Introduction

The quality of colonoscopy is influenced by patient and operator factors. Studies investigating operator attentional fatigue on adenoma detection rate (ADR) have yielded conflicting results.

Aims & Methods

This study aims to compare ADR, polyp detection rate (PDR) and sessile serrated lesion detection rate (SSLDR) and complication rates between elective colonoscopies performed early and late in the list, across multiple operators (n>15), over 10 years, in a tertiary Brisbane (Australia) hospital.
A retrospective study was conducted on elective colonoscopies performed between 2012 and 2022 at The Prince Charles Hospital. ‘Early procedures’ are defined as procedures starting in the first hour of the first procedure starting and ‘late procedures’ are defined as procedures starting within the last hour of the last procedure starting. This pre-defined early and late procedures were used as surrogate markers for operator fatigue, accounting for a possible non-linear relationship between time and onset of fatigue, as well as the variation in number of colonoscopies performed every hour. Timing and the order of colonoscopy were used as alternative surrogates, in a secondary analysis.
Logistic regression was performed on all categorical outcome variables, and Poisson regression was modelled on continuous outcome variables, due to their non-linear distribution. Multivariate analysis included all potential confounding variables related to outcomes of interest. Statistical significance was defined as a P-value less than 0.05. Statistical analyses were performed using the SAS® statistical software (SAS Institute Inc, USA).
Exclusion included; surveillance for IBD and elective ESD.

Results

In total, 15 proceduralist (11 gastroenterologists, 4 surgeons) performed 39, 206 elective colonoscopies from 2012 to 2022. The baseline characteristics of patients, year and procedure sessions are similarly distributed between the early and late groups.
Multivariant analysis found significant associations between late procedures and lower adenoma detection rates, polyp detection rates and SSL detection rates. After adjusting for age, gender, proceduralist, year, morning or afternoon procedure, and bowel preparation quality, PDR, ADR and SDR remain 20-28% lower in late colonoscopies compared with early ones. We also found that age, gender, proceduralist, year and bowel preparation independently affected ADR in multivariant analysis. However, morning or afternoon ADRs did not significantly differ after adjusting for other variables. Proceduralist were found to be the only additional significant variable for SDR in multivariant analysis. There was no significant association between timing of procedures and major adverse events.

Conclusion

This is the largest single-centre study on operator fatigue and its effect on colonoscopy quality. Although the minimum standards for ADR was achieved, ‘late procedures’ have statistically significant lower ADR, PDR and SSLDR compared to ‘early procedures’. There was no difference in the complication rates between ‘early procedures’ and ‘late procedures’.
This study highlights the detrimental effect of fatigue in later procedures affecting the quality of colonoscopies and may be applicable in principal to other procedural based clinical work. Importantly, this may or perhaps will or is having an impact on patient safety, satisfaction, outcomes and clinical KPI’s for the operators and hospital administrators.

Disclosure

None

Introduction

Etiology of acute pancreatitis can be established in the majority of patients. However, in 25 % of cases it remains elusive. Biliary microlithiasis/sludge is detected in up to 75% of patients with idiopathic acute pancreatitis (IAP). As recurrent biliary pancreatitis can be prevented, the underlying etiology of IAP should be established. The aim of this multicentre retrospective study was to develop a machine learning-aided non-invasive prediction tool to guide clinicians in the selection of idiopathic acute pancreatitis patients to be referred to endosonography to establish etiology and allow treatment.

Aims & Methods

We retrospectively used routinely recorded clinical and laboratory parameters of 218 consecutive patients with confirmed acute pancreatitis admitted to our tertiary care hospital between 2015 and 2020. Patients who did not receive endosonography as part of the diagnostic work-up and whose pancreatitis episode could be adequately explained by other causes than biliary sludge and microlithiasis were excluded. We trained supervised machine learning classifiers using H2O.ai automatically selecting the best suitable predictor model to predict microlithiasis/sludge. The predictor model was validated in two independent retrospective cohorts from two tertiary care centers (TU Munich & University Medical Center Göttingen).

