1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. References
  8. Appendix A: Adapting the efficiency formula to incorporate costs

While numerous reports have sought ways of improving the efficiency of surgical operating lists, none has defined ‘efficiency’. We describe a formula that defines efficiency as incorporating three elements: maximising utilisation, minimising over-running and minimising cancellations on a list. We applied this formula to hypothetical (but realistic) scenarios, and our formula yielded plausible descriptions of these. We also applied the formula to 16 consecutive elective surgical lists from three gynaecology teams (two at a university hospital and one at a non-university hospital). Again, the formula gave useful insights into problems faced by the teams in improving their performance, and it also guided possible solutions. The formula confirmed that a team that schedules cases according to the predicted durations of the operations listed (i.e. the non-university hospital team) suffered fewer cancellations (median 5% vs 8% and 13%) and fewer list over-runs (6% vs 38% and 50%), and performed considerably more efficiently (90% vs 79% and 72%; p = 0.038) than teams that did not do so (i.e. those from the university hospital). We suggest that surgical list performance is more completely described by our formula for efficiency than it is by other conventional measures such as list utilisation or cancellation rate alone.

As healthcare providers strive for better value-for-money from their investments [1, 2], hospital operating theatres have been identified as potential areas for cost reduction, especially as it has been estimated that ∼ 46% of patients discharged from hospital have undergone surgery [3]. Poor scheduling of operations can result in cancellation of procedures [3, 4]. This is costly both to the patient and to the hospital. Two common causes of cancellation are under-optimisation of the patient's medical condition (which can be effectively addressed by good pre-assessment) [5] and over-running of a surgical list [4].

Not surprisingly, a number of authoritative reports (e.g. from the Association of Anaesthetists of Great Britain and Ireland, AAGBI [6], and the Audit Commission [7–9]) have examined how best to manage operating lists. The recommendations made are broadly similar and include measures such as: effective administrative systems, accurate records for analysis and audit, optimally managing staff time, and good pre-assessment of patients to optimise medical conditions. This list is not exhaustive.

Collectively, these recommendations seem very reasonable but one problem remains: none of these reports offers an explicit definition of ‘efficiency’. In the absence of a clear definition, it may not be known, in quantitative terms, that implementing the recommendations has achieved the desired effect. One reason for the deficiency may be that there are potentially many elements contributing to ‘efficiency’. These include: cancellation rate, under-utilisation, over-running, complication rates, low costs, and perhaps some less tangible measures such as patient satisfaction or teamwork and communication [10].

Any chosen measure of efficiency should ideally have the following attributes: it should be widely applicable and be relatively simple to calculate. The professionals who work in theatre should be able to influence the measure directly through their activity. Thus, crude ‘profitability’ as an index of efficiency may have limited value if it is determined only by re-imbursement rates set elsewhere and not by the activity of the surgical team [11]. In its extensive report [7], the Audit Commission chose to focus heavily on utilisation of the list as a prime measure (i.e. the proportion of scheduled list time used for anaesthesia and surgery). Unfortunately, it is possible to achieve quite impressive rates of utilisation of > 100% simply by over-running lists. Far from being an accomplishment, over-running is a common – if not the main – cause of cancellations on the day of surgery, which is inefficient [4]. Cancellation rate itself therefore might be thought a useful measure, but excellent figures can be achieved by under-utilisation, which again is not efficient [12].

In the US, analysis of operating list management has historically been more established and sophisticated than it has in the UK, perhaps because of the more direct financial incentives involved [13–15]. Recently, Dexter and colleagues were invited by a leading anaesthetic journal to summarise their extensive work in this field [16]. They proposed a simple formula for (in)efficiency [12]:

  • image

The costs in this formula are in the main influenced by list under- or over-runs because these in turn are related to wage costs. However, we have recently criticised this formula because it yields a quantity whose units of measurement are in absolute dollars: efficiency is instead better thought of in terms of a percentage of an input or of an optimum value [17]. Their formula also is biased against larger institutions since here, the overall costs will be higher and so therefore will the absolute magnitude of the sum in the formula. Even if the formula were modified to take into account the size of institution or prevailing level of costs, as we have suggested [17], true costs might be difficult to calculate, especially in those countries where much work is still done according to historical or block contracts rather than as fee-for-service.

In the context of such difficulties, it was our aim to describe a new formula for surgical list efficiency. We also wished to test the applicability of our new formula using hypothetical (albeit realistic) scenarios and real data from elective lists of the same surgical specialty.


