Related Experiment Video
Updated: Feb 17, 2026
Author Spotlight: Developing a Point-of-Care Hemoglobin Estimation Method for Anemia Management
Published on: January 19, 2024
External validation of a smartphone app model to predict the need for massive transfusion using five different
E I Hodgman1, M W Cripps, M J Mina
1From the Division of Burns, Trauma and Critical Care, Department of Surgery (E.H., M.C., H.P.), University of Texas at Southwestern Medical Center, Dallas, Texas; Department of Clinical Pathology (M.M.), Harvard Medical School, Boston, Massachusetts; Division of Trauma and Critical Care, Department of Surgery (E.M.), School of Medicine, University of Washington, Seattle, Washington; Division of Trauma, Critical Care, and Acute Care Surgery (M.S.), School of Medicine, Oregon Health & Science University, Portland, Oregon; Division of Trauma and Critical Care, Department of Surgery (K.B.), Medical College of Wisconsin, Milwaukee, Wisconsin; Division of General Surgery, Department of Surgery (M.J., P.M.), School of Medicine, University of California San Francisco, San Francisco, California; Division of Trauma, Department of Surgery (J.M.), School of Medicine, University of Texas Health Science Center at San Antonio, San Antonio, Texas; Division of Trauma and General Surgery, Department of Surgery (L.A.), School of Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania; Biostatistics/Epidemiology/Research Design Core (M.R., E.F., D.d.J.), Center for Clinical and Translational Sciences, Division of Epidemiology, Human Genetics, and Environmental Sciences (M.R.), School of Public Health, and Center for Translational Injury Research, Division of Acute Care Surgery, Department of Surgery (J.H., B.C., E.F., D.d.J., C.W.), Medical School, University of Texas Health Science Center at Houston, Houston, Texas.
A validated model shows moderate ability to predict the need for massive transfusion (MT) in trauma patients. This predictive tool can be improved with machine learning as more data become available.
Area of Science:
- Trauma care
- Transfusion medicine
- Clinical prediction models
Background:
- A previous model predicted massive transfusion protocol (MTP) activation using single-institution data.
- External validation of this model was performed using the PRospective, Observational, Multicenter, Major Trauma Transfusion database.
Purpose of the Study:
- To externally validate a predictive model for MTP activation and massive transfusion (MT) administration.
- To assess the model's predictive ability across multiple definitions of MT.
Main Methods:
- The existing model was used to calculate the predicted probability of MTP activation or MT delivery.
- Five distinct definitions of MT were applied: 10 PRBCs/24h, Resuscitation Intensity score ≥ 4, critical administration threshold, 4 PRBCs/4h, and 6 PRBCs/6h.
- Receiver operating curves (ROCs) were generated to compare predicted probabilities with observed outcomes.
Main Results:
- The dataset included 1,245 patients, with varying proportions meeting each MT definition.
- The model demonstrated consistent predictive ability regardless of the MT definition used.
- Areas under the curve (AUC) ranged from 0.694 for MTP activation prediction to 0.695–0.711 for MT administration prediction.
Conclusions:
- The app model shows moderate predictive ability for MT need in an external, homogenous trauma population.
- The model's capacity for iterative recalibration via machine learning suggests potential for improved accuracy with accrued data.
Related Concept Videos
01:15Blood Transfusion
Blood Transfusion Overview
A blood transfusion is a medical procedure used to replace blood lost due to injury, surgery, or to treat conditions such as anemia or cancer. During a transfusion, donor blood is...
02:45Blood Transfusion and Agglutination
History
The history of blood transfusion dates back to the 17th century, when early attempts were made in animals. In 1818 James Blundell, a British doctor, performed the first successful human blood transfusion. Later in 1900, Karl...