Capacity Management improvements using AI methodologies
Given the capacity challenges hospitals are facing, efficient use of OR, ICU and ward resources will be vital to provide the timely care that patients need. The increased demand for care, partially delayed due to the COVID-19 pandemic, requires optimization of the planning and use of resources. We investigate AI-driven approaches that provide true, actionable insights based on available data. Machine learning technologies based on specific individual patient characteristics allow for the creation of customized models to optimize OR, ICU and ward utilization. These predictive models, implemented in an adaptive user interface, will support surgery planners by creating an optimized surgery schedule. Such schedule is essential for an effective use of capacity management, leading to improved patient and staff satisfaction.
Nursing notes analysis to extract nurse worry
Nurses spend a lot of time with patients which helps them develop a good judgement about the patient. One aspect of that judgment is called nurse worry, which is subjective information on how well patients are doing. We hypothesized that nurse worry could be extracted from their daily notes and might have predictive power on certain aspects of deterioration of the patient. Subsequently, we have applied and compared different AI/NLP techniques to extract nurse worry from the textual content of nursing notes and evaluated that worry in the context of augmenting Early Warning Scores (EWS).
Nursing notes analysis to prevent in-hospital falls
Nurse notes contain a wealth of information, but often not in a structured way. Nevertheless, a lot could be learned from these notes through the use of appropriate algorithms. Currently the nurse reports are being used in a project on fall prevention. At the moment, it is difficult to predict whether a patient will run that risk, because the only known indicator is whether someone has fallen before. Therefore we are searching the nursing reports in a better way with smart algorithms to improve the fall risk estimate.
Patient Similarity for Cardiac Ischemia Decision support
The timely recognition and treatment of cardiac ischemia is of vital importance. A patient similarity approach may improve treatment decisions by comparing the current patient with a cohort of previous comparable patients. The comparison of their treatments and outcomes may lead to a well-considered and motivated decision. The patient similarity project aims to identify cohorts based on a set of patient characteristics (such as demographics, vitals, labs, medical history, and treatments) to enable applications like case-based comparisons for clinical decision support and treatment comparisons across similar cohorts.