Effective tech for our Health Service
Zella has prepared a summary for this blog post. Note that pipeline is a term used in AI to describe the set of repeatable, programmed processing steps which prepare the data, validate it, and submit it to a process which in this case generates predictions.
To date, most applications of Artificial Intelligence to healthcare have been applied to clinical questions about diseases and patient care. Now that many hospitals have electronic health records (EHRs), there is potential to use AI for operational purposes. In this study, we present a prediction pipeline that uses live EHR data for patients in a hospital emergency department (ED) to generate forecasts of emergency admissions. This involved using Machine Learning to predict each individual patient's probability of admission.We worked closely with hospital bed managers to understand how to make these patient-level predictions most useful to them. From their point of view, knowing the probability that a particular patient will be admitted is less valuable than knowing in aggregate how many patients to plan for. We therefore developed a pipeline that begins by applying ML to live data for each patient currently in the ED, and then follows a series of steps to convert the ML predictions into aggregate predictions for the total number of admissions. Our predictions outperformed a six-week rolling average benchmark that is conventionally used in hospitals to predict daily admission numbers.
Zella is the corresponding author of the paper. To discuss this work please contact her at zella.king@ucl.ac.uk.