Session 1: General

 Aims

Most major procedures performed by paediatric surgeons occur with low frequency.  Conventional statistical analysis depends on sample size to allow detection of significant differences.  It is therefore assumed that variation in operator performance in a low frequency speciality will be impossible to demonstrate.  Sequential analysis adopts a different approach to data which is not dependent on probability distributions and sample size; this has been adopted by cardiac surgeons.  The analysis may be interpreted graphically, with series which cross limits being designated as out of control. We aimed to analyse leak rates for oesophageal atresia (OA) surgery comparing conventional analysis with statistical process control charts.

Methods

All OA repairs in a regional centre were prospectively recorded. Four consultants had repaired more than 10 OA, and their sequence of repairs and leaks were recorded in chronological order.  The leak rate and 95% confidence intervals were calculated for each consultant.

Sequential analysis was performed, setting the target rate for OA leaks at 5%, with a signal rate at 10%, α and β were both set at 10%.  CUSUM charts were generated for each consultant.

Results

The leak rate and 95% confidence intervals for these proportions were:  Consultant A: 6/49 12% (95%ci 5-24%), Consultant B: 6/32 19% (95%ci 8-35%), Consultant C: 3/19 16% (95%ci 5-37%). Consultant D: 2/16, 12% (3-36%).  All leak proportions were within the 95% confidence intervals of each other.

CUSUM charts show that the leak rate for Consultant B was out of control after patient 12 in their series.  All other consultant’s series were in control.

Conclusions

Sequential analysis allows discrimination of operator outcomes at low frequency when conventional statistical analysis suggest there is no difference in outcomes.  For a low frequency specialty such as paediatric surgery, sequential analysis should be the preferred technique for analysis of outcomes.

Aim            Adaptation after short bowel syndrome (SBS) may require home parenteral nutrition (HPN). We hypothesised that survivors after necrotising enterocolitis (NEC) achieve enteral autonomy, weaning off HPN, sooner than patients whose surgical diagnoses are often associated with intestinal dysmotility.

Methods   We reviewed patients with SBS on HPN analysing: diagnosis, gestational age, bowel length, enteral autonomy, and mortality, determining correlation-coefficients for gestational age and producing a Kaplan-Meier plot to compare the age of achieving enteral autonomy between two surgical groups. 

Results      Our HPN-service managed 142 patients between 1987–2015. Fifty-nine (51%) had SBS. On 1st January 2020, 45/59 (76%) patients were alive: NEC (18/20), Gastroschisis (9/14), Atresias (7/10), Miscellaneous (6/9), Volvulus (5/6). We compared two groups: [NEC] (18) and [GAMV] (27).

Median intestinal lengths were identical in both groups at 40 cm (IQR 30-50). The [NEC] group had a shorter median gestation: 27 weeks (IQR 26–30.5) vs 36 weeks (IQR 34–40) for the [GAMV] group, (p=0.002 Mann-Whitney). There was no correlation between gestational-age and age of enteral autonomy (Pearson r =0.2, p=0.3). There were 2 liver transplants in each group and one additional patient in each group had a combined liver-small bowel transplant.

All patients in the [NEC] group achieved enteral autonomy with a median of 645 (IQR 452–861) days. In the [GAMV] group, in which 3 remain on HPN, the median time to enteral autonomy will exceed 991 (IQR 535–1586) days giving a difference in medians of at least 346 days (p<0.03 Mann-Whitney). The percentage on HPN, against time, appears in Figure1.

Conclusion    Surgical neonates with SBS who require HPN and who had NEC can expect to achieve enteral-autonomy 1 year sooner than those with other diagnoses. This advantage for the [NEC] group does not correlate with gestational age but may reflect better intestinal motility. 

Aim of the study

A core outcome set (COS) for infants born with gastroschisis was developed in 2019 and aims to optimise reporting of relevant outcomes and facilitate results comparison across studies. For a COS to be useful the outcomes included must be achievable and available, ideally from routinely collected data. We assessed the availability of outcomes within the gastroschisis COS from routine electronic patient records (EPRs), to inform on the COS usefulness in practice.

