Development and validation of 'Patient Optimizer' (POP) algorithms for predicting surgical risk with machine learning.

BACKGROUND: Pre-operative risk assessment can help clinicians prepare patients for surgery, reducing the risk of perioperative complications, length of hospital stay, readmission and mortality. Further, it can facilitate collaborative decision-making and operational planning. OBJECTIVE: To develop effective pre-operative risk assessment algorithms (referred to as Patient Optimizer or POP) using Machine Learning (ML) that predict the development of post-operative complications and provide pilot data to inform the design of a larger prospective study. METHODS: After institutional ethics approval, we developed a base model that encapsulates the standard manual approach of combining patient-risk and procedure-risk. In an automated process, additional variables were included and tested with 10-fold cross-validation, and the best performing features were selected. The models were evaluated and confidence intervals calculated using bootstrapping. Clinical expertise was used to restrict the cardinality of categorical variables (e.g. pathology results) by including the most clinically relevant values. The models were created with logistic regression (LR) and extreme gradient-boosted trees using XGBoost (Chen and Guestrin, 2016). We evaluated performance using the area under the receiver operating characteristic curve (AUROC) and the area under the precision-recall curve (AUPRC). Data was obtained from a metropolitan university teaching hospital from January 2015 to July 2020. Data collection was restricted to adult patients undergoing elective surgery. RESULTS: A total of 11,475 adult admissions were included. The performance of XGBoost and LR was very similar across endpoints and metrics. For predicting the risk of any post-operative complication, kidney failure and length-of-stay (LOS), POP with XGBoost achieved an AUROC (95%CI) of 0.755 (0.744, 0.767), 0.869 (0.846, 0.891) and 0.841 (0.833, 0.847) respectively and AUPRC of 0.651 (0.632, 0.669), 0.336 (0.282, 0.390) and 0.741 (0.729, 0.753) respectively. For 30-day readmission and in-patient mortality, POP with XGBoost achieved an AUROC (95%CI) of 0.610 (0.587, 0.635) and 0.866 (0.777, 0.943) respectively and AUPRC of 0.116 (0.104, 0.132) and 0.031 (0.015, 0.072) respectively. CONCLUSION: The POP algorithms effectively predicted any post-operative complication, kidney failure and LOS in the sample population. A larger study is justified to improve the algorithm to better predict complications and length of hospital stay. A larger dataset may also improve the prediction of additional specific complications, readmission and mortality.

Study protocol for PREPARE: a phase II feasibility/safety randomised controlled trial on PeRiopErative Penicillin AlleRgy TEsting.

INTRODUCTION: Patient-reported antibiotic allergy labels (AALs) are common. These labels have been demonstrated to have a negative impact on use of appropriate antibiotics and patient-related health outcomes. These patients are more likely to receive suboptimal antibiotics, have increased rates of surgical site infections and are more likely to be colonised with multidrug-resistant organisms. Increasing recognition that antibiotic allergy forms a key part of good antimicrobial stewardship has led to calls for greater access to antibiotic allergy assessment.PREPARE is a pilot randomised controlled trial of beta-lactam allergy assessment and point of care delabelling in perioperative patients utilising a validated antibiotic allergy assessment tool that has been repurposed into a smartphone application. The aim of the study is to assess the feasibility and safety of this approach in the perioperative outpatient setting. METHODS AND ANALYSIS: Adult participants requiring elective surgery and are likely to require prophylactic intravenous antibiotics will be recruited. During the intervention phase, participants will be randomised to the intervention or control arm, with control patients receiving usual standard of care. Those randomised to intervention undertake a risk assessment via the smartphone application, with those deemed low risk proceeding to direct oral provocation with either a penicillin or cephalosporin. Study outcomes will be evaluated in the postintervention phase, 30 and 90 days after surgery.Feasibility of intervention delivery and recruitment will be reported as proportions with respective 95% CIs. Participants who experience an antibiotic adverse event will be reported by group with respective 95% CIs and compared using modified Poisson regression model with robust SE estimation. ETHICS AND DISSEMINATION: This protocol has received approval from the Austin Health human research and ethics committee, Heidelberg, Victoria, Australia (HREC/17/Austin/575). Results will be disseminated via publication in peer-reviewed journals as well as presentation at international conferences. TRIAL REGISTRATION NUMBER: ACTRN12620001295932.

The epidemiology of rapid response team activation amongst patients undergoing major gastrointestinal surgery.

