Introduction of paediatric laparoscopic cholecystectomy in Scotland: a national review of incidence,outcomes and training implications

Introduction

The first paediatric laparoscopic cholecystectomy (LC) performed in the UK was in King’s College Hospital, London in 1991. ¹ In Scotland this procedure was first performed in 1998 at the Royal Hospital For Sick Children, Edinburgh and was soon adopted amongst the other national paediatric surgical units in Scotland (Royal Hospital’s for Sick Children in Glasgow and Aberdeen). Specialist paediatric surgical services in Scotland are centralised to these three surgical centres which have to cover a large land mass and a disparate population of 5.5 million people.

It has been widely reported that paediatric surgeons are performing more LCs as biliary disease is on the increase. A recent study based on English Hospital Episodes Statistics data has reported a rising incidence of LC; now 2.7/100 000. ² Despite the increasing incidence; LC is still considered an infrequent operation encountered by the UK paediatric surgeon. A report by paediatric surgeons based at Chelsea and Westminster Hospital in London identified 20 LCs being performed over a 6 year period. This resulted in a consultant paediatric surgeon on average performing 1 LC a year. ³

A large North American study reported that high surgeon volume was a key factor in determining outcome rather than speciality. This study identified an overall complication rate of 15%; with gastrointestinal complications (6.4–7.2%) and specifically; bile duct injury (0.8%) and liver laceration (0.05%). ⁴ Their recommendation was that in order to optimise outcomes in LC; surgeon volume should be the chief consideration. It remains an area of debate regarding who is best equipped to treat this group of patients. British paediatric surgeons perform the procedure infrequently and adult general surgeons may have limited paediatric surgical experience.

Currently in the UK adult surgical trainees have limited or no exposure to paediatric surgical training during their higher specialist training. Adult trainees subspecialise early on in their training and do not have exposure to the broad surgical experience necessary to look after children. Compounding this lack of experience is that adult anaesthetists also have limited training exposure to children.

Key technical concerns regarding adult surgeons operating on children is a child’s varied anatomy, reduced abdominal cavity size thus restricting surgical operating fields. In addition most adult surgeons due to their training do not have adequate experience to manage a child pre and post operatively.

This diminishment of adult anaesthetic and surgical trainees and consultants experience of managing and operating on children in district general hospitals and major centres is leading to further centralisation of children’s surgery to specialist paediatric surgical centres.

We present a national data series to determine the incidence, indications, complications and training implications for LC among children <16 yrs when performed by paediatric surgeons; and consider whether surgeons in a low volume speciality such as paediatric surgery are competent to perform LC especially when the literature suggests high volume surgeons; may be more suitable.

Methods

A retrospective case note review was undertaken of all children who underwent LC performed by paediatric surgeons in Scotland between 1998–2015 (18 years) in the three paediatric surgical centres (Aberdeen, Edinburgh and Glasgow). These three children’s hospitals are stand alone tertiary centres and do not have any attached adult surgical services.

A database recording each laparoscopic cholecystectomy performed in the three centres has been prospectively recording clinical and operative data since the first procedure in 1998. Patients were identified from this database and all case notes were analysed. All patients were reviewed within three months of their operation and there was no loss to follow up of any patient.

Outcome measures in this study include; patient demographics, diagnosis, use of pre-operative imaging, grade of operating surgeon, operating time and complications.

Age and sex specific annual incidence rates were derived using the National Records Scotland (NRS) Database mid-year population estimates. ⁵ Mid year population estimates are made using the cohort component method which is a standard demographic method; factors such as net migration, births and deaths are also factored into the population estimates. ⁶ The annual estimates report the breakdown of the Scottish population by geographical region, single year of age and gender. National incidence of laparoscopic cholecystectomy were calculated as total number of cholecystectomies divided by NRS mid year population estimates (0-16 years) for each year of this study. Sex specific population estimates were used to calculate the incidences in boys and girls. All rates were reported with associated 95% confidence intervals using poisson approximation.

Trends in the observed case mix were tested using univariate linear regression and students t-test using GraphPad Prism version 6.00 for Windows, GraphPad Software, La Jolla California USA, www.graphpad.com. A p value of <0.05 was considered significant.

Results

Between 1998–2015; 141 paediatric laparoscopic cholecystectomies (LC) were performed; median age was 12 years (range: 2–16 years) and the median post-operative stay was 2 days (range 1–8.5 days). The most common indications were cholelithiasis (63, 44%), spherocytosis (57, 40%) and cholecystitis (11, 7.8%). All patients underwent preoperative ultrasound and in cases where a dilated common bile duct (CBD) was demonstrated a magnetic resonance cholangiogram (MRCP) was performed. No gallstones were identified in the CBD on MRCP.

The annual incidence of cholecystectomy increased from 0.10/100,000 to 0.88/100,000 (p = 0.069; 95% CI 0.60–1.07/100,000). Sex specific incidences were identified; 0.00–0.90/100,000 (Figure 1, p = 0.098; 95% CI 0.71–1.39/100,000) in girls and 0.20–0.86/100,000 in boys (p = 0.28; 95% CI 0.38–0.86/100,000). Cholecystectomy was more frequently required in girls (63%; p = 0.04). Seventy-five percent of cases were performed by consultants (n = 17 consultants, median = 5 cases, p < 0.05) and 25% by trainees (Figure 2). In two centres the primary operator in 90% of cases was a consultant surgeon.