Methods

The availability of each outcome for gastroschisis patients managed in over a 2 year period, was determined from local and national (BadgerNet) EPRs. Each outcome was categorised for data availability and conformance with published COS definition. The COS comprises: mortality, sepsis episodes, growth (weight, length/height and head circumference as z-scores for local growth data), number of operations, time on parenteral nutrition, liver disease (conjugated bilirubinaemia), number of severe gastrointestinal complications and quality of life (as per PedsQL questionnaire).

Results

Records for 15 patients were analysed.

7/8 outcomes were available in the EPR for all cases. Quality of life questionnaire (PedsQL) was not documented for any infant.

 5 outcomes (mortality, sepsis, severe gastrointestinal complications, and parenteral nutrition, number of operations) could be recorded from EPRs as per COS definitions.

The other 2 outcomes did not conform with COS definitions:

• growth  –  measurements were taken at variable time points, not conforming to suggested uniform points of data collection
• liver disease – total rather than conjugated bilirubin measured in all patients

Conclusions

The majority of gastroschisis COS outcomes are routinely available within EPRs. Quality of life is not routinely recorded for infants born with gastroschisis. For 2 of 8 outcomes, routinely available data did not meet the precise definition within the COS. These findings will inform the design of future research into gastroschisis.

Aim

Respiratory morbidity in patients with exomphalos major (EM) ranges from short term, self-limiting respiratory difficulties relating to abdominal wall closure, to long term chronic lung disease and death. Increasingly, a subset of infants with severe respiratory insufficiency (RI) is being recognised. The aim of this study was to investigate RI and its prognostic factors in infants with EM.  

Methods

A 13-year (2005-2018) single centre retrospective review of infants with EM (fascial defect >5cm +/- liver herniation) was performed, to compare EMRI (EM with respiratory insufficiency, defined as diagnosis of pulmonary hypoplasia (PH) and/or pulmonary hypertension (PHTN) on chest radiograph or ECHO respectively), to EM without RI (EMWRI). Data were stored and analysed using GraphPad Prism 8. Results are presented as median (range).

Results

Forty neonates with EM were treated during the study period. Overall mortality was 20% (8/40). The group of infants with EMRI and EMWRI were comparable in parameters including gestation, birth weight, sex, comorbidities and number of procedures to close defect (Table 1). None of the EMRI infants had documented prenatal diagnosis of PH. 

EMRI infants were significantly more likely to be intubated at birth and have a higher mortality. Causes of death in the EMRI cohort were exclusively related to PH, PHTN, chronic lung disease or a combination. Cause of death in the infant who died in the EMWRI cohort was sepsis with multi-organ failure.

Conclusions 

- RI leads to high morbidity and mortality in EM, with implications for antenatal and postnatal counselling of families.

- EMWRI is not associated with high mortality, suggesting that RI is a predictor of poor survival.

- Further prospective studies are needed to investigate for predictive factors of the EMRI group, and to assess if invasive surgical management could be avoided.

Aim

Paediatric Surgical training in the UK has evolved organically to provide a diverse experience of rare conditions.  We aimed to present the unique professional and personal demands of this vocation to prospective applicants and to committees spearheading the future shape of training.

Method

A survey was created containing objective questions on Paediatric Surgical training in the UK with input from all authors.  Current and recent national training number holders were asked to complete the survey via a link distributed nationally by e-mail.  Data are given as number(%) or number(interquartile range).

Results

50 responses were obtained from all regions of the UK. 31 were current trainees, 19 had recently completed training (CCT).  Respondents started higher surgical training 2007-2019, aged 25-39 years.

21(42%) obtained a training number at the first attempt, with 35(70%) achieving their first choice of region for training.  25(50%) surgeons took time out-of-programme (OOP) for research or further experience with one third completing a higher degree during training.  15/44 (35%) took at least 6 months OOP for maternity/paternity or compassionate leave. 11(22%) trainees have worked less-than-full-time.  34(68%) respondents lived away from partners or dependants for at least 3 days a week, for more than 3 consecutive months.  

11(44%) trainees reported extending their training, 4 to achieve required index case numbers and 3 to complete examinations.  Median age at CCT was 37(36.5-39).  64% spent 2 or more years post-CCT before commencing a substantive consultant position and 36% moved region to do so.

