Problem

Our unit is a large district general hospital with approximately 5500 deliveries annually. The caesarean section rate in 2016 was 27.2% and 11% of all deliveries were performed as planned elective caesarean sections (ELCS).

There are two maternity operating theatres available with separate theatre staff teams for elective and emergency obstetric cases. ELCS lists are scheduled every morning from 0800 to 1300 hour on Mondays to Thursday and on alternate Fridays. There are two consultant-delivered lists per week and the remainder are performed by junior surgeons operating independently on low or moderate risk cases. After 1300 hours, there is one operating theatre available for emergencies. Each ELCS list has three slots which are booked in antenatal clinic using a master Excel spreadsheet accessed by antenatal clinic midwives. Patient details are entered into the spreadsheet, along with any identified risk factors and the indication for caesarean section.

We identified two main problems with ELCS lists at our Trust. First, the existing ELCS booking process did not involve any selection of what cases were allocated to a particular list. This resulted in unselected ELCS lists, often leading to inappropriately complex case mix on lists carried out by more junior surgeons. Second, unselected ELCS lists of high-complexity cases could over-run beyond the allocated time for list. This resulted in the third ELCS case either being cancelled or performed after 1300 hours in the emergency maternity theatre, creating a clinical risk of a simultaneous obstetric emergency without a free operating theatre.

Our objective was to improve the ELCS booking process by using a novel risk scoring system, ELECTIVIST, to reduce the number of lists with an inappropriately complex case mix. The overall aim was to use existing theatre capacity in a more intelligent way, reduce the need for extra ELCS lists and reduce list over-run, particularly that which impacted on emergency obstetric theatre use. There was no existing information on our case mix complexity and theatre over-run so our objective was to audit this data before and after implementation to assess the impact of ELECTIVIST over a 6-month intervention time period.

Background

Caesarean section is one of the most commonly performed operations worldwide. In the UK in 2016–2017, 27.8% of women delivered by caesarean section with 15.5% of women having an ELCS.1 Most ELCS are performed between 39 and 40 weeks of gestation and routine elective surgery booking procedures are therefore unsuitable. ELCS cases vary widely in surgical complexity but there are little published data regarding the impact of this on surgical time and theatre usage.

One study showed that operative times ranged from 13 to 108 min and identified previous Caesarean deliveries, increased scar tissue, tubal ligation and surgical experience as factors that increased ELCS operative times. In particular, the number of previous caesarean deliveries increased the mean operative time linearly from 39.5 min in subjects with no previous caesarean deliveries to 52.8 min in subjects with three or four previous caesarean deliveries.2 Another study showed that total operative time was significantly longer among obese patients compared with women with normal BMI with prolonged operation time of over 60 min associated with a history of previous caesarean sections, grade of the surgeon and higher estimated blood loss. The authors concluded that there were multiple factors influencing ELCS operative time but some of these factors related to patient characteristics were identifiable early during the preoperative planning period and could be used to plan an appropriate theatre list size. They hypothesised that this could impact ELCS list performance and improve theatre safety and efficiency, while reducing costs and maximising training opportunities.3

The Obstetric Anaesthetists’ Association and Association of Anaesthetists of Great Britain & Ireland (OAA/AAGBI) guidelines for obstetric anaesthetic services recommend that elective caesarean sections should neither interfere with delivery unit work nor be interrupted by emergencies and that such interruptions and delays (to both the scheduled and the emergency workload) should be audited and monitored with necessary action taken to correct recurrence.4 However, ELCS list theatre over-run appears to be a common problem for many maternity units with several presented audits at various units describing common themes.

A UK national survey of ELCS availability in 2008 showed that many units are struggling to meet the OAA/AAGBI recommendations regarding elective surgery provision with 37% of units frequently performing ELCS outside normal working hours if not done on the scheduled day list and 21% units cancelling one or more ELCS cases per week. This survey reported ‘unit activity’ as the main perceived problem with ELCS list efficiency in 63% of respondents, followed by ‘organisational’ factors at 47%.5 One audit of ELCS lists in an Australian maternity unit demonstrated other factors contributing to theatre efficiency. Their findings showed that the surgical component of the operating list was 54% of the total theatre time, with 12% on anaesthetic time and 34% on non-surgical and non-anaesthetic activities designated as preparation time. They also found that operator seniority impacted on total case time and cases with a junior operator were more likely to be delayed, with an average over-run of 37 min for an ELCS list performed by a junior operator. The recommendations from these findings suggested improving case mix planning and turnover times between cases.6

