Towards proactive palliative care in oncology: developing an explainable EHR-based machine learning model for mortality risk prediction.

BACKGROUND: Ex-ante identification of the last year in life facilitates a proactive palliative approach. Machine learning models trained on electronic health records (EHR) demonstrate promising performance in cancer prognostication. However, gaps in literature include incomplete reporting of model performance, inadequate alignment of model formulation with implementation use-case, and insufficient explainability hindering trust and adoption in clinical settings. Hence, we aim to develop an explainable machine learning EHR-based model that prompts palliative care processes by predicting for 365-day mortality risk among patients with advanced cancer within an outpatient setting. METHODS: Our cohort consisted of 5,926 adults diagnosed with Stage 3 or 4 solid organ cancer between July 1, 2017, and June 30, 2020 and receiving ambulatory cancer care within a tertiary center. The classification problem was modelled using Extreme Gradient Boosting (XGBoost) and aligned to our envisioned use-case: "Given a prediction point that corresponds to an outpatient cancer encounter, predict for mortality within 365-days from prediction point, using EHR data up to 365-days prior." The model was trained with 75% of the dataset (n = 39,416 outpatient encounters) and validated on a 25% hold-out dataset (n = 13,122 outpatient encounters). To explain model outputs, we used Shapley Additive Explanations (SHAP) values. Clinical characteristics, laboratory tests and treatment data were used to train the model. Performance was evaluated using area under the receiver operating characteristic curve (AUROC) and area under the precision-recall curve (AUPRC), while model calibration was assessed using the Brier score. RESULTS: In total, 17,149 of the 52,538 prediction points (32.6%) had a mortality event within the 365-day prediction window. The model demonstrated an AUROC of 0.861 (95% CI 0.856-0.867) and AUPRC of 0.771. The Brier score was 0.147, indicating slight overestimations of mortality risk. Explanatory diagrams utilizing SHAP values allowed visualization of feature impacts on predictions at both the global and individual levels. CONCLUSION: Our machine learning model demonstrated good discrimination and precision-recall in predicting 365-day mortality risk among individuals with advanced cancer. It has the potential to provide personalized mortality predictions and facilitate earlier integration of palliative care.

The impact of elective surgery postponement during COVID-19 on emergency bellwether procedures in a large tertiary centre in Singapore.

The coronavirus disease 2019 (COVID-19) pandemic drove many healthcare systems worldwide to postpone elective surgery to increase healthcare capacity, manpower, and reduce infection risk to staff. The aim of this study was to assess the impact of an elective surgery postponement policy in response to the COVID-19 pandemic on surgical volumes and patient outcomes for three emergency bellwether procedures. A retrospective cohort study of patients who underwent any of the three emergency procedures [Caesarean section (CS), emergency laparotomy (EL), and open fracture (OF) fixation] between 1 January 2018 and 31 December 2021 was conducted using clinical and surgical data from electronic medical records. The volumes and outcomes of each surgery were compared across four time periods: pre-COVID (January 2018-January 2020), elective postponement (February-May 2020), recovery (June-November 2020), and postrecovery (December 2020-December 2021) using Kruskal-Wallis test and segmented negative binomial regression. There was a total of 3886, 1396, and 299 EL, CS, and OF, respectively. There was no change in weekly volumes of CS and OF fixations across the four time periods. However, the volume of EL increased by 47% [95% confidence interval: 26-71%, P = 9.13 × 10-7) and 52% (95% confidence interval: 25-85%, P = 3.80 × 10-5) in the recovery and postrecovery period, respectively. Outcomes did not worsen throughout the four time periods for all three procedures and some actually improved for EL from elective postponement onwards. Elective surgery postponement in the early COVID-19 pandemic did not affect volumes of emergency CS and OF fixations but led to an increase in volume for EL after the postponement without any worsening of outcomes.