Conclusion

Paediatric Surgical training in the UK is challenging and can exert significant pressures on a young workforce.  Prospective trainees should have access to this data.  Committees responsible for changes to training should consider whether this pathway promotes a diverse and representative consultant body.

Aim of the Study

The optimal timing of neonatal stoma closure is uncertain. The ToSCiN study aims to determine the feasibility of a potential future randomised controlled trial comparing early and late stoma closure.  In this part, we undertook a survey of UK healthcare professionals to determine current practice and attitudes.

Methods

An online survey was distributed to neonatal surgeons, neonatologists, specialist nurses and dieticians across the UK. The survey explored current stoma closure practice and key considerations relative to expediting or delaying surgery using a 19-item closed-question survey with optional/validating free-text responses and six open-ended questions.

Results

One hundred and sixty six professionals (52% surgeons, 40% neonatologists, 8% other) from 28 centres responded. Forty-seven percent (78/166) self identified as proponents of early stoma closure (28% late and 25% unsure). Across 4 representative clinical scenarios (2 preterm, 2 term), the most frequently reported target time to stoma closure was 6 weeks, although more respondents favoured longer intervals for preterm as opposed to term infants. Forty-one percent of respondents favoured reaching a pre-defined weight prior to preterm stoma closure, with 2500g being the most common threshold. The majority (73%) of respondents expressed a preference for stoma closure prior to discharge for preterm scenarios with less consistency for term scenarios. Reasons for expediting stoma closure identified in qualitative thematic analysis were: poor nutrition, high output, stoma complications, social concerns and intravenous access. Reasons to delay stoma closure related to: thriving/enteral autonomy, comorbidities, underlying gut pathology/technical concerns and logistics e.g. access to theatre.

Conclusions

This large study of practice and practitioner views demonstrates relatively high concordance of opinion with regards to initial target time for stoma closure and and has identified factors that would expedite or delay closure. These data will help inform feasibility and design of a future trial.

Aim of the Study: The fragility index has been proposed some years ago as a tool to assess the robustness of randomised controlled trials. The fragility index evaluates how many patients in a randomised controlled trial need to have a different outcome to change the result of a trial from statistically significant to insignificant and vice versa. We therefore analysed randomised controlled trials in paediatric surgery from the last decade for their robustness.

Methods: We searched MEDLINE for paediatric surgical randomised controlled trials and screened 516 records. Out of these 84 dealt with paediatric surgery and were published between 01/01/2010 and 31/12/2019 thus they were eligible for full-text analysis. We included 61 trials and calculated the fragility index for dichotomous outcomes between two groups.

Main Results: The 61 included trials reported 27 primary, 50 secondary, and 104 additional aims resulting in 181 comparisons. Of them 57 were statistically significant and 124 were not. Median trial size was 63 participants (Q_25%-Q_75%: 41–131) and median total events were 9 (Q_25%–Q_75%: 4-20). Median fragility index was 1 (Q_25%-Q_75%: 1–4) and in 14 comparisons the fragility index was 0. A fragility index ≥10 was found in 5 comparisons. Median reverse fragility index was 3 (Q_25%–Q_75%: 2-4), but 18 comparisons had a reversed fragility index of 1, resulting in statistical significance with one additional patient experiencing the investigated outcome. 

Conclusions: More than 50% of statistically significant comparisons in paediatric surgical randomized controlled trials could be turned to insignificant by adding just one event to the smaller comparator, whereas in 14 comparisons not even one event could be added without changing the result. In contrast, reverse fragility was often higher. Consequently, trial results need to be interpreted cautiously if their fragility index is low particularly if the event count was low, too.

Aim of the Study:

The aim was to compare the safety of paediatric laparoscopic cholecystectomy (LC) at Paediatric-Centres (PCs) and Adult-Centres (ACs).

Method:

A literature review of the last 20 years was performed using relevant MeSH terms and following PRISMA guidelines. Statistical analysis was performed using Stata-16 (Stata Corp, Texas, 77845, USA).

