Defining efficiency requires more fidelity

We would like to congratulate Pandit and colleagues [1] for their formula which attempts to define the efficiency of surgical operating lists through the incorporation of three main elements: maximising utilisation, minimising over-running and minimising cancellations. They quite rightly suggest that this formula offers a more complete description of list performance than some more conventional measures. However, we believe that it still has significant limitations and will not always accurately represent the efficiency of a complex process. In 2003 our Trust was one of five sites chosen to pilot ‘New Ways of Working in Anaesthesia’, an initiative originally designed to address predicted and unavoidable man-power shortages within the speciality over the next decade [2]. Two Anaesthesia Practitioners (AP) have now been in full time employment with us for a year and during this time we have looked at their impact on theatre utilisation. One of their roles is to facilitate theatre turnround during primary hip arthroplasty lists by allowing for the early sending of patients and the administration of regional anaesthesia prior to the completion of the preceding case. We have seen a significant decrease in the median theatre ‘down time’ (the time between the exit of one patient from the operating theatre to the entry of the next patient) from 23 to 10 min and this has allowed us to book one extra primary hip arthroplasty per list. Clearly such an achievement is of great financial interest to a Foundation Trust which generates its income through Payment by Results. We believe that this reduction in theatre ‘down time’ represents a marked improvement in theatre utilisation and we have demonstrated significant increases in ‘profitability’. Unfortunately, applying the Pandit formula to our model would not credit us with improved theatre efficiency as, assuming we still completed all our booked cases and finish on time, our score could never be more than 100%, a result we could have achieved without an AP and by doing one less case. In addition, if we use an AP without booking extra cases and finish early, unless we can find additional cases to do on an ad-hoc basis, we would be defined as inefficient even though we may have done the same amount of work as an ‘efficient’ list without an AP. In their introduction, Pandit et al. state that: ‘the professionals who work in theatre should be able to influence the measure directly through their activity’. However, in this scenario the actions of the staff would not influence this efficiency measure. Our findings in this subspecialty with both hip and knee primary arthroplasty certainly contrast strongly with the author's statement that ‘reducing the time taken between operations – turnover time – does not generally affect overall performance, except in rare cases of initially unusually long turnover times’. Performing just one additional hip replacement per week equates to an additional 50 patients treated per year and an additional income to a Trust of approximately £325 000 per annum. A further limitation of their formula lies in their decision to define the period of the ‘scheduled list’ as running for a fixed interval (3.5, 4 or 8 h) from the ‘start of anaesthesia’ for the first patient on the list. We have found that several lists in our theatres often begin (start anaesthetic time) more than 15 min after the timetabled start of the session. If the theatre and staff are all prepared to begin a 4 h list at 08.30 h but the patient (or anaesthetist) does not arrive until 09.00 h then, if all cases are completed in 4 h and the session ends at 13.00 h, the list efficiency calculated by applying Pandit et al.’s formula will be 100%. We would suggest that, in reality, this constitutes 87.5% utilisation of the allocated list time (08.30–12.30 h) coupled to a late start and a 30 min overrun. We doubt that the surgeon undertaking the afternoon list in that theatre who is kept waiting because the morning list overran would be happy to see his colleague being credited with a 100% efficient morning list. A ‘real world’ measure of efficiency must take into account the fact that operating sessions need to be held to agreed start and finish times because staff are required to be at work on time and need to be able to leave promptly at the end of their shift. It is also worth noting that defining an ‘overrun’ as a finish time beyond 110% of the scheduled list duration, as Pandit et al. have, means that an 8 h list can continue for up to 48 min beyond its planned finish time before being classified as late running. If list overruns of this magnitude were accepted in our theatres on a regular basis organisational chaos would ensue! While we acknowledge that this paper enhances our understanding of the concept of theatre efficiency we question whether any formula which yields a single value can provide a satisfactory description of the effectiveness of an operating list. We certainly feel that their measure falls short of the requirements for an effective Key Performance Indicator of operating theatre efficiency and would caution against its premature adoption without more rigorous testing in a wider range of clinical scenarios.