Biomarker-Guided Brain CT Imaging in Traumatic Brain Injury
Abstract
The Problem: Deciding on which traumatic brain injury (TBI) patient needs evaluation with a brain computed tomography (CT) scan remains a challenge for both civilian and military medical practice. Brain CT scans take a picture of the brain to look for bleeding caused by trauma. This is important because patients with bleeding in the brain may need surgery to stop the bleeding and remove blood clots. However, CT scans are not useful for patients who do not have bleeding in the brain, and unnecessary CT scans expose patients to unnecessary medical radiation (can contribute to causing cancer) and financial cost. Furthermore, in combat zones, getting a brain CT requires evacuating a wounded Warrior from a combat zone to a facility that has a CT scan. Therefore, the decision regarding which wounded Warrior needs a brain CT scan is very important and consequential. Based on work funded by the Department of Defense (DoD), the U.S. Food and Drug Administration (FDA) approved two blood tests in January 2021 to help clinicians in deciding whether a TBI patient needs a brain CT. A positive blood test means the clinician should order a brain CT, and a negative blood test means there is no need to get a brain CT. There are four important gaps in current knowledge that serve as barriers to using these tests for military medicine and routine patient care. First, studies examining the accuracy of these tests were done in a select group of TBI patients and not all-comers. Therefore, it is not clear how well these tests will work in TBI patients who were not included in previous studies. Second, the FDA cleared a one size fits all cut-off point for determining who has a positive or negative blood test. This approach results in many false positive tests (70% have positive tests but only 10% have bleeding in the brain). We are worried that when this test is used by doctors, if there are a lot of false positives, doctors may decide to skip this blood test and perform a brain CT right away. Third, before the development of this new blood test, clinicians sometimes used a series of questions to decide which TBI patients need a brain CT. It is not clear how these new blood tests fit in with that approach. Fourth, no one has studied whether the use of these new blood tests actually decreases unnecessary brain CT scanning. Our research proposal will address all four gaps in current knowledge and identify the best way to use these tests to decrease unnecessary brain CT scanning. Proposed Solution: We will use data from patients with TBI and machine learning methods to develop a new algorithm (i.e., set of rules) for using the blood tests to determine who should receive a CT scan. Our goal is to decrease the number of false positive blood tests (therefore reducing unnecessary CT scans), without missing any patients who have bleeding in the brain. Project Objective: We propose to conduct the following studies: (1) study the accuracy of the new blood tests in a population of all-comers with TBI; (2) use machine learning techniques to create an algorithm that decreases the number of patients with false positive blood tests; (3) implement the algorithm at our hospital and see whether it leads to a reduction in brain CT scan use; and (4) continue to refine the algorithm by learning from new patients we see at our hospital who get this blood test clinically. Focus Area: The proposed work fits within the Prevent sub-area of the Fiscal Year 2021 Traumatic Brain Injury and Psychological Health Research Program Translational Research Award Focus Areas since it involves improving the utility of blood tests for the evaluation and diagnosis of TBI. We are using FDA-approved tests as encouraged by the DoD funding announcement. Who This Will Help: This test will help everyone who has TBI—both Service Members and civilians. Timeline to Patient Impact: We will implement our new algorithm in year 2 of the project and hope to begin impacting pa
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Dec 28, 2022
- Source ID
- W81XWH2211024
Entities
People
- Frederick Korley
Organizations
- United States Army
- University of Michigan