Dr. Nadia Gabarin is completing her transfusion medicine traineeship, supported by the Canadian Blood Services Elianna Saidenberg Transfusion Medicine Traineeship Award Program.
Two new research studies from the Michael G. DeGroote Centre for Transfusion Research (MCTR, formerly known as the McMaster Centre for Transfusion Research), part of Canadian Blood Services’ supported research network, examine the potential of predictive models to address complex issues facing health-care providers.
The first study, published in the Journal of Thrombosis and Haemostasis, developed and validated a model for diagnosing a bleeding disorder that can happen when there aren’t enough platelets to clot blood. The disorder, called immune thrombocytopenia (ITP), can be very difficult to diagnose because there is no definitive way to test for ITP and because many other conditions can also result in a low platelet count. As a result, ITP is a diagnosis of exclusion—meaning it is diagnosed through a process of elimination, after tests have ruled out all other similar conditions.
The model developed by the research team is called the Predict-ITP Tool. Like other clinical prediction models, it supports clinical decision-making by predicting the likelihood that a patient may have the disease being evaluated, based on a number of patient variables. To create the model, the research team used data from a McMaster registry of ITP patients, which included adult patients with a low platelet count who were referred to a McMaster hematology clinic. The variables in the final model included platelet count variability, lowest platelet count, maximum mean platelet volume, and history of major bleeding. The Predict-ITP tool was successfully able to estimate the likelihood that a patient with a low platelet count would be diagnosed with ITP. The next steps for this research will include further validation of the tool so that it can eventually be used by physicians when they see a patient.
The second study published in the journal Transfusion examined whether prediction models could be used to guide a hospital’s daily ordering of platelets from a blood supplier. For hospitals, it can be hard to predict the number of platelets that will be needed each day because demand can vary, influenced by the platelet transfusion needs of different patient populations. Adding to the challenge is the fact that platelets can only be stored for short amount of time before they expire (Canadian Blood Service’s platelets have either a 5- or 7-day shelf life, depending on the type of platelets). Although ordering too many platelet units than needed could lead wastage, hospitals need to maintain an adequate supply of platelets to meet patient needs.
The research team, which included researchers from MCTR and Canadian Blood Services, developed a prediction model using data on the historical demand of platelets as well as additional transfusion information from hospitals. This work, which relied on data from the McMaster Transfusion Research Utilization, Surveillance, and Tracking (TRUST) database, illustrates that prediction models can guide daily ordering decisions for platelets, with the aim of reducing wastage and shortage rates. Further evaluation of this model is needed so that it may be implemented into practice.
Through discovery, development and applied research, Canadian Blood Services drives world-class innovation in blood transfusion, cellular therapy and transplantation—bringing clarity and insight to an increasingly complex healthcare future. Our dedicated research team and extended network of partners engage in exploratory and applied research to create new knowledge, inform and enhance best practices, contribute to the development of new services and technologies, and build capacity through training and collaboration. Find out more about our research impact.
The opinions reflected in this post are those of the author and do not necessarily reflect the opinions of Canadian Blood Services nor do they reflect the views of Health Canada or any other funding agency