  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. References
  8. Appendix A: Adapting the efficiency formula to incorporate costs

Developing and defining the formula

Widdison summarised the optimum aims of the elective operating theatre management succinctly: ‘…[a theatre] is used most efficiently when as much of the time available is utilised, when there are no over-runs and no patients are cancelled’[18]. This, and our own experience, guided the translation of this sentiment into terms of an empirical formula:

  • image

The ‘fraction of scheduled time utilised’ means that if a list scheduled for 8 h finishes in 6 h, this quantity = 0.75 and the ‘fraction of scheduled time over-running’ for this list is zero. The ‘fraction of scheduled time over-running’ means that if a list scheduled for 8 h over-runs by 2 h, this quantity = 0.25, and the fraction of scheduled time utilised for this list = 1. Thus the first two terms operate in a mutually exclusive manner: i.e. a single list cannot be both under- or over-utilised at the same time. The ‘fraction of scheduled operations completed’ means that if four of five of the patients booked onto the list have their operations (i.e. one patient is cancelled), this quantity = 0.80.

The formula theoretically yields a result for efficiency ranging from 0 to 1.0 (or 0–100% if this result is multiplied by 100). The value of 100% is obtained when all booked cases are complete at the scheduled time. The formula can also give ‘credit’ for a list that completes its own booked cases early (e.g. four cases) and accepts and completes extra cases (e.g. a fifth case from another list). Thus a number > 1 in the last term (i.e. the fraction of patients completed = 1.25 for this example) could translate as an efficiency > 100% for that particular list.

Clearly, an efficiency of 100% is the ultimate goal, but in practice is unlikely for every list. An efficiency of 80% is equivalent to one patient in five being cancelled on a list and seems unacceptable. An efficiency of 90% equates to one patient in 10 cancelled and is better. Such considerations lead us to suggest that 85% efficiency is the very minimum that should be attained, with > 90% being highly desirable.

Applying the formula to hypothetical data sets

We applied the formula to a number of hypothetical scenarios based upon our experience as reflecting common occurrences in clinical practice. The aim was to explore the ability of the formula to yield sensible descriptions for events (i.e. under- or over-running or cancellations), and also to assess any limitations.

Applying the formula to real data sets

We then applied the formula to real data sets. We have previously described that in general surgery and urology at a centre, elective surgical lists were predictably over-booked, and so both rates of list over-runs and patient cancellations were high [4]. We wished to assess a different specialty, and we therefore selected data from two gynaecology teams in a university hospital, where (as in our previous report) lists are not planned according to estimated duration of operation, to assess if similar considerations applied. For comparison, we also selected data from one gynaecology team in a non-university hospital whose policy is to plan lists according to estimated times [19].

The university hospital has ∼ 1500 beds, and serves a population of ∼ 250 000. The non-university hospital has ∼ 460 beds and serves a population of ∼ 200 000. The surgical specialties in the two centres are similar except the university hospital, as a tertiary referral centre, also provides cardiac (but not thoracic) surgery, neurosurgery, specialist paediatric services, renal/pancreas (but not liver) transplant surgery, and major head and neck surgery. Both hospitals teach doctors in training and medical students.

We selected 16 consecutive lists from a 3-month period from mid-January to mid-April 2007 (i.e. a ∼ 16-week period that excluded major holiday periods). For each list, we used the theatre computer logs in each hospital to identify the start time (defined as the start of anaesthesia), the end-time (defined as the time of arrival of the last patient in recovery), the number of patients originally booked for surgery and the number cancelled. At the university hospital, the scheduled time for a ‘half-day’ list was 4 h and for a ‘full-day’ list, 8 h (there were no breaks for lunch). At the non-university hospital, the scheduled lists were all half-day lists of 3.5 h. The Audit Commission has noted that ∼ 40% of centres had no formal definition of start times for a scheduled list [7], so to account for the possibility that different lists had different start times, we used the scheduled time as being 3.5, 4 or 8 h as appropriate from the start of anaesthesia for that list.

For each list, we calculated three main variables. The utilisation rate was the total time from the start of anaesthesia to the arrival of the last patient in recovery, divided by the scheduled time for the list. The cancellation rate was the number of patients cancelled, divided by the number booked on that list. The efficiency was calculated using our new formula. For each of the three teams, we combined these variables from individual lists to yield averages for the team. Although it was not our main purpose to compare the teams' performance, for completeness we assessed the differences between the medians for the university and non-university hospitals using the Mann–Whitney U-test. Statistical significance was taken at a level p < 0.05.


