Machine Learning-Based Risk Prediction of Critical Care Unit Admission for Advanced Stage High Grade Serous Ovarian Cancer Patients Undergoing Cytoreductive Surgery: The Leeds-Natal Score

Achieving complete surgical cytoreduction in advanced stage high grade serous ovarian cancer (HGSOC) patients warrants an availability of Critical Care Unit (CCU) beds. Machine Learning (ML) could be helpful in monitoring CCU admissions to improve standards of care. We aimed to improve the accuracy of predicting CCU admission in HGSOC patients by ML algorithms and developed an ML-based predictive score. A cohort of 291 advanced stage HGSOC patients with fully curated data was selected. Several linear and non-linear distances, and quadratic discriminant ML methods, were employed to derive prediction information for CCU admission. When all the variables were included in the model, the prediction accuracies were higher for linear discriminant (0.90) and quadratic discriminant (0.93) methods compared with conventional logistic regression (0.84). Feature selection identified pre-treatment albumin, surgical complexity score, estimated blood loss, operative time, and bowel resection with stoma as the most significant prediction features. The real-time prediction accuracy of the Graphical User Interface CCU calculator reached 95%. Limited, potentially modifiable, mostly intra-operative factors contributing to CCU admission were identified and suggest areas for targeted interventions. The accurate quantification of CCU admission patterns is critical information when counseling patients about peri-operative risks related to their cytoreductive surgery.

193 BALANCING EDUCATION AND PRACTICE: A REFLECTION FROM A SIMULATION EDUCATOR DURING THE COVID-19 PANDEMIC

What? I consider myself privileged to divide my work time between my roles as a clinical simulation educator and as an intensive care nurse in a large teaching hospital. I find that working alternate weeks in educational and clinical roles can be challenging because both demand complementary but different skills. However, I am thrilled to have the opportunity to continue caring for patients alongside supporting and learning with colleagues. Balancing these roles during a pandemic presented me with new challenges and rewards, and reflection on these experiences has given me some fascinating insights. As the COVID-19 pandemic progressed and the number of patients requiring admission to the Critical Care Unit increased, the units were expanded and staff were redeployed from other areas to provide support. These ‘surge’ staff required rapidly developed simulation-based training to allow them to work in this unfamiliar environment within a restricted scope of practice. Being involved with delivering this training as well as working with surge staff in Critical Care afforded me a deeper understanding of the surge role and the unique challenges it presented. Once surge training was completed and I returned to delivering our standard simulation-based education courses, my experiences of working clinically continued to enrich my teaching because I felt somewhat familiar with some of the challenges our learners were facing as the pandemic continued. So what? Over the last year, I have felt conflicted at times; when working clinically during the peak of the pandemic, there was very little time to facilitate learning at the bedside, and during my educator weeks I relished the opportunity to support and teach but felt guilty for spending time away from colleagues and patients in Critical Care Unit. However, continuing with both roles better equipped me to answer questions and to provide support during surge training, particularly for those staff who had not yet spent time on the units. When assisting with other courses as a faculty member, I was able to deeply empathize with participants who encountered situations that I had become familiar with in practice – for example, communicating with others when wearing full personal protective equipment – which helped me to validate and normalize some of the experiences shared during debrief discussions. Through continuing to reflect on my time spent working in these environments during the pandemic so far, I hope to present my learning and recommendations for optimizing practice under challenging circumstances.

Predicting Community Discharge for Occupational Therapy Recipients in the Neurological Critical Care Unit

Importance: Occupational therapy in the neurological critical care unit (NCCU) may enable safe community discharge by restoring functional ability. However, the influence of patient characteristics and NCCU occupational therapy on discharge disposition is largely unknown. Objective: To examine how patient factors and receipt of occupational therapy predict discharge disposition for NCCU patients. Design: Retrospective cross-sectional cohort study of electronic health records data from adults admitted to the NCCU between May 2013 and September 30, 2015. Setting: NCCU in a large urban academic hospital. Participants: Adults age 18 yr or older (N = 1,134) admitted to the NCCU. Outcomes and Measures: Using logistic regression with discharge disposition as the dependent variable, we entered sex, age, length of stay (LOS), baseline Glasgow Coma Scale score, Elixhauser Comorbidity Index, and receipt of occupational therapy services as predictor variables. Results: Of NCCU patients, 39% received occupational therapy. Younger age, shorter LOS, lower comorbidity burden, and not receiving occupational therapy services increased the likelihood of discharge to the community. Men who received occupational therapy were less likely to be discharged to the community than men who did not receive occupational therapy. As age increased, differences in the probability of community discharge decreased between recipients and nonrecipients of occupational therapy services. Conclusions and Relevance: Our results suggest that patients receiving occupational therapy services in the NCCU may have a lower likelihood of community discharge. However, these findings may result from therapist's consideration of the safest discharge location to ensure the greatest balance between independence and support. What This Article Adds: This study's findings suggest that receipt of occupational therapy in the NCCU is associated with higher likelihood for noncommunity discharge (i.e., to inpatient rehabilitation, skilled nursing, or long-term care). However, activity limitations and comorbidity burden may be greater for recipients of occupational therapy, and these NCCU patients are presumably less prepared for community discharge.

Clinical and Physiologic Factors Associated With Mode of Death in Pediatric Severe TBI

Aims and Objectives: Severe traumatic brain injury (sTBI) is the leading cause of death in children. Our aim was to determine the mode of death for children who died with sTBI in a Pediatric Critical Care Unit (PCCU) and evaluate factors associated with mortality.Methods: We performed a retrospective cohort study of all severely injured trauma patients (Injury Severity Score ≥ 12) with sTBI (Glasgow Coma Scale [GCS] ≤ 8 and Maximum Abbreviated Injury Scale ≥ 4) admitted to a Canadian PCCU (2000–2016). We analyzed mode of death, clinical factors, interventions, lab values within 24 h of admission (early) and pre-death (48 h prior to death), and reviewed meeting notes in patients who died in the PCCU.Results: Of 195 included patients with sTBI, 55 (28%) died in the PCCU. Of these, 31 (56%) had a physiologic death (neurologic determination of death or cardiac arrest), while 24 (44%) had withdrawal of life-sustaining therapies (WLST). Median (IQR) times to death were 35.2 (11.8, 86.4) hours in the physiologic group and 79.5 (17.6, 231.3) hours in the WLST group (p = 0.08). The physiologic group had higher partial thromboplastin time (PTT) within 24 h of admission (p = 0.04) and lower albumin prior to death (p = 0.04).Conclusions: Almost half of sTBI deaths in the PCCU were by WLST. There was a trend toward a longer time to death in these patients. We found few early and late (pre-death) factors associated with mode of death, namely higher PTT and lower albumin.