Analyzing data from more than 700,000 patients at two large academic medical centers, researchers in the Montefiore Einstein Department of Anesthesiology have developed a prediction instrument to better understand which patients are at a high risk of requiring blood transfusions intraoperatively. The prediction tool was externally validated and is highly accurate, making it an important instrument that can be used in future clinical practice to optimize blood available for transfusions during surgery.

The research project, led by the Department of Anesthesiology’s Digital Health Laboratory, is a collaboration between clinician scientists at Montefiore and Beth Israel Deaconess Medical Center. It is also a multidisciplinary effort between the Department of Anesthesiology and physicians in the Department of Transplant and Hepatobiliary Surgery, the Department of Pathology, and the Department of Cardiothoracic and Vascular Surgery. Department of Anesthesiology Chair Dr. Matthias Eikermann approached his inter-disciplinary colleagues including Transplantation Surgery Division Chief Dr. Milan Kinkhabwala, Surgical Director for Advanced Pulmonary Failure and Lung Transplantation Dr. Stephen Forest, and Clinical Pathology Attending Physician Dr. Michael Karasick about collaborating on the project to ensure input from all members of the perioperative team.

“We are using our digital health data, which is available from the medical records, to make better decisions in the operating room and in the intensive care unit,” Dr. Eikermann explained. “Any unneeded blood transfusion we manage not to administer improves the value of perioperative and critical care medicine and has a positive impact on our patients’ recovery.”

In analyzing the large set of patient data, the researchers developed the prediction model utilizing 25 preoperative predictors including patient demographics, the type of surgery, patient comorbidities, and laboratory results. The factors that put patients at highest risk of requiring an intraoperative blood transfusion were patients who were anemic preoperatively, those undergoing complex surgeries, patients with very lengthy surgeries, and patients who had multiple comorbidities. The researchers found that patients undergoing cardiopulmonary bypass surgeries had the highest risk of requiring blood transfusions, while those undergoing visceral transplant surgeries and gynecological surgeries were also at a higher risk. In addition, emergency procedures or surgeries that took place outside typical business hours were predictive for intraoperative blood transfusions.

Attending Anesthesiologist Dr. Se-Chan Kim, an author on the study, explained that this is the first research study analyzing intraoperative transfusion needs for patients undergoing a wide range of surgeries – making it a particularly exciting tool for implementing into future clinical practice.

"We can use this prediction tool in the future to calculate the need of blood transfusions intraoperatively, which will ensure the right amount of blood is issued to the patient,” Dr. Kim said. “This way, we can know exactly who needs blood transfusions during surgery and reduce the waste of blood products for those who do not.”

Utilizing blood products efficiently is of the utmost importance, as there is a blood supply shortage internationally. In addition, the prediction tool will allow hospitals and healthcare systems to allocate their resources in a cost-effective manner, as one unit of packed red blood cells and the process of transfusion costs more than $1,000 in the United States.

Of note, the prediction tool does not require advanced information technology in order to implement, making it accessible for many healthcare systems. Expanding the international visibility of the project, one of the study’s authors, Annika Eyth – previously a Research Fellow in the Montefiore Einstein Anesthesiology Digital Health Lab – received the Kosaka Top Clinical Research Award at the 2024 International Anesthesia Research Society Meeting for her work on the project. The research team’s ultimate goal is to have the prediction tool widely implemented at academic institutions across the globe, which will not only help perioperative teams provide the best possible care for their patients, but also help conserve the important resource of red blood cells.

“This interprofessional collaboration between the blood bank, the Transplant and Cardiothoracic Surgery teams, and Anesthesiology under Dr. Se-Chan Kim’s leadership is a great indicator of our success,” Dr. Eikermann said.