Summer holidays have just begun, and you’re working in the emergency department when 6-year-old Anisha is brought in by her concerned parents.
She’s fallen off the climbing frame and has a noticeably deformed right wrist.
After ensuring some good analgesia, you review her X-ray, and she has a significantly displaced distal radius and ulna fracture. Anisha’s mum wants to know whether she’ll need an operation.
You know that the FORCE trial showed that buckle fractures heal well within intervention, but do severely displaced radius fractures need operative manipulation or fixation?
Luckily, you have just read the recent paper on the CRAFFT trial.
Perry, D.C., Zimmermann, A., Achten, J., Nicolaou, N., Metcalfe, D., Kounali, D., Lyttle, M.D., Frempong, S., Appelbe, D., Mason, J. and Kandiyali, R., 2026. 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. The Lancet.
Distal radius fractures account for approximately half of all fractures in children. Severely displaced radial fractures look dramatic on X-ray and have traditionally been managed by operative manipulation and sometimes fixation with plates or K-wires.
In recent years, emerging evidence has demonstrated the huge potential for bone remodelling in paediatric populations and has raised the question: ‘Do we need to reduce these fractures?’
What did we know?
Previous observational studies have demonstrated that even completely displaced distal radius fractures can heal well over time without reduction, with good restoration of function and no long-term deformity. But before the CRAFFT trial, there were no previous randomised controlled trials comparing surgical reduction to non-surgical casting.
Who were the patients?
The CRAFFT study was conducted across 49 hospitals in the UK. Patients were recruited between August 2020 and May 2024.
The children participating were between 4 and 10 years old and had severely displaced distal radial fractures (metaphyseal or Salter–Harris II), with or without an associated ulnar fracture. They were eligible for the trial if the treating clinician considered that surgical reduction would usually be performed.
Of 1227 eligible patients, 750 participants were randomised to either non-surgical casting or surgical reduction. 477 children were excluded (54 met the exclusion criteria, and 423 were excluded due to parental choice or lack of clinical equipoise).
Exclusion criteria included injuries more than 7 days old, complex wrist fractures, additional fractures elsewhere or factors that would make trial adherence or follow-up difficult, such as a language barrier.
456 (61%) participants were male, 294 (39%) were female, and the median age of participants was 7·9 years (IQR 6·5–9·5). 329 (44%) of the 750 participants had completely off-ended fractures.
What was the intervention?
The study compared surgical reduction with non-surgical casting without purposeful manipulation of the fracture position.
Surgical reduction included fixation, at the surgeon’s discretion. Non-surgical casting can be below- or above-elbow.
750 participants were randomly assigned, 375 to the non-surgical casting group and 375 to the surgical reduction group.
The study was a non-inferiority study. This type of trial is designed to show that the intervention in question (non-surgical casting) is not unacceptably worse than surgical reduction. Meaning ‘is casting almost as good as surgical reduction?’, with acceptable pre-defined limits.
What were the outcomes measured?
The primary outcome for this study was functional recovery, assessed using the Patient-Reported Outcomes Measurement Information System (PROMIS) Upper Extremity Score for Children Computer Adaptive Test at 3 months post-randomisation.
Lower scores represented a worse outcome. Discussion with parents during study design led to a PROMIS score difference of 5 being the minimum required to demonstrate significance. A conservative non-inferiority margin of -2.5 was used.
Secondary outcomes included:
- Pain measured using the Wong–Baker FACES Pain Rating Scale.
- Health-related quality of life was measured using EQ-5D-Y.
- Complications including loss of reduction, infection, re-fracture, unplanned surgery, and adverse events related to the cast or surgery.
- Details of treatments received and interventions related to any complications.
- Parental work absence, out-of-pocket costs, and school absence.
- Cosmesis was assessed using a parent-reported VAS.
What were the results?
The study did not demonstrate non-inferiority of non-surgical casting – the primary outcome measure.
At three months, the mean PROMIS Upper Extremity score was 44·9 (SD 8·7) in the non-surgical casting group and 46·6 (SD 8·8) in the surgical reduction group. The adjusted mean difference was –1·64 (95% CI –2·84 to –0·44; p=0·0073), favouring surgical reduction. However, as the difference was less than 5, the authors deemed it small enough to fall below the threshold parents would consider sufficient to justify surgical intervention.