  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. References
  8. Appendix A: Adapting the efficiency formula to incorporate costs

Figure 1 shows the common theoretical distributions for efficiency calculated using the formula (i.e. the formula represented graphically). The formula predicts that efficiency will increase with increasing utilisation of the list, reaching a maximum at the scheduled list duration, and then declining with list over-run. Cancelled operations ‘set the envelope’ for the maximum efficiency (i.e. there are ‘isopleths’ set by the cancellation rate).


Figure 1.  Graphical representation of the efficiency formula we present. Efficiency is plotted against the percentage list utilisation (and a parallel x-axis indicates the corresponding actual hours completed for an 8-h list). The percentage of booked operations completed (rates of 100%, 80% and 20% are shown, corresponding to cancellation rates of 0%, 20% and 80%, respectively) form ‘isopleths’ for the relationship.

Download figure to PowerPoint

Applying the formula to hypothetical data sets

If a list finishes on time, with no over-running or cancellations, this yields an efficiency according to our formula of 100%. If a list is cancelled at the last minute and all patients cancelled, the efficiency is 0%. This last statistic might be used when applying the formula to a suite of operating theatres, and this will reduce the mean efficiency of the suite as a whole.

If an 8-h list finishes at 6 h, completing all its cases, the efficiency is 75%. If an 8-h list finishes at 6 h, with four of five patients completed, the efficiency is 60%.

If an 8-h list finishes at 10 h, completing all its cases, the efficiency is 75%. If an 8-h list finishes at 10 h, completing four of five cases, the efficiency is 60%.

A value of 0% efficiency can be achieved if the list over-run is equal to its scheduled time (i.e. if an 8-h list over-runs by 8 h and is completed only at a total of 16 h). If the list over-run exceeds the original scheduled time, the equation yields a ‘negative’ value for efficiency. This is not meaningful when applied to a single list, but where a theatre suite is examined, this might be appropriate as such occasions will properly reduce average efficiency of the suite.

Applying the formula to real data sets

Table 1 indicates summary data for the three gynaecology teams and Fig. 2 shows the graphical presentation of the results. The overall cancellation rates in Teams A and B (from the university hospital) was high, their lists frequently over-ran and the calculated efficiency was correspondingly low. The pattern of distribution of points in Fig. 2 broadly indicates that Team A might have had a higher cancellation rate if list over-runs had been strictly prevented (i.e. the main problem was probably over-booking and over-running). The point distribution for Team B indicates, in contrast, both long over-runs and cancellations. Team C (from the non-university hospital) rarely over-ran, its cancellation rates were very low and calculated efficiency was acceptable. The differences between the medians for the university vs non-university hospital efficiencies were statistically significant (p = 0.038).

Table 1.   Summary of key indicators for surgical list performance for gynaecology Teams A and B (from a university hospital) and Team C (from a non-university hospital). The last two rows show the subgroup analysis for half-day lists from Teams A and B (all data for Team C were from half-day lists). The columns are, in order, as follows: the number of patients booked for surgery (and mean number/list and [range]); the number of patients cancelled (%); the number of operations completed (with the mean/list). The next column, the average duration of operation for the team, was calculated by dividing the hours utilised by the number of operations completed. For the full-day lists, the average duration for full- and half-day lists is shown, with the data for the full-day lists alone in parentheses. The next columns are: the list utilisation; the number of lists under-running (%); the number of lists over-running (%). The final column lists the median efficiency calculated according to our new formula (interquartile range) and [range]. Lists were classed as ‘under-running’ if they finished > 10% before their scheduled end-time (i.e. if they were < 90% utilised). Lists were classed as ‘over-running’ if they finished > 10% after their scheduled end-time (i.e. if they were > 110% utilised).
TeamNo. of patients booked (mean/list) [range]No. of patients cancelled (cancellation rate, %)No. of operations completed (mean/list)Mean duration of operation; min (mean for full-day lists alone)List utilisationNo. of lists under-running; (%)No. of lists over-running (%)Efficiency
Team A92 78570101%4679%
 (5.8)  (8%) (5.3) (83) (92–119%) (25%) (38%) (68–86%)
 [4–9]    [59–141%]   [44–97%]
Team B79106990110%2872%
 (4.9)  (13%) (4.3) (183) (92–119%) (17%) (50%) (59–81%)
 [3–7]    [56–127%]   [56–95%]
Team C57 35470101%3290%
 (3.6)  (5%) (3.4)  (83–110%) (19%) (6%) (60–95%)
 [3–5]    [60–146%]   [55–100%]
Team A35 2335493%3375%
 (half-day) (4.4)  (6%) (4.1)  (79–129%) (38%) (38%) (61–82%)
n = 8 [4–5]    [59–142%]   [44–91%]
Team B36 2344694%2273%
 (half-day) (5.1)  (6%) (4.9)  (58–108%) (29%) (29%) (58–92%)
n = 7 [4–6]    [56–127%]   [56–95%]

