Grant: Travel Awards - Institute Community Support

Healthcare systems globally are challenged with providing quality care to diverse patients, efficiently utilizing available resources. It is crucial to meet each patient's unique needs - some require specialized hospital care, while others can be effectively treated at home or local health centers, aiding in personalized care and resource savings. Home transfusion, introduced in 1986, has had a slow global uptake. However, with the COVID-19 pandemic showing us the importance of more patient-friendly and creative care solutions, there is renewed interest in expanding these home blood transfusion services. While there are concerns about safety and costs, recent studies have shown that adverse events are quite rare. Finding the right patients for home transfusion is essential as it requires a one-on-one care approach, highlighting the need to wisely use resources for the most suited patients. In this study, we will introduce a smart technology approach to find out which patients could benefit from receiving blood transfusions at home instead of at a hospital. We have a four-step plan: First, we will use a machine learning technique to find similar patients using electronic health records, which helps in identifying those who could do well with home care while having a low risk of serious adverse events. Next, using the information from the first step, we will create simple guidelines to identify such patients even with limited data in a practical setting. Third, we will develop a prioritization mechanism to determine which patients should be prioritized when resources are limited. Lastly, we will design a simulation model to evaluate how resources are used and how well patients' needs are met in different situations. Through this project, we aim to provide key insights on patient characteristics, resources and capacity evaluations of home transfusion, and will create opportunities to refine existing transfusion programs through data-driven techniques.