Despite evidence that ELCS list over-run is a widespread concern for many units, there is a paucity of literature regarding effective ELCS booking systems. A recent quality improvement project presented a four-point categorisation tool which ranked caesareans based on their complexity and calculated the average amount of time required for each category from A to D. This information was then passed to the booking and scheduling department who could then allocate the appropriate amount of time to each case based on the degree of complexity and likely time needed to complete the procedure. This system reduced over-run from 80% to 40% of ELCS lists and demonstrates valuable proof of efficiency improvement when ELCS booking is improved.7

Baseline measurement

We designed a novel ELCS risk scoring system, ELECTIVIST, comprising a risk assessment by the booking clinician who then allocated a complexity score based on this to each case on a scale from 1 to 6. We designated the maximum risk score for any one patient as 6 and set a standard of a maximum total score on any one ELCS list of 6 and a maximum of three cases per list. Our baseline measurement was a retrospective analysis of all ELCS operations performed in 2016 at our unit.

We collected data on the risk scores for each individual case and the total score per ELCS list from the existing booking Excel spreadsheet. This measurement was to assess our current case mix complexity and to see how many lists had a a score of >6 and were therefore overbooked. We collected data on the surgical and theatre time from hospital records to assess whether the risk score correlated with surgical operating time, total theatre time and incidence of over-running with impact on emergency theatre capacity. We defined theatre over-run as surgical time finishing past 1300 hours and theatre over-run impacting on emergency theatres as an ELCS case performed in the emergency maternity theatre after 1300 hours. We also collected data on the number of extra ELCS lists performed to accommodate capacity in 2016. We then performed a comparative analysis of our actual case complexity and theatre capacity to assess the monthly and annual impact of ELECTIVIST on ELCS capacity prior to implementation. We planned to repeat this analysis at 6 months to evaluate the impact of the new booking form on case mix complexity per list, total theatre time and incidence of over-running.

The results of this baseline measurement are described in figure 1.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

(A) ELCS cases and lists subdivided by ELECTIVIST risk score with average surgical and theatre times. (B) Number of ELCS lists with total ELCS list risk scores (C) ELCS lists subdivided into underbooked (score <6), appropriate (score=6), overbooked (score >6), empty and extra lists. (D) Total risk score of all ELCS cases performed in 2016 and total risk score of list capacity in 2016. (E) Total risk score of all ELCS cases performed in 2016 and total risk score of list capacity in 2016 subdivided into months. ELCS, elective caesarean section.

The baseline measurement showed that only 16% of ELCS lists were of appropriate complexity with 32% of lists overbooked with inappropriately complex case mix. Theatre over-run occurred in 21% of ELCS lists and theatre over-run which impacted on emergency theatres occurred in 6% of ELCS lists. There was widespread variation in the case mix complexity per ELCS list, from 1 to 11 (figure 1B). In 2016, the number of extra ELCS lists performed was 8 (figure 1C). This baseline measurement also showed that ELCS lists with a total risk score of >6 had a significantly higher chance of theatre over-run (relative risk (RR): 2.2) and an even higher chance of over-running with impact on emergency theatres (RR: 2.9). The capacity analysis showed that annually case complexity scores were exceeded by capacity (figure 1D) and that complexity per month exceeded capacity in 2 months. This suggested that annually no additional capacity would be required in order to accommodate the complexity of ELCS cases.

Design and implementation

PDSA 1: design of the ELECTIVIST booking form

The project team initially comprised an specialty trainee and two consultants in obstetrics and gynaecology with personal clinical experience of the problems caused by the current ELCS booking system. We discussed the issues contributing to ELCS list theatre over-run and decided to focus on the booking process as the factor which would be most easily improved by application of a systematic approach.