Figure 2.  Panels a– c: plots of efficiency vs % list utilisation for Teams A–C, respectively, for the 16 lists analysed for each team. The horizontal line indicates the minimum desirable efficiency of 85%. Also plotted on each panel is the ‘isopleth’ for 0% cancellation (100% patients completed), which enables the lie of the data points to be assessed subjectively. Panel d: boxplot of the data in panels a–c. The horizontal line of each box is the median, the lower and upper borders of the box are the 25th and 75th centiles, respectively, the error bars are the limits of the 10th and 90th centiles, and outlying points are shown.

Download figure to PowerPoint

As Teams A and B had included data from all-day lists but Team C data were confined to half-day lists, we also undertook a subgroup analysis of the half-day lists for the former teams (Table 1). The cancellation rates were closer to the rates of Team C, but it seems that this was only achieved at the expense of over-running (∼ 30–40% of these half-day lists over-run vs just 6% for Team C). Poor list planning was also reflected in a relatively high rate of under-utilisation for the half-day lists of Teams A and B (38% and 29%, respectively). Thus, the overall efficiency remained as low for the half-day lists of Teams A and B as for their full-day lists.

The mean duration of an operation undertaken by the teams suggested some differences in case mix. For their half-day lists, Teams A and B seemed to book quite short operations (Table 1). Team C conducted longer operations within its half-day lists, which were of similar length to the average duration conducted by Teams A and B on their lists as a whole. But when the full-day and half-day lists of these two teams were considered separately, it seemed that Teams A and B booked much longer operations on their full-day lists. In particular, Team B booked quite long cases on its full-day lists (Table 1).


  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. References
  8. Appendix A: Adapting the efficiency formula to incorporate costs

This paper presents a formula for describing the efficiency of an operating list. We have applied it to both hypothetical data and actual lists in a meaningful manner. We suggest that our efficiency measure is less misleading than more common alternative measures such as list utilisation (which can be quite close to 100% yet yield very varied cancellation rates; Table 1) or cancellation rate itself (which can be minimised by systematic over-running; see half-day list data for Teams A and B).

Before we discuss the implications of our new measure for the management of operating lists, it is important to consider the limitations and strengths of our approach.

Strengths and weaknesses of the formula

In some respects, our formula is not at all novel but simply a self-evident truth that reflects the fact that any concept of ‘efficiency’ must inevitably incorporate the factors of list under-running, over-running and cancellations. Despite the ‘truism’ of our formula, we are nonetheless surprised that many previous reports (including those by authoritative bodies such as the AAGBI [6], Audit Commission [7–9], Office for National Statistics [20], and public health bodies elsewhere [21, 22]), who have all examined the concept of surgical list efficiency in considerable detail, have not referred to or described the simple formula we present here.

There are a number of possible deficiencies in the formula, which we discuss and explain below.

The formula gives no specific ‘credit’ for booking a list appropriately (e.g. using known or published average times for operations), yet completing these in a shorter than expected time. This may seem surprising, but McIntosh et al. [12] have explained that, while low utilisation because of inadequate booking of cases is certainly inefficient, low utilisation because of easier-than-expected surgery or anaesthesia (although welcome) should not of itself be considered efficient unless it facilitates additional cases to be undertaken on that list (and thereby prevent cancellations elsewhere or generate extra income). Our formula is consistent with their analysis, as it also gives additional ‘credit’ for extra cases being added onto lists that finish their own scheduled cases early (i.e. these lists will potentially yield efficiencies of > 100%).

Our formula does not address the cost of operations, but simple amendments can incorporate costs into our equation (see Appendix A). A detailed costing analysis was outside the scope of this study, but we predict that those units with highest efficiency according to our basic formula will also turn out to have the lowest costs.