Results

28 categorized patients’ variables recorded at admission were identified to compute the predictor model with an accuracy of 0.84 [95% CI 0.791, 0.9185], positive predictive value of 0.84 and negative predictive value 0.80 in the identification cohort (218 patients). In the validation cohort, the robustness of the prediction model was confirmed with an accuracy of 0.76 [95% CI 0.673, 0.8347], positive predictive value of 0.76 and negative predictive value of 0.78 (117 patients).

Conclusion

We present a robust and validated machine learning-based predictor model consisting of routinely recorded lab work at admission that can predict biliary sludge and microlithiasis as cause of acute pancreatitis and select patients for EUS.

Clinical Case Summary

Case: W present a case of a 43-year-old man with hepatic cirrhosis complicated by hypertensive gastrointestinal bleeding, in a program to eradicate esophageal varices. One week after an elective endoscopic variceal ligation (EVL), he presented with fever, anorexia, pain in the dorsolumbar region, headaches, and psychomotor retardation. Subsequently, there was the appearance of focal motor seizures. The analyses showed C-reactive protein of 113 mg/L, without leukocytosis. Peripheral blood cultures were negative. A head computed tomography (CT) showed mild left parietal and occipital cortical hypodensities. Analysis of the cerebrospinal fluid was compatible with central nervous system (CNS) infection. Cerebral magnetic resonance revealed multiple supra and infratentorial rounded lesions, compatible with abscesses and ventriculitis. He started antibiotics with a meningeal dose. Hospitalization was complicated by SARS-COV 2 infection, so a chest CT was performed after 1 month of hospitalization, and was identified a mediastinal abscess with 5 cm with fistulization from the esophagus, compatible with iatrogenic esophageal perforation after EVL. Other possible causes of brain abscesses were excluded. He completed several weeks of antibiotic and parenteral nutrition, with normalization of the neurological status and resolution of brain lesions. About 1 month after targeted antibiotic treatment, he performed an endoscopy, revealing a deformed area in the distal esophagus but no evidence of an orifice or suppuration. Conclusion: Neurological manifestations in patients with liver cirrhosis are usually due to hepatic encephalopathy (HE). However, CNS infection, especially in patients who underwent recent procedures, such as EBL, should be considered in the differential diagnosis, given that early treatment influences the prognosis. Clinically, CNS infection can be very variable, however, it is important to be aware of atypical signs of HE, such as focal neurological deficits.

References

Krige JE, Bornman PC, Shaw JM, Apostolou C. Complications of endoscopic variceal therapy. S Afr J Surg. 2005 Nov;43(4):177-88, 190-4.

Disclosure

Nothing to disclosure

Introduction

Pain in acute pancreatitis is the symptom to be treated urgently. There are currently no recommendations for analgesic management. Intravenous lidocaine has both analgesic and anti-inflammatory effects. Its use has never been evaluated for analgesia in acute pancreatitis. The aim of our study was therefore to evaluate the efficacy and safety of intravenous lidocaine in the management of acute pancreatitis resistant to level I and II analgesics.

Aims & Methods

We conducted a retrospective, monocentric study between January 2015 and December 2021. Patients visiting the emergency department for proven acute pancreatitis and presenting with pain resistant to group I and II analgesics, hospitalized in the hepato-gastroenterology department and treated with intravenous lidocaine, were included. Pain was assessed by retrospective collection of numerical pain scales (NPS) at the emergency department, and at each analgesic re-evaluation. Treatment efficacy was defined as the absence of need for a Level III analgesic (opioids).