OPEN IN VIEWER

Figure 1.

Incidence of paediatric laparoscopic cholecystectomy per 1,00,000 in boys and girls in Scotland 1998–2015.

OPEN IN VIEWER

Figure 2.

Laparoscopic cholecystectomy performed by consultants and trainees In the three national paediatric surgical centres.

Our data show that trainees started to perform laparoscopic cholecystectomy as primary operators in 2002. There was an increasing incidence of procedures performed by trainees during this study period (p = 0.06). There was no statistical significant difference demonstrated in operating time for LC, when performed by a consultant or trainee (p = 0.32). The median operative time for LC was 90 minutes.

The overall complication rate was 4.3% including; 1 wound infection, 4 cases of atelectasis (requiring simple chest physiotherapy) and 1 conversion occurred due to unclear anatomy caused by dense adhesions due to previous gastroschisis closure. No major complications defined as common bile duct injury or mortality were identified. No patient was readmitted within 30 days of surgery.

Discussion

Most reports of LC in British children have been of single centre experiences and hence the numbers of paediatric LC reported by UK institutions have been low numbers. We present the largest UK series to date. A recent review using Hospital Episode Statistics (HES) data suggested that there has been an effective epidemic of gallstone disease in children in England since 1997. This study demonstrated an increase in incidence in both sexes and the authors concluded that there was a three-fold increase in the incidence of paediatric cholecystectomy in England since 1997. ²

The North American literature also demonstrates an increase in LC. However many of these publications include biliary dyskinesia as an indication for LC with some reports showing this may be the underlying diagnosis accounting for as many as 50% of LC. ² In the UK biliary dyskinesia is not a commonplace diagnosis and in our study no patient underwent LC for biliary dyskinesia.

Our data shows that the incidence of LC in Scotland is lower than that of England (0.88 v 2.7/100, 000). Our sex specific incidences are also much lower as well when compared to the English data. It is unclear why we have identified a lower incidence when compared to England. It may be due to the fact that the English HES data includes LC performed for biliary dyskinesia.

A recent North American study suggested that surgeon volume was more important in influencing good outcomes than speciality. In this study they identified that low volume surgeons had more complications and longer lengths of stay than high volume surgeons. A high volume surgeon in this study performed >37 LC/year. ⁴ No single Scottish paediatric surgeon would qualify to be a high volume surgeon by these criteria. In fact the highest volume Scottish paediatric surgeon performed 31 in the entire study period of 18yrs. Our data has demonstrated an overall complication rate of 4.3% which is less than the 15% from the study by Chen and colleagues. No major complications were recorded in our study. In particular we experienced no CBD injuries, although one anatomical variation was encountered involving a short cystic duct with an overlying right hepatic artery. Anatomical variations have been reported to occur in other paediatric series with a frequency of 3%; highlighting the need for precise anatomical understanding to avoid CBD injury. ⁷ Our data suggests a quick recovery from LC with patients having a median postoperative stay of 2 days and cosmetically it is clearly advantageous.

In Scotland the first LC performed by a trainee was in 2002 in Edinburgh. Correspondingly trainees rotating through Edinburgh received greater exposure to LC with 45% cases performed by trainees (Figure 2). In the other two centres; 10% of cases have been performed by trainees.

There is a differing approach to LC in Edinburgh compared to the other centres with subspecialisation occurring in this centre. In Edinburgh 90% of all LCs were performed by only two of the consultant paediatric surgeons from the department. These two consultants had a subspecial interest in minimally invasive surgery and in fact undertook 36% of all LC’s in this national experience (50 cases of LC). Concentrating the departmental experience of LC to two of the consultants in Edinburgh allowed greater training opportunities.

A systematic review of proficiency in laparoscopic surgery was performed by Dagash and colleagues. For LC they reported that proficiency is reached after 30 procedures (range 8–200). ⁸ In the initial experience of surgeons performing laparoscopic interventions there was wide variation in operating times, conversion rates and duration of hospital stays. However in the late experience there appears to be a convergence towards improved outcomes. It has been reported that it takes 3 years to be considered competent in performing LC. ⁹ This is clearly unlikely in paediatric surgery where the volume of LC is much lower. Our data demonstrates the median number of total LC’s performed by consultants is 5 over the total study period of 18 years. Despite the relatively low number performed; trainees were given opportunities to perform LC. To increase training opportunities for current trainees in paediatric surgery we would advocate exposure to adult cholecystectomy during paediatric surgical training and subspecialisation within departments will also aid in their training.

Since laparoscopic cholecystectomy was introduced in Scotland, over an 18 year period no bile duct injury occurred when performed by paediatric surgeons. On average 8 LC’s per year are performed in Scotland spread amongst the 3 national children’s hospitals. We have demonstrated that despite this low national caseload; paediatric surgeons are undertaking laparoscopic cholecystectomy with a complication rate of 4.3% which is lower than suggested in the literature.

In the UK population it remains to be clarified who is best suited or “proficient” to perform LC in the paediatric population which as shown in the literature is an expanding cohort but nevertheless is still a small overall volume of cases.