Our formula makes no allowance for case mix, but focuses only on the overall number of cases. So, if longer cases were inherently more unpredictable, then our formula might favour lists involving shorter cases. Team B in particular undertakes long cases on its full-day lists, and this is where its inefficiencies mainly seem to lie (Table 1). However, neither we in our previous paper [4] nor Wright et al. [23] found that this was the case. More subtly, certain types of surgery (e.g. laparoscopic surgery) were more unpredictable, regardless of their absolute duration [4, 23]. It is difficult to establish precisely the extent to which this factor may have influenced our data (e.g. Team B did not frequently undertake long laparoscopic cases), but case mix might conceivably be important in another way: certain cases (regardless of their duration) might carry disproportionate financial incentive, but our formula does not take into account such situations. Conversely, it is arguably a strength that, in according all cases equal weight, our formula is consistent with patient perception. It matters little to the patient whether their scheduled operation is long or short, or more or less variable; the disruption and stress of cancellation are probably the same [24].

Our approach is designed specifically to address elective operating. Emergency cases can impact upon elective lists in a number of ways. An emergency might be added to an elective list and as a result cause cancellation of an elective case; or an elective list might rarely be cancelled in its entirety to facilitate emergencies. As our formula only considers elective cases, these scenarios will result in a reduced calculated efficiency. This may, in fact, be appropriate as it might be an incentive for the institution to manage emergency cases in more appropriate ways (e.g. a dedicated emergency theatre).

Comments on our analysis of the real data sets

Our analysis was simply intended to be an illustration of how the formula might be applied rather than a case-controlled comparison of the performance of surgical lists in different centres. Nonetheless, we perhaps chose our teams fortuitously, as they reflect clear differences with a variety of problems and possible solutions.

The desirable objective of at least 85% efficiency is clearly achievable (Team C, Table 1), an outcome which is perhaps not surprising: a list which is appropriately booked is likely to be more efficient than lists which are not [19].

Graphical analysis indicates many of the data points for Team A lie on the ‘100% isopleth’ but efficiency is lost through frequent over-running (Fig. 2). This indicates that better scheduling (i.e. by using mean duration of operations to plan lists) would be an important measure to improve the efficiency of this team.

Better scheduling is also likely to improve the performance of Team B. However, Fig. 2 suggests that other measures may also be needed. Team B undertakes longer cases on average on its full-day lists, and might consider adopting strategies such as booking the predicted longest case first, which has been shown to reduce over-running and cancellation [3, 19, 25]. The very high cancellation rate suggests that pre-assessment or optimisation of patients before surgery might also need to be improved.

Team C should not be complacent. There are instances of under-utilisation, which may have been due to sporadic cases of patient non-attendance, rather than being indicative of systematic under-booking or cancellation due to medical reasons. This might be addressed by having a ‘standby’ list of patients who might be called in at short notice [26–28]. The average duration of a case in Team C seems quite long for a half-day list: Team C has demonstrated its ability to deliver services efficiently, and these longer cases may be facilitated with less risk of over-run or cancellation by securing an all-day list (if the capacity is needed), which will offer greater flexibility.

There are clearly many differences between university and non-university hospital settings that might account for the differences in outcomes that we report, and it may be that many unidentifiable factors might have influenced our results.

Our sample size may be criticised as being unrepresentative of the respective teams. However, the 16-week sample period represents almost a third of the lists in the year, taking into account Bank Holidays, other holiday periods and compulsory list cancellations due to audit meetings, etc., and we feel this was appropriate.

Comparison with other measures of theatre performance

The Audit Commission has emphasised high utilisation of theatre time as an important objective in itself [7]. This has been reinforced by the Scottish Executive's National Theatres Project definition of (in)efficiency as ‘the level of unutilised hours[of theatre/list time]’[9]. There are, however, a number of potential measures of ‘utilisation’. One is the time a theatre is used as a proportion of the theoretical maximum of 24 h per day. Any value of this measure < 100% represents failure to utilise all possible time available. A second measure is the actual duration of the list(s) as a proportion of the total time scheduled (this is the utilisation measure we used in our formula). Any value < 100% indicates list under-running; any value > 100% indicates over-running. A third measure is time actually spent inducing anaesthesia and operating, as a proportion of the actual duration of the list (i.e. any value < 100% represents time lost because of gaps during the list or a slow turnover of patients). Finally, it is possible to record the time for each case actually spent making incisions and performing surgery, as a proportion of the total duration of the case [29]. This measure focuses purely on the surgeon's activity and excludes the time for anaesthetic induction, time taken to position the patient and awaken the patient at the end of surgery.