Results

154 patients, hospitalized in our gastroenterology unit for acute pancreatitis and presenting with pain refractory to level I and II analgesics treated with intravenous lidocaine were included. Patients were predominantly male, with a median age of 48.5 years [37.0; 58.0]. The main etiology of acute pancreatitis was ethylic in 36% of cases. Acute necrotizing pancreatitis was present in 69% of patients. 44/154 patients received a bolus of opioid morphine in the emergency department before the administration of lidocaine. Lidocaine was prescribed in a continuous way at the dose of 2g/24h (intravenous administration with electric syringe). The median NPS before prescription of lidocaine was 8.0 [7.00; 9.00]. Lidocaine treatment was effective in 88.7% of patients (136/154). 1.3% of patients experienced lidocaine-related adverse events such as skin rash and dysgeusia. In univariate analysis, patients requiring intensive care or developing an acute respiratory distress syndrome were significantly more likely to fail lidocaine (respectively OR = 5.83; CI [1.69; 20.06]; p=0.01 and OR = 4.23; CI [1.44; 12.43]; p=0.009).

Conclusion

Intravenous lidocaine administration appears to be effective and safe for pain control in 89% of patients failing to level I and II analgesia. Further prospective studies are needed to confirm these results.
​

Introduction

The introduction of lumen-apposing metal stents (LAMSs) has revolutionized the field of therapeutic endoscopic ultrasound (EUS). In the last decade, its utilization for off-label indications has markedly increased, particularly among patients with surgically altered anatomy to establish access in the excluded segment of the gastrointestinal tract. Indications for anastomosis creation include EUS-directed transgastric endoscopic retrograde cholangiopancreatography (EDGE), EUS-directed transgastric intervention (EDGI), or endoscopic reversal of Roux-en-Y gastric bypass (RYGB) for specific indications (see below).

Aims & Methods

This study aims to evaluate the effectiveness and safety of LAMS in patients with surgically altered anatomy for EDGE, EDGI, and reversal of RYGB. We collected data on all LAMS procedures conducted in patients with surgically altered anatomy, between October 2019 and February 2024 at the University Hospital of Ghent.
We used our prospective LAMS database and retrieved missing data from retrospective analysis of the medical patient records. The main outcome data of interest were technical success, clinical success, and adverse events. Technical success was defined as a successful placement of LAMS in the desired position. Clinical success was defined as either successful endoscopic access to the intended excluded gastrointestinal segment or achievement of the intended therapeutic goal when RYGB reversal was necessary. Adverse events were classified according to the AGREE criteria.

Results

A total of 73 LAMS procedures were performed: 70 cases of EUS-guided gastro-gastrostomy in patients with RYGB and 3 cases of EUS-guided gastro-enterostomy in patients with hepaticojejunostomy. Indications for LAMS procedures included 55 (75.3%) cases of EDGE, 13 (17.8%) cases of EDGI (such as gastroduodenoscopy and EUS), and 5 (6.8%) cases of RYGB reversal. Reversal of RYGB was required for one cachectic patient due to metastatic cholangiocarcinoma, three patients with hyperinsulinemic hypoglycemia, and one patient with therapy-resistant small intestinal bacterial overgrowth (SIBO).
The overall technical and clinical success rates were very high at 97.3% and 95.9%, respectively. Adverse events were observed in 11 out of 73 (15.1%) cases, including 7 (9.6%) minor events (AGREE score < 3) and 4 (5.5%) major events (AGREE score ≥ 3). Major events included stent migration (N=1), persistent fistula (N=2), and respiratory desaturation (N=1). All events could be successfully managed, and there were no procedure-related deaths. In 69.9% of cases, our method involved a dual-step approach, strategically implemented to await fistula maturation and reduce the risk of perforation. The median LAMS dwell time was 38 days (range 7, 431). LAMS was removed in 72.6% of the cases. Loss of LAMS patency occurred in 1 out of 73 (1.4%) cases and was endoscopically managed.

Conclusion

In conclusion, our study demonstrates that EUS-GG and EUS-GE offer an effective and safe approach to create anastomoses between gastrointestinal (GI) segments when performed by experts. With high technical (97.3%) and clinical (95.9%) success rates, LAMS proves beneficial for accessing excluded GI segments and reversing RYGB. Adverse events (15.1%) were mostly minor and manageable, with no procedure-related mortality. The extended patency duration and low re-intervention rate highlight LAMS as a minimally invasive and durable option in therapeutic endoscopic ultrasound procedures. Further prospective studies are required.