Aim The aim of the present study was to identify the significant risk factors predicting the occurrence of postoperative pneumonia after major or ultra‐major operations (OT), and construct a predictive score to identify high‐risk patients for postoperative pneumonia. Patients and Methods A total of 6237 patients who underwent major/ultra‐major operations in a single institute were recruited to participate in the present study. Data from 1742 patients who underwent major or ultra‐major operations from July 2012 to June 2013 were retrieved from the Surgical Outcome Monitoring and Improvement Program database. Thirty‐eight variables were analysed by univariable analysis method. Significant factors with a P‐value ≤0.05 were further analysed by using the multivariable logistic regression model. A scoring system was then formulated by using these significant risk factors to identify high‐risk patients. The efficacy of this scoring system was examined by applying this to another 4495 patients. Results Thirty‐eight variables were included in this study (7 patients’ demographic variables, 20 preoperative and 11 operative or disease‐related variables). Five of 38 variables were found to be significant, including (i) Dependence of activity of daily living; (ii) Ascites; (iii) General anaesthesia; (iv) preoperative Dyspnoea; and (v) American Society of Anaesthesiology (ASA ) score ≥3. They formed the basis of the DAGDA score. The maximum DAGDA score was 18, with an area under the ROC curve score of 0.774 (95 per cent confidence interval: 0.711–0.836). The sensitivity and specificity of the DAGDA score with a cut‐off point of 8 were 73.6 and 73.4 per cent, respectively. Validation of the DAGDA score was examined by using another cohort group of 1329 patients (July 2010–June 2011 database), with an area under the ROC curve score of 0.751. Similar results were achieved when the scoring system was applied to the July 2013–June 2014 (1671 patients) and July 2014–June 2015 (1495 patients) database. Conclusion Dependence of activity of daily living, ascites, general anaesthesia, preoperative dyspnoea and ASA score ≥3 are significant risk factors associated with postoperative pneumonia after major and ultra‐major operations. The DAGDA score can be used to predict patients at high risk of developing postoperative pneumonia.

Aim: The reported incidence of postoesophagectomy chylothorax is 1.2-4 per cent. The diagnosis is based on clinical suspicion: milky chest drain fluid and/or high volume chest drain output. Confirmatory tests involve sending fluid for laboratory investigations. Chylothorax is a potentially lethal condition and carries a high morbidity rate. An early diagnosis can potentially help to identify those who might require operation, dietary restrictions or other treatments. The aim of the present study was to evaluate chylothorax in postoesophagectomy patients. Patients and Methods: We extracted data from patients who underwent oesophagectomy at Tuen Mun Hospital from 2004 to 2018 for the present retrospective study. Chest drain fluid from every patient was sent every postoperative day for testing for chyle, regardless of quantity and whether or not it was milky. The characteristics of the patients, disease, operations, perioperative conditions, postoperative complications, hospital stay, interventions, morbidity and mortality and chest drain outputs were compared. Results: The incidence of chylothorax in our cohort was 9 per cent (14/155). There were nine biochemical chyle leaks (64.3 per cent) and five clinical leaks (35.7 per cent). The biochemical leakage group had a significantly higher 1-month morbidity rate compared to the nonchylothorax patients (89.9 vs 50 per cent, P = 0.03). All biochemical leaks were resolved with conservative management. The entire chylothorax group also had a longer hospital stay (25.5 vs 17 days, P = 0.04) compared to the nonchylothorax group. Thoracotomy was done in two patients with clinical leaks, and the remaining leaks were resolved with conservative management. In subgroup analysis, the biochemical leak subgroup was found to have a higher 1-month morbidity rate (88.9 versus 50 per cent, P = 0.03) compared to the nonchylothorax group. Conclusion: By routinely screening for chylothorax, the incidence was found to be 9 per cent, higher than that reported in the literature. The biochemical leak subgroup was found to have a higher 1-month morbidity rate. This could suggest that biochemical leakage of chyle might have a clinical impact on the recovery of postoesophagectomy patients.

Thoracoscopic oesophagectomy is usually performed in stages and intrathoracic oesophagogastric anastomosis often requires mini-thoracotomy or extension of the thoracoscopic incisions. This paper describes a new technique whereby such an operation could be completed in one stage and the need to extend the thoracoscopic incisions is obviated.

No abstract

Aim: To investigate the clinical presentation, the risk factors and clinical outcomes in post-operative pulmonary embolism in General Surgery patients. Method: Data from the Surgical Outcomes Monitoring and Improvement Program on all major and ultra-major surgeries from the Department of Surgery of the New Territories West Cluster of Hospital Authority from July 2008 to June 2019 were reviewed. Clinical information and case notes during hospitalization of patients with pulmonary embolism were reviewed.Demographics, past medical history, drug history, pre-operative condition, operative parameters and post-operative events of patients were analysed. Subgroup analyses for emergency and elective operations were performed. Significant risk factors were further analysed by using multivariate Cox regression to adjust for potential confounding factors.Results: A total number of 20 572 patients were reviewed; 34 patients (0.17%) were diagnosed post-operative pulmonary embolism with CT pulmonary angiogram. The most common clinical presentation was dyspnoea or desaturation (61.8%), followed by tachycardia (50.0%). Deep vein thrombosis (DVT) was present in 17.6% of patients. 30-day overall mortality rate was 14.7%. Disseminated malignancy (16.7% vs 2.4%; p = .034) and cancer operation (83.3% vs 41.2%; p = .005) were associated with more post-operative pulmonary embolism in subgroup analysis of elective operations.Upon multivariate Cox regression analyses, increased age (odds ratio = 1.037; 95% confidence interval = 1.009-1.066; p = .010), post-operative renal events (odds ratio = 3.639; 95% confidence interval = 1.465-9.038; p = .005) and surgical complications required interventional radiology drainage within 30 days (odds ratio = 6.306; 95% confidence interval = 2.463-16.144; p < .001) were significantly associated with more post-operative pulmonary embolism. Conclusion:The most common clinical presentation of post-operative pulmonary embolism was dyspnoea or desaturation (61.8%), followed by tachycardia (50.0%). DVT was present 17.6% of patients. 30-day overall mortality rate was 14.7%. Increased age, malignancy, post-operative infection and post-