Time at the surgical ‘coal-face’ has been reduced by introduction of the European Working Time Directive (EWTD) significantly impacting training opportunity. Our null hypothesis was that duration of surgery is significantly longer if a trainee were performing the operation despite supervision or level of trainee experience.

Cemented hip hemiarthroplasty was chosen as our index procedure as complexity is largely comparable between cases. 461 patients were identified on the hospital trauma database. Data were augmented by information regarding level of surgeon, assistant and time of surgery from the hospital theatre database.

There was no significant difference in registrar and consultant operative times, mean time 69 and 72 minutes respectively. SHOs were significantly slower (mean 80 minutes, p=0.0006). Junior (ST5 or less) registrars were significantly slower (mean 81minutes, p=0.0002) whereas senior registrars were not. Supervision level had no effect on duration of senior registrar operations but when junior registrars were consultant supervised they were not significantly slower (mean 75 minutes, p=0.09).

Supervised operating therefore reduces time variability and should be promoted within a climate of training. Increase in mean operative time in registrars and SHOs is insignificant within a day's operating and is unlikely to lead to cancellations of cases.

A number of studies suggest revision of unicompartmental knee replacement (UKR) to total knee replacement (TKR) is straightforward. We hypothesise that this is not always the case in terms of complexity, cost and clinical outcome.

We identified 23 consecutive patients revised from UKR to TKR by 2 consultant surgeons (2005–2008). These were matched by age, sex and comorbidity to a cohort of primary TKRs (42 patients) performed during the same period. Data were collected regarding demographics, cost (surgical time & implants) and 1 & 5-year follow-up of clinical outcome (OKS) and outpatients attended.

There was no statistically significant difference in cost of implants for revision UKR to TKR vs. primary TKR (p=0.08), however operative time was significantly higher in the revision group. One year mean OKS was significantly higher in the primary TKR group (mean 30 vs. 23 p=0.03), but 5-year follow up showed no significant difference (mean OKS 27 vs. 32 p=0.20). The revision group had statistically significantly greater number of follow-up appointments (mean 6 Vs. 2 p<0.0001).

Revision of UKR to TKR is not a universally straightforward procedure, carrying significant overall cost implications. Clinical outcomes, although significantly different at 1 year are almost the same at 5 years.

In previously published work, MR arthrogram (MRA) has sensitivities and specificities of 88–100% and 89–93% respectively in detection of glenoid labrum tears. Our practice suggested higher frequency of falsely negative reports. We aimed to assess accuracy of this costly modality in practice.

We retrospectively reviewed MRA reports of 90 consecutive patients with clinical shoulder instability who had undergone arthroscopy. All had history of traumatic anterior dislocation and had positive anterior apprehension tests. All underwent stabilisation during the same procedure. We compared the findings, using arthoscopy as gold standard in identification of glenoid labral tears.

83/90 patients had glenoid labrum tears at arthroscopy. Only 54 were correctly identified at MRA. All normal labra were identified. This gave sensitivity of 65% and specificity of 100% in identification of all types of glenoid labrum tear. The majority had anterior glenoid labral tears, which were detected at an even lower rate of sensitivity (58%).

Sensitivity of MRA in this series is significantly lower than previously published. This study highlights the importance of an accurate history and clinical examination by specialist shoulder surgeons in the management of glenohumeral instability. The need for this costly investigation may not be as high as is currently the case.