Machine learning for real-time aggregated prediction of hospital admission for emergency patients.

Machine learning for hospital operations is under-studied. We present a prediction pipeline that uses live electronic health-records for patients in a UK teaching hospital's emergency department (ED) to generate short-term, probabilistic forecasts of emergency admissions. A set of XGBoost classifiers applied to 109,465 ED visits yielded AUROCs from 0.82 to 0.90 depending on elapsed visit-time at the point of prediction. Patient-level probabilities of admission were aggregated to forecast the number of admissions among current ED patients and, incorporating patients yet to arrive, total emergency admissions within specified time-windows. The pipeline gave a mean absolute error (MAE) of 4.0 admissions (mean percentage error of 17%) versus 6.5 (32%) for a benchmark metric. Models developed with 104,504 later visits during the Covid-19 pandemic gave AUROCs of 0.68-0.90 and MAE of 4.2 (30%) versus a 4.9 (33%) benchmark. We discuss how we surmounted challenges of designing and implementing models for real-time use, including temporal framing, data preparation, and changing operational conditions.

What was the impact of the first wave of COVID-19 on the delivery of care to children and adults with congenital heart disease? A qualitative study using online forums.

OBJECTIVES: Globally, healthcare systems have been stretched to the limit by the COVID-19 pandemic. Significant changes have had to be made to the way in which non-COVID-19-related care has been delivered. Our objective was to understand, from the perspective of patients with a chronic, life-long condition (congenital heart disease, CHD) and their parents/carers, the impact of COVID-19 on the delivery of care, how changes were communicated and whether healthcare providers should do anything differently in a subsequent wave of COVID-19 infections. DESIGN AND SETTING: Qualitative study involving a series of asynchronous discussion forums set up and moderated by three patient charities via their Facebook pages. PARTICIPANTS: Patients with CHD and parents/carers of patients with CHD. MAIN OUTCOME MEASURES: Qualitative responses to questions posted on the discussion forums. RESULTS: The forums ran over a 6-week period and involved 109 participants. Following thematic analysis, we identified three themes and 10 subthemes related to individual condition-related factors, patient-related factors and health professional/centre factors that may have influenced how patients and parents/carers experienced changes to service delivery as a result of COVID-19. Specifically, respondents reported high levels of disruption to the delivery of care, inconsistent advice and messaging and variable communication from health professionals, with examples of both excellent and very poor experiences of care reported. Uncertainty about follow-up and factors related to the complexity and stability of their condition contributed to anxiety and stress. CONCLUSIONS: The importance of clear, consistent communication cannot be over-estimated. Our findings, while collected in relation to patients with CHD, are not necessarily specific to this population and we believe that they reflect the experiences of many thousands of people with life-long conditions in the UK. Recommendations related to communication, service delivery and support during the pandemic may improve patients' experience of care and, potentially, their outcomes.