The relevance of each of these measures depends on one's perspective. Some surgeons might consider the time spent making incisions as the most relevant, but clearly times taken to prepare and position the patient are an inherent part of the surgical process and cannot reasonably be discounted. Reducing the time taken between operations – turnover time – does not generally affect overall performance, except in rare cases of initially unusually long turnover times [30, 31]. The degree to which theatre suites as a whole are used or mothballed is clearly of interest to strategic planners, but good utilisation of the published list is of more interest locally in organising theatre times, and this is why we have focused on this measure in our formula. The Audit Commission recommended a target utilisation in this respect of > 90%, but Table 1 confirms our suspicion (and that of others [23]) that this can very easily be achieved with over-running, resulting in unacceptably high levels of patient cancellation (and probable additional costs to the service; Appendix A). It is interesting that a mean utilisation of 97% has been suggested to be optimum for performance [32, 33]: Team C, the most efficient by our formula, had exactly this mean utilisation (means for Teams A and B were both 103%).

We also argue that cancellation rate itself is not a sufficiently useful measure of performance in isolation as, again, low rates can be achieved by systematic over-running of lists (see the half-day lists for Teams A and B in Table 1). Published cancellation rates can also be manipulated to yield favourable outcomes. For example, some institutions might only record a ‘cancellation’ if it occurred after the patient entered hospital premises; cancellation by telephone before arrival might go unrecorded [34]. Clearly, such manipulations are unhelpful because the institution might never know where the true problems lie or where improvements might be made. It is a strength of our formula that it is not influenced by such manipulations, as these will only result in lower list utilisation rates, which in turn will be reflected in reduced efficiency (despite low cancellation rates).

Using the formula for operating list management

The formula may be used to inform decision-making on a day-to-day basis. We encountered a urology list consisting of one radical cystectomy and five check cystoscopies. Using our previously published data [4], the predicted time for this (534 min) somewhat exceeded the scheduled capacity of the list (480 min). Unfortunately, the cystectomy took longer than expected so that by the scheduled end-time of the list, two cystoscopies (∼ 66 min of surgery) remained outstanding. If the list had been terminated at that point to prevent over-run, the calculated efficiency would have been just ∼ 66%. We chose instead to complete all cases, thus over-running by 1 h, which achieved an efficiency of 87.5%. On this occasion, our formula indicated it was better to complete the list despite a modest over-run.

On the other hand we encountered a gynaecology list consisting of four major cases, each estimated to take 3 h. The list was clearly over-booked and at ∼ 8 h, the scheduled end-time for the list, only three cases were complete. If the list had been stopped and the last patient cancelled, the efficiency at that point would have been ∼ 75%. At the time, the team opted to continue with the last operation, which finished ∼ 4 h later. The final efficiency therefore was just ∼ 50%. In this instance – and in contrast to the urology list described above – the decision to proceed was probably counter-productive and did not help the efficiency of the institution as a whole, as judged by our formula.

Our formula will better describe the effect of any changes in policy or care pathways so that interventions or changes in practice can be better assessed.

The NHS is currently introducing a system of Payment-By-Results (PBR) for hospitals [35]. In this complex system, central government sets a national standard level of re-imbursement (tariff) for each procedure. In turn, each hospital sets its own cost for that procedure (its reference cost). If the latter exceeds the former, the hospital incurs financial losses unless it can introduce efficiencies to lower its own costs. However, even in those instances where the published reference cost of the hospital is lower than the national tariff, the hospital will only make a profit if its assessment of its reference cost was accurate; if it was flawed, then again losses will be incurred. So under PBR, successful hospitals will be those that (i) know their costs accurately, and (ii) can minimise these costs through real and measurable efficiencies [35]. Because our suggested formula defines ‘efficiency’ explicitly, its use might become instrumental in achieving both these aims.


  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. References
  8. Appendix A: Adapting the efficiency formula to incorporate costs

We present a novel formula that describes surgical operating theatre efficiency. We recommend its wide adoption by the NHS as a key performance indicator. The measure of efficiency yielded by the formula will be especially useful in assessing the true impact of interventions designed to improve performance. We predict that institutions with highest reported efficiency will also be generally those with lowest true overall costs and so in the best financial balance, especially after PBR is introduced. The formula emphasises that the most efficient hospitals or teams are not those that simply undertake the most procedures; the most efficient teams are those that plan their lists to fit within the time that is available to them. The preliminary data we report here suggest that this is indeed the case. The notion is readily amenable to confirmation, if our formula is more widely tested and adopted.


