You are working the ambulatory section of your ED and evaluating a 6-year boy with a possible distal radius injury. He lives with autism spectrum disorder and is uncomfortable in X-ray rooms, and you wonder if there is a role for PoCUS to help exclude a fracture.
Ultrasonography or Radiography for Suspected Pediatric Distal Forearm Fractures
Snelling PJ, Jones P, Bade D, Bindra R, Byrnes J, Davison M, George S, Moore M, Keijzers G, Ware RS. Ultrasonography or Radiography for Suspected Pediatric Distal Forearm Fractures. New England Journal of Medicine. 2023 Jun 1;388(22):2049-57
How was the study designed?
Patients: 5-15 years old with suspected distal forearm injury.
- They excluded patients with obvious deformities, open injuries, multiple injuries, congenital bone disorders, neurovascular compromise, or suspicion of non-accidental trauma.
Intervention: PoCUS performed by a trained HCP (not just a physician)
- The training was 2-hour didactic, 1 training session, and 20 proctored PoCUS, with at least 10 positive scans, X-rays were permitted in the following situations
- Identification of a cortical breach fracture (apart from an isolated ulna styloid fracture
- Fracture <1cm from physis
- Pronator quadratus hematoma sign present
- Angulation greater than ~5 degrees
- Physis widened or narrowed
- Periosteal hematoma
- High clinical suspicion
Comparison: standard of care X-rays
Outcome: Physical function of the injured upper limb at 4 weeks (28 ± 3-days), as reported by a standardized survey tool
This was a non-inferiority RCT study with a clinical non-inferiority margin of 5 points on the outcome scale. This was a non-blinded study.
270 children were randomized, and 266 completed the follow-up. The difference in function scores was 0.1 point (95% CI, −1.3 to 1.4) between US and X-rays. The lower boundary of the 95% confidence interval (-1.3) was higher than the noninferiority margin of −5 points that the authors pre-specified, so they concluded that PoCUS was non-inferior to the X-ray.
Critical appraisal of “Ultrasonography or Radiography for Suspected Pediatric Distal Forearm Fractures”
The study was well designed, they met their intended sample size, and the groups were well balanced, except for a higher percentage of males in the X-ray group.
The primary outcome was obtained by expert consensus. It reflected other published studies indicating that the PROMIS outcome score’s minimally significant clinical difference is 5 for distal radius injuries. It might have been helpful to engage patient representatives to verify this.
The most glaring issue is the use of expert panel determination of the final outcome. Specifically, they used the intervention of interest (PoCUS) to determine the final outcome. Using the results of the reference test while already knowing the results of the intervention test can lead to overestimating the intervention test’s accuracy, especially when subjective measurements are used. Not everyone in the PoCUS group also received the standard reference X-ray, although the primary functional outcome (not radiological outcome) makes this less important.
Bottom line on “Ultrasonography or Radiography for Suspected Pediatric Distal Forearm Fractures”
As you are not comfortable doing PoCUS for forearm injuries just yet, you spend some time to help make the patient comfortable in the X-ray room. However, once you are trained in this procedure, it might be part of your diagnostic toolbox for minor distal forearm injuries. It may also be helpful in resource-limited settings where X-rays are not readily available.
Expert clinical commentary with Dr. Arun Sayal on “Ultrasonography or Radiography for Suspected Pediatric Distal Forearm Fractures”
While the notion of using PoCUS in a resource limited setting to diagnose distal radius fractures in children is appealing, if X-ray is available, there is limited utility in using PoCUS. There are a multitude of important clinical questions I have that this study did not fully address: What proportion of patients had PoCUS only vs PoCUS plus X-ray in the ED? What proportion had an X-ray performed at follow up confirming or refuting the diagnosis? What about DRUJ injuries (dismissing isolated ulnar styloid fractures as not needing an X-ray that PoCUS may miss may result in missing a serious DRUJ injury)? Did they account for teenage patients with closed or near-closed growth plates in which even a few degrees of fracture angulation would be clinically significant? How did they measure angulation on US? Was the distal fragment angulated dorsally or volarly (management should be tailored accordingly)? How did they assess if the fracture was bicortical? Did they determine whether the physis was wide or narrow on PoCUS (which is difficult to discern, especially for junior PoCUS operators)?
Given all these unanswered questions and the likelihood that our orthopedic colleagues (who may be assessing some of the patients in follow up) would not be supportive of a PoCUS only approach as a way of diagnosing distal radius fractures, I think it will be a very long time before PoCUS replaces XR as the primary modality for diagnosing distal radius fractures in children.
Research Methodology Hot Take with Dr. Shelley McLeod on “Ultrasonography or Radiography for Suspected Pediatric Distal Forearm Fractures”: Understanding non-inferiority trials
From a research methodology perspective, this trial was well done. They used a variety of health care practitioners (nurse practitioners, physiotherapists, emergency physicians), ultrasound machines, and the trial took place in various hospital ED settings, so there is good external generalizability. The non-inferiority margin represents the maximum clinically acceptable difference between the two groups, and I think the authors did a good job describing how the non-inferiority margin of 5 points was chosen by experts from the BUCKLED trial group and review of the literature, but I agree with Dr. Mohindra that patients and their loved ones should have been involved in this decision. The inclusion and exclusion criteria seem appropriate, and both per-protocol and intention-to-treat analyses are commonly presented.
The primary outcome was physical function of the arm at 4 weeks, which is an important patient-centered outcome. The only concern I had was the lack of control for possible confounding. Clearly, differences in ED and hospital management in terms of subsequent therapeutic interventions after the initial diagnostic method may have had a huge influence in the subjective, patient reported primary outcome.
Drs Mohindra, Sayal and Mcleod have no conflicts of interest to declare