The OMENS trial: opportunistic evaluation of musculo-skeletal physician care among orthopaedic outpatients unlikely to require surgery.

STUDY OBJECTIVE For orthopaedic outpatients judged as unlikely to require surgery, how does the cost-effectiveness of treatment from a musculo-skeletal medicine physician compare with that of a conventional orthopaedic surgeon-led service? DESIGN Partially randomised, pragmatic trial comparing management from a single musculo-skeletal medicine physician with orthopaedic surgeon-led management, with the timing and nature of treatments at the discretion of individual clinicians. Main outcomes were: change in patient reported health (SF-36 and EuroQol), and marginal health service costs. SETTING Outpatient department of an Edinburgh orthopaedic hospital. SUBJECTS One thousand three hundred and twenty six people; aged over 18, referred to and later attending the outpatient department with a 'non-surgical' musculo-skeletal condition, who provided baseline information. Eight hundred and twenty nine were randomly allocated to management by either the musculo-skeletal medicine physician (medical group) or a conventional orthopaedic surgeon-led service (surgical group); 497 were assigned to the treatment groups according to their general practitioner's preference. RESULTS There are aspects of the study design concerning statistical power, potential selection bias and generalisability, which mean that any results must be treated with caution. Although there was no statistically significant difference in health gain between the randomly allocated groups (with similar findings in the non-randomised medical and surgical groups), the patient groups were heterogeneous and the confidence intervals did not rule out clinically important differences. The medical group received more outpatient treatment, 20% vs. 10% (difference 95% CI 5% to 15%) with a higher rate of physiotherapy referral, 45% vs. 30% (15%, 9% to 22%). The surgical group received more inpatient surgical care 6% vs. 17% (-11%, -15% to -7%). Mean healthcare costs in the medical group were 179 Pounds per patient compared with 287 Pounds in the surgical group (-108 Pounds, -25 Pounds to -191 Pounds). CONCLUSIONS Interpretation should take into account the practical difficulties and circumstances of the study. Although no differences in health gain were found between these two management policies for 'non-surgical' orthopaedic outpatients, the trial was less statistically powerful than planned and the confidence intervals included clinically important differences. Differences in treatment emphasis gave estimated costs savings to the Health Service of 108 Pounds per patient treated by the musculo-skeletal medicine service.