However, at 6 months and 12 months, there was no statistically significant functional difference between the groups (adjusted difference -0.47 (95% CI -1.67 to 0.73; P=0.44) and -0.41 (95% CI -1.52 to 0.70; P=0.47), respectively). This demonstrated no statistically significant difference in long-term recovery between the groups.
Even in children with completely off-ended (displaced) fractures, non-surgical casting was non-inferior to surgical reduction (adjusted mean difference -1.39 (95% CI -2.86 to 0.09; p=0.066).
Secondary outcomes showed a small (non-clinically significant) favour in terms of pain scores for the surgical reduction group, yet an increase in the number of complications, including pressure damage, wound infections, scarring, and nerve irritation.
Cost-wise, non-surgical casting demonstrated a significant mean per-patient saving of £1665 (95% CI £1487 to £1843).
School absence was slightly worse in the surgical reduction group (mean 3·6 days) than in the non-surgical casting group (mean 3·0 days).
At 12 months, cosmesis was not of meaningful significance. The families considered recovery complete regardless of appearance. The non-surgical casting group valued not having any scars, while it was sometimes brought up for cosmetic reasons for the surgical reduction group. No child had revision surgery for cosmesis.
How good was the paper? – CASP checklist for Randomised Controlled Trials.
Did the study address a clearly formulated research question?
Yes.
Was the assignment of participants to interventions randomised?
Yes
Randomisation was carried out, overseen by the local researcher in the ED or Orthopaedic Clinic. Participants were randomly assigned 1:1 using a secure web-based system.
Were all participants who entered the study accounted for at its conclusion?
Yes
Were the participants blinded to the intervention?
No, neither the participants nor the treating clinicians were blind to the intervention. This is acknowledged as a limitation, but it is common in surgical trials. The analysis was blinded.
Were the study groups similar at the start of the trial?Yes
Stratification was used to ensure balanced allocation across treatment groups.
- Age 4–6 years vs 7–10 years
- Fracture location (metaphyseal vs physeal)
- Displacement severity (completely off-ended vs not completely off-ended).
Apart from the intervention, did the groups receive the same care?
Yes, regarding the follow-up schedule. Some specifics of treatment were left to the local centres, including whether to use general anaesthesia or sedation for surgical reduction, the choice of cast type, and follow-up procedures.
Were the effects of the intervention reported comprehensively?
Yes. The study was adequately statistically powered.
The primary and secondary outcome measures are clearly outlined and documented, with appropriate statistics and reporting.
Was the precision of the estimate of the intervention or treatment effect reported?
Yes. There was a very strict non-inferiority margin measured at 3 months for the primary outcome (-2.5 PROMIS points). Confidence intervals crossed this margin.
Do the benefits of the experimental intervention outweigh the harms and costs?
The further from randomisation, the smaller the difference between the two groups in the measured outcomes. The study reports that recruitment for follow-up is complete for a cohort of patients, extending to 3 years post-enrolment.
Whilst surgical reduction demonstrated superior functional scores at 3 months compared to non-surgical casting, the difference was small enough to be of no clinical significance to parents and carers and resolved in the long term. The use of non-surgical outcomes demonstrated significant cost savings.
Can the results be applied to your local population/in your context?
As a UK-based study, these results will be most applicable to clinicians working in the UK or similar settings. They may be less applicable to physicians working in different healthcare settings, particularly regarding the economic findings.
Would the experimental intervention provide greater value to the people in your care than any of the existing interventions?
The study demonstrated that non-surgical casting can help avoid some complications associated with surgical casting. The transient functional and cosmetic differences were small enough that parents did not deem them significant.
Anisha had a cast applied and went home later that afternoon. Follow-up X-rays demonstrated excellent remodelling.
Will this change my practice?
This study supports the potential for effective remodelling in paediatric bony injuries, and the associated options for less invasive treatment with similar functional outcomes and fewer complications.
The authors believe this is a move towards more conservative treatment of paediatric fractures, in keeping with previous (and possibly future!) clinical trial outcomes.
References
Closed treatment of overriding distal radial fractures without reduction in children – PubMed
Remodeling of Salter-Harris type II epiphyseal plate injury of the distal radius – PubMed