  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. References
  8. Appendix A: Adapting the efficiency formula to incorporate costs
  • 1
    Berwick DM, Measuring NHS. productivity. How much health for the pound, not how many events to the pound. British Medical Journal 2005; 330: 9756.
  • 2
    Cegan PC. The easiest cut: managing elective surgery in the public sector. Medical Journal of Australia 2005; 182: 6056.
  • 3
    Gordon T, Paul S, Lyles A, Fountain J. Surgical unit time utilization review: resource utilization and management implications. Journal of Medical Systems 1988; 12: 16979.
  • 4
    Pandit JJ, Carey A. Estimating the duration of common elective operations: implications for operating list management. Anaesthesia 2006; 61: 76876.
  • 5
    Rai M, Pandit JJ. Day of surgery cancellations after nurse-led preassessment in an elective surgical centre: the first 2 years. Anaesthesia 2003; 58: 6929.
  • 6
    Association of Anaesthetists of Great Britain and Ireland. Theatre Efficiency, Safety, Quality of Care and Optimal Use of Resources. August 2003. London: AAGBI, 2003.
  • 7
    Audit Commission. Operating Theatres. Review of National Findings. London: HMSO, 2003.
  • 8
    Northern Ireland Audit Office. The Use of Operating Theatres in the Northern Ireland Health and Personal Social Services. Belfast: HMSO, 2003.
  • 9
    Scottish Executive. National Theatres Project: Final Report. Edinburgh: HMSO, 2006.
  • 10
    Healey A, Undre S, Vincent CA. Developing observational measures of performance in surgical teams. Quality and Safety in Health Care 2004; 13 (Suppl. 1): i33i40.
  • 11
    Gabel RA. Performance measurement. American Society of Anesthesiologists Newsletter 2003; 67: 56.
  • 12
    McIntosh C, Dexter F, Epstein RH. The impact of service-specific staffing, case scheduling, turnovers, and first-case starts on anaesthesia group and operating room productivity: a tutorial using data from an Australian hospital. Anesthesia and Analgesia 2006; 103: 1499516.
  • 13
    Overdyk FJ, Harvey SC, Fishman RL, Shippey F. Successful strategies for improving operating room efficiency at academic institutions. Anesthesia and Analgesia 1996; 85: 12324.
  • 14
    Strum DP, Vargas LG, May JH, et al. Surgical suite utilization and capacity planning: a minimal cost analysis model. Journal of Medical Systems 1997; 21: 30922.
  • 15
    Dexter F, Macario A. Application of information systems to operating room scheduling. Anesthesiology 1999; 91: 15018.
  • 16
    Shafer S. Case scheduling for dummies. Anesthesia and Analgesia 2006; 103: 13512.
  • 17
    Pandit JJ, Westbury S, Pandit M. Operating room efficiency in the National Health Service. Anesthesia and Analgesia 2007: in press.
  • 18
    Widdison AL. Can we predict when an operating list will finish? Annals of the Royal College of Surgeons (Suppl.) 1995; 77: 3046.
  • 19
    Kelly MG, Eastham A, Bowling GS. Efficient OR scheduling: a study to reduce cancellations. Association of Perioperative Registered Nurses (AORN) Journal 1985; 41: 5657.
  • 20
    Office for National Statistics. Public Service Productivity: Health, October 2004. London: Office for National Statistics, 2004.
  • 21
    New South Wales Health Department. Operating Theatre Management Report. Sydney: Health Department, 2002.
  • 22
    Saskatchewan Provincial Wait List Strategy Team. Surgical Wait List Management: a Strategy for Saskatchewan. Saskatchewan: Health Department of Saskatchewan, 2002.
  • 23
    Wright I, Kooperberg C, Bonar B, Bashein G. Statistical modeling to predict elective surgery time: a comparison with a computer scheduling system and surgeon-provided estimates. Anesthesiology 1996; 85: 123545.
  • 24
    Ivarson B, Kimbald PO, Sjörberg T, Larsson S. Patient reactions to cancelled or postponed heart operations. Journal of Nursing Management 2002; 10: 7581.
  • 25
    Sier D, Tobin P, McGurk C. Scheduling surgical procedures. Journal of the Operational Research Society 1997; 48: 88491.
  • 26
    Lawrentschuk N, Hewitt PM, Pritchard MG. Elective laparoscopic cholecystectomy: Implications of prolonged waiting times for surgery. Australian and New Zealand Journal of Surgery 2003; 73: 8903.
  • 27
    Wu R. Cancellation of operations on the day of surgery at a major Australian referral hospital. Medical Journal of Australia 2005; 183: 551.
  • 28
    Queensland Health. Policy Framework for Elective Surgery Services. Queensland Government, 2005.
  • 29
    Durani P, Seagrave M, Neumann L. The use of theatre time in elective orthopaedic surgery. Annals of the Royal College of Surgeons (Suppl.) 2005; 87: 1702.
  • 30
    Abouleish AE, Hensley SL, Zornow MH, Prough DS. Inclusion of turnover time does not influence identification of surgical services that over- and underutilize allocated block time. Anesthesia and Analgesia 2003; 96: 8138.
  • 31
    Dexter F, Abouleish AE, Epstein RH, Whitten CW, Lubarsky DA. Use of an operating room information system data to predict the impact of reducing turnover times on staffing costs. Anesthesia and Analgesia 2003; 97: 11926.
  • 32
    De Riso B, Cantees K, Watkins WD. The operating room: cost center management in a managed care environment. International Anesthesiology Clinics 1995; 33: 13350.
  • 33
    Weinbroum AA, Ekstein P, Ezri T. Efficiency of the operating room suite. American Journal of Surgery 2003; 185: 24450.
  • 34
    Pitches D, Burls A, Fry-Smith A. How to make a silk purse from a sow's ear – a comprehensive review of strategies to optimise data for corrupt managers and incompetent clinicians. British Medical Journal 2003; 327: 14369.
  • 35
    Audit Commission. Payment by Results. Key Risks and Questions to Consider for Trust and PCT Managers and Non-Executives, June. London: HMSO, 2004.
  • 36
    Calvert N, Hind D, McWilliams R, Davidson A, Beverley CA, Thomas AM. Ultrasound for central venous cannulation. economic evaluation of cost-effectiveness. Anaesthesia 2004; 59: 111620.
  • 37
    McWhinnie DL, Michaels JA, Collin J, Morris PJ. Resource implications of cancelled operations. British Medical Journal 1994; 398: 1389.
  • 38
    Tessler MJ, Kleiman SJ, Huberman MM. A ‘zero tolerance’ for overtime increases surgical per case costs. Canadian Journal of Anaesthesia 1997; 44: 103641.
  • 39
    Tait AR, Voepel-Lewis T, Munro HM, Gutstein HB, Reynolds PI. Cancellation of pediatric outpatient surgery: economic and emotional implications for patients and their families. Journal of Clinical Anesthesia 1997; 9: 2139.