Results:

Between 2000 and 2020, 114 articles describing 76,524 paediatric LC were identified. Most (57%) were performed at ACs. There was no significant difference in demographics between PCs vs ACs: [female-male ratio (2.21 vs 2.60), mean age (12.04 vs 11.3years,p-value=0.457), mean weight (48.2 vs 44.9kg,p-value=0.77), mean BMI (25.1 vs 22.7,p-value=0.265)] respectively.

Common indications at PCs and ACs were Cholelithiasis (27.4%vs32%, p-value=0.521) and Biliary dyskinesia (12.9%vs23.5%, p-value=0.9125). There was no significant difference between median inpatient-stay at PCs (3.75days) and ACs (2.85days) (p-value=1.0). There was a significant difference in intra-operative cholangiogram (IOC) use at PCs (24.6%) vs ACs (30.2%)(p-value<0.0001). Major complications at PCs and ACs included: bleeding (2.0% vs 1.6%, p-value=1.0), acute chest syndrome (5.5% vs 1.6%, p-value=0.056), and bile duct injury (BDI) (0.80% vs 0.37%, p-value<0.0001). Major complications at PCs without comparable AC data included: bile leak(2.49%), duodenal perforation(0.32%), retained stone(1.93%). Complication at PCs and ACs were managed with re-operation (2.37% vs 0.74%, p-value=0.0088) and conversion to open surgery (2% vs 4.64%,p-value=0.0019). Four studies were suitable for meta-analysis to compare complications, which showed no significant difference (p-value=0.92)(Fig.1).

Conclusion:

There was no difference in the demographics and overall complication rates between PCs and ACs. The number of paediatric LCs performed, IOC and Conversion rates were higher at ACs, whereas BDI and re-operation rates were higher at PCs. LC was found to be a safe operation at both PCs and ACs with comparable outcomes.

Figure 1: Forest plot comparing overall complications between PCs and ACs.

Aim of the study 

Recent evidence suggests simple laparoscopic inguinal herniorrhaphy is associated with higher rates of recurrence and testicular ascent. We instigated a highly standardised approach of laparoscopic inguinal herniotomy (LIH), with circumferential division of the sac and ‘purse-string’ closure with 4/0 monofilament polypropylene. An active follow-up programme was then pursued.

We aimed to review the outcomes of this technique and compare them to a cohort of open herniotomies (OIH).

Method

LIH patients were identified prospectively (2017-2019): OIH retrospectively from 2016. Risk factors for complications were defined as prematurity (<32 weeks), emergency presentation with incarceration and redo surgery for recurrence. Data are presented as median [IQR]. Comparison used Fisher’s exact and Mann-Whitney U tests: significance defined as p<0.05.

Results

125 inguinal herniae in 86 patients were included in the LIH group and 214 herniae in 179 patients in the OIH group. Groups were similar in terms of age and sex (Table 1). The LIH group had a significantly larger proportion of cases that were premature, had emergency surgery for incarceration or were redo surgery after previous OIH.

Follow-up was 15.0 months [8.4-18.6] vs. 48.9 [47.0-51.0], LIH vs. OIH. Hernia recurrence occurred in 2/125 (1.6%) vs. 4/214 (1.9%) (LIH vs. OIH), p=1.00. There was one known case of testicular ascent after OIH but none to date in the LIH group.

Conclusion 

Recreation of the open herniotomy operation laparoscopically appears to confer excellent outcomes, with low rates of recurrence despite a high proportion of patients having known risk factors. Further, long-term data on rates of testicular ascent after active follow-up are awaited.

Aim of the Study

Laparoscopic percutaneous extraperitoneal closure (LPEC) has recently been widely accepted for use in pediatric inguinal hernia repair. We previously reported that LPEC produced satisfactory results regarding recurrence rate and complications in general pediatric patients. Additionally, operative time was shorter and incidence of postoperative contralateral metachronous inguinal hernia (CMIH) was lower with LPEC than with traditional open approach (OR). However, the usefulness of LPEC in infants has not yet been fully verified. This single-center study aimed to compare LPEC with OR in patients under 1 year old.