BACKGROUND: Clinical deterioration requiring rapid response team (RRT) review is associated with increased morbidity amongst hospitalised patients. The frequency of and association with RRT calls in patients undergoing major gastrointestinal surgery is unknown. Understanding the epidemiology of RRT calls might identify areas for quality improvement in this cohort. OBJECTIVES: The objective of this study is to identify perioperative risks and outcome associations with RRT review following major gastrointestinal surgery. METHODS: We conducted a retrospective cohort study using electronic databases at a large Australian university hospital. We included adult patients admitted for major gastrointestinal surgery between 1 January 2015 and 31 March 2018. RESULTS: Of 7158 patients, 514 (7.4%) required RRT activation postoperatively. After adjustment, variables associated with RRT activation included the following: hemiplegia/paraplegia (odds ratio [OR]: 8.0, 95% confidence interval [CI]: 2.3 to 27.8, p = 0.001), heart failure (OR: 6.9, 95% CI: 3.3 to 14.6, p < 0.001), peripheral vascular disease (OR: 5.3, 95% CI: 2.7 to 10.4, p < 0.001), peptic ulcer disease (OR: 4.2, 95% CI: 2.2 to 8.0, p < 0.001), chronic obstructive pulmonary disease (OR: 4.0, 95% CI: 2.2 to 7.2, p < 0.001), and emergency admission status (OR: 2.6, 95% CI: 2.1 to 3.3, p < 0.001). Following the index operation, 46% of first RRT activations occurred within 24 h of surgery and 61% had occurred within 48 h. The most common triggers for RRT activation were tachycardia, hypotension, and tachypnoea. Postoperative RRT activation was associated with in-hospital mortality (OR: 6.7, 95% CI: 3.8 to 11.8, p < 0.001), critical care admission (incidence rate ratio: 8.18, 95% CI: 5.23 to 12.77, p < 0.001), and longer median length of hospital stay (12 days vs. 2 days, p < 0.001) compared to no RRT activation. CONCLUSION: After major gastrointestinal surgery, one in 14 patients had an RRT activation, almost half within 24 h of surgery. Such activation was independently associated with increased morbidity and mortality. Identified associations may guide more pre-emptive management for those at an increased risk of RRT activation.

The association between unanticipated prolonged post-anaesthesia care unit length of stay and early postoperative deterioration: A retrospective cohort study.

This study aimed to investigate whether there was an association between an unanticipated prolonged post-anaesthesia care unit (PACU) length of stay and early postoperative deterioration, as defined as the need for a rapid response team activation, within the first seven days of surgery. We conducted a single-centre retrospective cohort study of adult surgical patients, who stayed at least one night in hospital, and were not admitted to critical care immediately postoperatively, between 1 July 2017 and 30 June 2019. A total of 11,885 cases were analysed. PACU length of stay was significantly associated with rapid response team activation on both univariate (odds ratio (OR) per increment 1.57, 95% confidence intervals (CI) 1.45 to 1.69, P < 0.001) and multivariate analysis (OR per increment 1.41, 95% CI 1.28 to 1.55, P < 0.001). Patients who stayed less than one hour were at low risk of deterioration (absolute risk 3.7%). In patients staying longer than one hour, the absolute increase in risk was small but observable within six hours of PACU discharge. Compar\ed to a one-hour length of stay, a five-hour stay had a relative risk of 4.9 (95% CI 3.7 to 6.1). Other factors associated with rapid response team activation included non-elective surgery (OR 1.78, P < 0.001) and theatre length of stay (OR per increment 1.61, P < 0.001). PACU length of stay was also independently associated with predefined complications and unplanned intensive care unit admission postoperatively. In our cohort, an unanticipated prolonged PACU length of stay of over one hour was associated with an increased incidence of rapid response team activation in the first seven days postoperatively.

Giant right coronary artery aneurysm presenting with non-ST elevation myocardial infarction and severe mitral regurgitation: a case report.

INTRODUCTION: Coronary artery aneurysms are seen in 1.5-5% of patients presenting for coronary angiography, but giant aneurysms, defined as being greater than 2 cm in diameter, are rare. Given the paucity of cases and limited experience in diagnosis and management of the disease, each case is a learning tool in itself. CASE PRESENTATION: We report the rare case of a 78-year-old Caucasian man who presented to a peripheral emergency department with chest pain and was subsequently found to have a giant right coronary artery aneurysm. Following initial investigation and treatment he was referred to our hospital for definitive management. CONCLUSION: The case described illustrates one of the varied presentations and subsequent management of an ill-defined and heterogeneous disease process. Given the limited experience with giant aneurysms in the coronary circulation, this case provides valuable insight into the clinical presentation of the disease and gives an example of the management of the most recent such case at our hospital.