Appendix A: Adapting the efficiency formula to incorporate costs

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. References
  8. Appendix A: Adapting the efficiency formula to incorporate costs

Our formula for efficiency is

Eqn 1: Efficiency formula:

  • image

There is always a base cost to running an operating theatre (estimated by the Audit Commission as £17 400 as the national average) and in addition there is a cost per hour (for staffing, equipment, maintenance, material, drugs, heating, lighting, etc.) [7]. The cost per hour can be offset by performing cases, as these bring income (either directly as in the US [12], or indirectly in terms of block contracts or Payment-By-Results in the United Kingdom [35]). Thus, under-utilisation results in an extra cost to the institution as no income is generated by performing no work. Over-utilisation also incurs extra costs because of staff overtime and unbudgetted costs of additional surgical/anaesthetic materials in the over-run. McIntosh et al. have estimated that over-running is ∼ 1.5–2 times as expensive as under-running [12]. Using these considerations, our Efficiency Formula can be adapted:

Eqn 2; Cost formula:

  • image

where £ is the extra cost per hour of under-utilised time. We have used K£ to estimate the cost of over-running (where constant K is the assumed ratio of cost of over- to under-run), but if the actual cost per hour of over-running is known, a different constant can be used for this second term of the equation. For example, Calvert et al. [36] have estimated that theatres cost ∼ £500 per hour (including staff costs) so that over-running costs ∼ £1000 per hour if the multiplier suggested by McIntosh et al. [12] is correct. A cancelled operation has numerous additional or detrimental knock-on effects on the institution in terms of wasted tests, cross-matched blood, intensive care beds, etc. [37]. It may be possible to estimate these costs (e.g. one study estimated costs to the institution to be ∼ £600–900 per case over and above under- or over-utilisation costs [38]; another study reported that in addition to these costs, patients and families may face costs of £250–500 [39]), or this term may be ignored [12].

It is interesting that this adaptation of our base formula for efficiency to consider costs closely resembles the ‘cost formula’ of McIntosh et al. [12] which focused on overall costs:

Eqn 3: McIntosh et al. Cost formula:

  • image

Note that Eqn 2, our adapted Cost formula, predicts that over-running as a device to minimise cancellations will be an expensive strategy as it will increase overall costs. As McIntosh et al. state: ‘…inefficiency is minimised by minimising the hours of over-utilised operating room time…what matters is…reducing the hours worked late’[12].