Method

Following institutional review board approval, we conducted a retrospective chart review. We compared LPEC with OR in 172 infants who underwent OR from July 2003 to June 2008 and 404 infants who underwent LPEC from July 2008 to December 2019. Mean follow-up time was 165 months for OR and 70 months for LPEC (p<0.01). In consideration of this difference in follow-up time, log-rank testing was used to analyze long-term results.

Results

There were no significant differences in patient characteristics such as age, weight, sex, or preoperative laterality. Asymptomatic patent processus vaginalis was identified in 44.6% of unilateral hernia patients who underwent LPEC, and prophylactic surgery was performed accordingly. Mean operative times with OR and LPEC were 36.9 min and 21.7 min, respectively, for unilateral surgery (p<0.01), and 59.6 min and 29.8 min, respectively, for bilateral surgery (p<0.01). The frequency of postoperative recurrence was 1.52% with OR (3/198 sides) and 2.57% with LPEC (12/467 sides; p=0.27). The frequency of postoperative CMIH was 6.85% with OR (10/146) and 0.59% with LPEC (1/341; p<0.01).

Conclusion

In conclusion, LPEC was associated with reduced operative time and lower CMIH incidence, suggesting benefits of using LPEC in infants. Further investigation is needed to reduce the recurrence rate, which was slightly higher with LPEC.

Background

Childhood obesity is a serious public health challenge. Bariatric surgery is the most effective available intervention for weight loss and obesity-related comorbidities and is gaining popularity as a treatment modality for selected group of morbidly obese adolescents.(MOA)

AIM: To report the efficacy and outcome on MOA that underwent laparoscopic sleeve gastrectomy LSG at a British Specialist Paediatric Centre.

Method

A retrospective review of patients, managed by specialist Paediatric MDT team throughout, who underwent LSG from 2011 to 2019 was performed. 

Patients with at least 1 follow up beyond 6 months. 

Thirty patients were identified (M=8, F=22) 

At Surgery - Mean age 17 yrs, mean BMI 51.8 kg/m2 

Associated Comorbidities were: type 2 DM (n= 2), obstructive sleep apnoea, (n= 4). NAFLD (n= 15), PCOS (n= 11), hypertension (n=8), Blount’s disease (n= 1) and psychosocial issues (n= 5).

% Excess Weight Loss (%EWL), Comorbidities Resolution and Complications were evaluated. 

Results

At Surgery: Mean Age: 17, Mean BMI 51.8 Kg/m2, 

Associated Comorbidities 

The mean %EWL, Mean BMI & Mean BMI SDS changes were. 

 Co-morbidities Resolution at 1 year were: T2DM = 2/2, Hypertension = 5/8, OSA= 4/4. Improvement in Quality of life in all. 

Complications: One each :stapling of naso-gastric tube, collection & pneumonia managed conservatively,  dysphagia and iron-deficiency anaemia.  Decreased vitamin- D in 5 . 4 Patients showed Weight regain at 2 years.

Conclusion

Sleeve Gastrectomy can be safely performed by specialist Paediatric Team with appropriate support. It is effective in MOA in the short and mid-term. They need long-term monitoring by specialist MDT and appropriate transition to adult teams. Further studies  are required in British settings.

Aim

Acute abdominal pain in childhood is a common presentation. Knowing the accurate cost of treatment for children with suspected appendicitis would improve service design and inform attempts to improve diagnostics. No formal UK-based health economic analysis (HEA) has been published.

Our aim was to define the potential patient pathways and identify costs for each.

Method

Electronic patient records, health informatic coding data and operating lists were used to identify children referred to a single centre (April 2017 – March 2018 with possible appendicitis). Using input from a Paediatric Surgeon and a Health Economist, pathways representing patient journeys from presentation to discharge were constructed.

Financial software, patient level information and costing systems (PLICS), was used to extract costs (A&E, doctors, ward, pharmacy, theatre, radiology, pathology and overheads) and incomes for each patient. Statistical analysis was performed using SPSS.