We designed a draft booking form, listing all common potential risk factors that could affect surgical complexity based on our clinical experience. Many of these were generic risk factors but some were specific to our maternity population. For example, Buckinghamshire Healthcare NHS Trust is the site of the National Spinal Injuries Centre and pregnant patients with spinal cord injuries often receive their obstetric care in our unit. These comprise a high-risk population with complex anaesthetic considerations for elective caesarean delivery but these would not be widely applicable to other units. We considered two options for the risk scoring system. Initially, we considered a Red/Amber/Green rating with each list designated to have a maximum combination of low, moderate and high risk cases. However, we felt that this could be confusing for users and therefore we sought to create a numerical system which was easier to follow.

We wanted to allocate a complexity score to each patient while simultaneously capturing all potential risk factors in the booking process. We therefore designated a scoring system from 1 to 6 with low-risk factors scoring 1 and the highest risk factors scoring 6. These are summarised in figure 2.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

ELECTIVIST risk factors and scoring system. BMI, body mass index; CS, caesarean section.

Having stratified all risk factors into categories, the final score for each case was determined by the highest scoring risk factor and not by adding the risk factor scores together. For example, a patient with fetal breech presentation with a BMI of 45 would score 1 for breech but 4 for BMI of >40. The most important surgical risk factor is the BMI of 45 and therefore the final score allocated is 4. We designated the maximum risk score for any one patient as 6 and set a standard of a maximum total score on any one ELCS list of 6 and a maximum of three cases per list. Our rationale was that this would allow a variety of permutations of case mix. For example, a list could comprise three moderate risk cases with a score of 2 or a high-risk case with a score of 4 and two low-risk cases with a score of 1.

We sought feedback at this stage from additional stakeholders, including our local antenatal forum comprising antenatal midwives, other obstetricians and obstetric anaesthetists. Additional special features were then incorporated into the form to aid preoperative list planning, including the need for a consultant obstetrician, other surgical specialist input, preoperative anaesthetic review, antenatal corticosteroids and the requirement for cell salvage or neonatal cot availability. Using the feedback provided, we produced a final version of the booking form, which is provided in the online supplementary information A.

Supplementary file 1

[bmjoq-2018-000350-SP1.pdf]

Lessons and limitations

During this project, there were a number of positive learning points. There was excellent engagement of all stakeholders involved, from antenatal clinic midwives, obstetricians, obstetric anaesthetists, theatre staff and departmental managers. All the stakeholders recognised the ELCS list problems previously described and were highly motivated to engage in a quality improvement project that could present a solution. This meant that staff swiftly became acquainted with the ELECTIVIST risk scoring system and began to feedback to colleagues when lists were overbooked and over-ran. This aided rapid involvement of consultants at the time of booking if a list was overbooked.

We chose to modify the existing master spreadsheet meaning that antenatal clinic midwives felt comfortable with the additions, rather than having to learn a total different new system of booking. This allowed a smooth transition to the new ELECTIVIST spreadsheet and did not deter new users from engaging with it. In addition, ELECTIVIST did not require any new technological support or investment of funds which meant that the entire project was cost-neutral for the hospital. Prior to the repeat analysis, staff anecdotally were encouraged that there appeared to be considerably less over-run and if a list was appropriately booked, the entire theatre team began to recognise that the list was achievable in the time frame allocated. This meant that most of the lists that over-ran occurred in the first 3 months after implementation and that theatre efficiency continued to improve throughout.

The limitations of the project were in relation to the human interactions with the ELECTIVST system. The system was not designed to address capacity issues but to engage the booking clinicians in an immediate thought process about case mix complexity on any one list. At times when there was a capacity problem, the system had to be overseen by a designated consultant. For example, if a decision was made to perform a complex ELCS case on a particular week when there were very few remaining slots, clinicians tended to overbook the list on the basis that there was no other slot available and these lists inevitably over-ran as a result. It required a pathway for escalation to manage this more effectively and the authors acknowledge that at times, clinical need may take priority over the booking system.

We assessed for potential confounding factors which may have simultaneously contributed to the reduction in inappropriately complex case mix and theatre over-run. There was no significant difference in the grade of the operating surgeons between the baseline measurement and the repeat analysis. There were slightly fewer cases performed in the repeat analysis over a 6-month time period at 274 cases compared with 292 over the same time period in the baseline measurement in 2016. However, the average complexity score per case increased from 2.26 at baseline measurement to 2.36 in the repeat analysis. This is suggestive that the complexity of the cases was being assessed more accurately and that there was no significant difference in the number of cases performed to account for the improved ELCS list efficiency.