There were five different real-world ‘catchy’ problems presented. One of the problems was presented by NIHR CLAHRC NWL’s Paul O’Sullivan, NIHR CLAHRC NWL Hon Senior Lecturer in Improvement Science and Fellowship Co-Director and consultant at Chelsea and Westminster Hospital. The research question focussed on ‘predicting the residual bed occupancy in two or three days’. A pseudo-anonymised dataset was provided alongside to support the question.
“There were approximately 25 students participating in this event. Unsure of how many students would be attracted to our problem, the other problems seemed very interesting and appealing to students, the numbers outweighed our expectations. Over the course of the week we had 9 students tackling the CLAHRC NWL problem, 7 of whom attended the whole week.” – Gabriel Burcea, NIHR CLAHRC NWL Research Assistant in Data Science.
The question posed is of critical importance since the occupancy of the hospital, at different period of the day and week impacts on the flow of patients and quality of care. Being able to predict the occupancy of the hospital for the next 2 or 3 days of the week, not only will lead to a better management of the hospital resources but also to a better quality of patient care.
“Additionally, the research question led students to develop additional projects with different research questions emerging; predicting whether A&E patients will be either discharged or admitted to the hospital; and to predict the over different time periods. At the end of the week, we ended up presenting 3 projects, with 3 different questions. And for the extra mile and effort our group made, the winning team of Data Dive, Hackathon, Imperial was of course, ours.” – Gabriel Burcea, NIHR CLAHRC NWL Research Assistant in Data Science.
With the predictive modelling developed on Chelsea and Westminster Hospital dataset, we hope to apply this model to different hospitals in London.
“These are amazing opportunities for us to develop and amazing connections were made with students eagerly wishing further collaborations with us. We will be pursuing these connections through the Health Foundation project Paul is leading, and of course will be looking to publish on the findings.” – Gabriel Burcea, NIHR CLAHRC NWL Research Assistant in Data Science.