Results

Nine potential pathways were identified. In our cohort (n=228) complex appendicitis was more expensive than simple (£6794 v £4810) (p<0.001). Delaying treatment (‘inadvertent’ discharge/delayed diagnosis with interval appendicectomy) (n=15), led to higher costs in both simple (£6915 v £4810) and complex (£8319 v £6794) cases. Increased length of stay (LOS) was the largest contributor to increased costs. Mean LOS in simple cases was 4.1 days whereas in complex it was 7.1 days (p<0.001). 

Income, based on national tariffs, did not cover the service costs for any pathway. Cost associated with diagnostics, currently account for the lowest proportion of costs (3.18%).

 Conclusion

The cost of managing patients with possible appendicitis was greater than the national tariff for all pathways. Complex appendicitis cost more than simple appendicitis and increases bed occupancy. Diagnostics consumed the lowest proportion of costs across all pathways. 

Improving diagnostic tests and reducing the frequency of complex appendicitis would potentially reduce overall cost, length of stay and reduce morbidity. 

Introduction: Who best manages an “acute scrotum” in a time-critical manner is currently controversial. We reviewed current practice to try and identify current strengths and weakness. 

Method: Retrospective review of all emergency scrotal explorations admitted to a network(4 paediatric surgical units with population~5.75million).  Data are presented as median(IQrange) and compared using non-parametric tests. P<0.05 was regarded as significant.

Results: 560 scrotal explorations were performed between Jan.2016 and Dec.2018. Overall age at surgery was 11 years (2days-17years). Infants <1 year (n=25) were regarded as separate category and therefore excluded. Commonest pathology was hydatid torsion (n=257,48%) and then testicular torsion (n=128,24%). Time to surgery was 26(13-53) and 9(5-24)(P<0.001) respectively. Symptom onset to time of presentation was 24(9-48) and 5(3-24);P<0.0001.    

The “nearest” hospital was one of the four units in 302/535(56%) cases. In the others (n=233) the distance to a nearer general hospital was 3.2(0.2-40) miles compared to the actual distance travelled of 6.0(2.4-78)(P=0.25) miles.  

Only 39/128(30%) with testicular torsion reached theatre within 6 hours of onset. 27/71(38%) of those who did not reach the 6 hour window had a general hospital closer than the definitive paediatric surgical unit. Overall 47/128(37%) had a closer local hospital and of these actual distance travelled was 14(9.1-24) miles.

Overall rate of testicular loss (orchidectomy/testicular atrophy) was 34/128 (26.5%) with a time to theatre of 34(17.7-76.13) hrs and age of 13.02(5.9-14.5)years. There was a single example of testicular loss with a time to theatre of <6 hours.

Conclusion: 

• Late presentation is the biggest component to overall time to surgery. 
• High proportion bypass a closer unit to obtain surgery from tertiary unit. 
• Targeted public health education and access to local hospital facilities are needed to improve outcomes of testicular torsion. 

Background: Intra-abdominal collections (IAC) are a common complication following appendicectomy, one of the most commonly performed emergency abdominal procedures in childhood.  Some respond to conservative management while others may need drainage.

Aim: To compare the outcomes of medically and procedurally managed post appendicectomy IACs and suggest a method of standardising the need for intervention

Methods: A single centre, retrospective review of children aged ≤16 years presenting between 2014 and 2019. Patient demographics, management and outcome data were collected. IAC volume and surface area were calculated assuming a prolate spheroid or true ellipsoid depending on the number of dimensions reported.

Results: 60 patients (18%) of patients developed an IAC post appendicectomy.  Medical management was undertaken in 44 (73%), drainage in 12 (20%) and surgical washout in 4 (7%). Collection size was associated with failure of medical management across a range of measurements; maximum diameter (p=0.028), volume (p=0.002) and surface area (p=0.001).  Collections with a volume of 2ml/kg were significantly less likely to fail medical management than larger collections (0/33 vs 6/11; p<0.0001).

Discussion:  Not all post appendicectomy IACs require drainage, the relationship between collection volume and need for drainage is more closely assessed using a volume or surface area calculation than a single dimension measurement, particularly when adjusted for weight of the child.  A cut off of 2ml/kg appears to be a good objective measure for intervention and provides a communication tool for discussion among the multidisciplinary team.  A prospective study or randomised, controlled trial on this topic would be extremely timely.