A new study of nearly 1,400 U.S. patients with severe traumatic brain injuries (TBI) found that some patients for whom life support was withdrawn may have survived and recovered some level of independence a few months after injury.

Because families are typically asked to decide whether to withdraw life support within 72 hours of a TBI, researchers suggest that delaying these decisions may benefit some patients. 

The study was published in Journal of Neurotrauma.

According to the authors, the findings suggest there is a cyclical, self-fulfilling prophecy taking place: Clinicians assume patients will do poorly based on outcomes data. This assumption results in withdrawal of life support, which in turn increases poor outcomes rates and leads to even more decisions to withdraw life support. However, there are currently no medical guidelines or precise algorithms that determine which patients with severe TBI are likely to recover.

In the Mass General Brigham-led study, investigators analyzed potential clinical outcomes for TBI patients enrolled in the Transforming Research and Clinical Knowledge in TBI (TRACK-TBI) study for whom life support was withdrawn. Using data collected over a 7.5-year period on 1,392 of these patients in intensive-care units at 18 U.S. trauma centers, the researchers created a mathematical model to calculate the likelihood of withdrawal of life-sustaining treatment, based on properties like demographics, socioeconomic factors, and injury characteristics. Then, they paired individuals for whom life-sustaining treatment was not withdrawn (WLST-) to individuals with similar model scores, for whom life-sustaining treatment was withdrawn (WLST+).

Based on follow-up of their WLST- paired counterparts, the estimated six-month outcomes for a substantial proportion of the WLST+ group was either death or recovery of at least some independence in daily activities. Of survivors, more than 40 percent of the WLST- group recovered at least some independence. In addition, the research team found that remaining in a vegetative state was an unlikely outcome by six months after injury. Importantly, none of the patients who died in this study were pronounced brain-dead, and thus the results are not applicable to brain death.

The authors suggest that further studies involving larger sample sizes that allow for more precise matching of WLST+ and WLST- cohorts are needed to understand variable recovery trajectories for patients who sustain traumatic brain injuries. 

“Our findings support a more cautious approach to making early decisions on withdrawal of life support,” said corresponding author Yelena Bodien of the Department of Neurology’s Center for Neurotechnology and Neurorecovery at Massachusetts General Hospital and of the Spaulding-Harvard Traumatic Brain Injury Model Systems. “Traumatic brain injury is a chronic condition that requires long-term follow-ups to understand patient outcomes. Delaying decisions regarding life support may be warranted to better identify patients whose condition may improve.”