- Paul Glasziou, professor of evidence based medicine
- Centre for Evidence-based Medicine, Department of Primary Health Care, University of Oxford, Oxford OX3 7LF
As I was about to start my morning general practice clinic, the receptionist told me of a patient who was coming in with an injury and handed me a faxed report from radiology, which stated: “possible fracture of the radial head.” The “possible” suggested it was undisplaced or minimally displaced, so I wondered if it needed treating at all. While waiting for the first patient to arrive and sipping my tea, I checked the orthopaedics texts in my room and did a PubMed search.
Searching for the evidence
I went to the Clinical Queries section of PubMed Central (which is bookmarked on my Firefox toolbar) and used the narrow version of the “therapy” filter (which filters for randomised trials). I entered search terms to describe the condition “fracture and radial and head,” which brought up seven studies. Two of these studies were not trials and three were not relevant (two looked at different types of internal fixation, and one looked at different methods of reduction), which left two that were relevant. I used the most recent study (2002)1 because it was more relevant to this patient's problem and I had access to the full text. I had access only to the abstract of the second trial,2 but this seemed to be consistent with the findings of the first trial. My search took only a few minutes.
Assessing the evidence
The trial1 randomised 60 patients into two groups. One group was treated with rest in a broad sling for five days followed by mobilisation, and the other was treated with immediate mobilisation and an exercise programme that started 24 hours after injury. The randomisation process was haphazard rather than randomised, but baseline characteristics were not significantly different between the two groups (although the immediate mobilisation group seemed to have had more minor fractures). Follow-up assessments were done by an independent observer who was blinded to treatment, but the authors did not mention loss to follow-up. As is usual, the trial was not perfect, but none of the flaws seemed sufficient to invalidate the results.
All fractures in both groups united without problems. At the end of the first week, the early mobilisation group had slightly more pain (10 v 6 on a 25 point scale) but better range of motion (flexion of 112° v 98°). However, most measures showed no difference at that time or at further follow-up (four and 12 weeks).
Outcome for the patient
When I saw the patient, it was clear that she had tenderness over the radial head but still had a full—though somewhat painful—range of elbow movement. I explained the choices to her—that immobilisation would help a little with the pain but make no long term difference, and it would mean that temporarily her arm would be a little stiffer. She opted for no immobilisation with a simple bandage for comfort.
Of course, I could have called the local orthopaedic registrar, and that is often the wise thing to do. But once I had seen the patient I was sure that this was just a “minor” fracture that could be dealt with in primary care.
Clinical training and practice is replete with heuristics—rules of thumb—in both diagnosis and treatment. For example, “beware the unilateral red eye,” “if a child fails to speak by 16 months then assume that the child is deaf until proved otherwise,” and “where there is pus, let it out.” One treatment rule we use for soft tissue injury is RICE (rest, ice, compression, and elevation). However, is this rule based on good research? Recently some colleagues and I questioned this advice, and did a systematic review of whether immobilisation is best for the treatment of limb injuries. We identified 49 trials, and we found that less rest and more mobility seemed to be better, even in the 14 studies of fractured limbs.3 Subsequently, a trial showed that short arm plasters were preferable to long arm plasters for forearm fractures in children. Some people have suggested we change the rule from RICE to MICE (mobilisation, ice, compression, elevation). However, applying this general principle to individual patients and injuries is not always straightforward.
I was puzzled about exactly when I can apply the MICE heuristic to injuries, but a helpful reviewer suggested that, “The majority of hand fractures and radial head fractures are now treated by mobilisation. Many minor lower limb fractures are similarly treated. In many distal radial fractures and most proximal humeral and clavicle fractures, immobilisation is simply for pain relief.” With that clarified, I am now left wondering how many other mnemonics provide misleading advice? Indeed, a colleague—Carl Heneghan—pointed out that the C of RICE is also questionable, with a trial in ankle sprains showing that compression bandages provide no advantage and lead to a greater use of analgesia.4
We welcome contributions of evidence based case reports. These reports should describe a clinical dilemma raised by a real patient and show how evidence can be applied at all stages of patient care
Thanks to the reviewers for helpful comments, in particular, CM Court-Brown for the advice in the final paragraph.
Competing interests: None declared.
Provenance and peer review: Not commissioned; externally peer reviewed.