspects of the trial, they knew that families said a 5-point difference mattered. They chose a bit more conservative margin of -2.5. This is one of those trials where the statistics say, ‘maybe not’… but the patients say, ‘we don’t care.’ And we care about the POOs, those patient-oriented outcomes, more. Lack of Masking (and why it matters here) One thing worth pausing on is the lack of masking. In this trial, neither clinicians nor families were masked to the treatment, and the primary outcome was a parent-reported function measure using PROMIS. That combination matters because it opens the door to expectation bias, which is a form of observer bias. If your child has gone to theatre, had an anaesthetic, and the bone has been “put back in place,” it’s very natural to feel that something definitive has been done. It feels like proper treatment. On the other hand, if the arm has been put in a cast and left looking a bit bent, that can feel like something has been left unfinished even if it’s entirely appropriate. When we see a small early functional advantage for surgery, it’s worth asking how much of that is true benefit, and how much might be shaped by perception. Because if you’ve just watched your child go through a procedure, you might understandably feel like they’re doing better, regardless of what’s happening at the level of bone healing. There’s a lot of additional information in the supplemental section, but one part includes parental satisfaction scores, and there were really no big differences between the two groups. Selection Bias / Equipoise Problem Another interesting aspect is who made it into the trial. Of the 1,227 children who were eligible, only 750 were randomized. Many families declined to participate, and a notable proportion of clinicians felt there wasn’t enough equipoise to even offer enrollment. Some of these exclusions based on clinician equipoise seem warranted like concern for neurovascular compromise. When it came to the families who declined consent, most of them declined because they had a preferred treatment…which more preferring surgical reduction. ...