John K. Yue, MD

Neurosurgery Resident (PGY-6), UCSF

Residents Corner

Prognostic models hold high power in medicine and healthcare. The National Cancer Institute defines prognosis as “The likely outcome or course of a disease; the chance of recovery or recurrence.1” which is as succinct as anyone’s common understanding. In the setting of serious illness with unpredictable or unknown clinical trajectories, patients and their families often request numbers to facilitate the cognitive processing of risk, e.g., “What is the chance that my loved one will recover? When will my loved one wake up? When will they know what’s happening around them? How long does it take to go back to normal?” We know these questions seemingly invoke objective statistics to aid in decision-making, but in reality are quite subjective and dependent on one’s personal definitions and expectations. As doctors, we have the natural ability to consult our knowledge of the disease state, adjust for patient-specific risk and protective factors, readily empathize with the patient and family, without forgoing hope for the best possible outcome. Our domain-specific knowledge and personal investment enable us to provide the highest quality care for each patient. Our experience, perception, and internal feedback from these complex care encounters could either inform or bias our decision-making, and often does both.

It is important to note the subtle qualifiers within the definition of prognosis that grant its wielder the ability to advise, inform, and educate others, and just as easily to misinterpret and misapply, which (intent aside) can become misleading. “The likely outcome or course of a disease; the chance of recovery or recurrence.” “Likely” and “chance” can be quantified with numbers, as the “course” of a disease can be intensively supported by medical citations. However, as clinicians and surgeons we also know that “recovery” and “clinical course” occur along a continuum, and what “recovery” means to the treating clinician may differ from what it meant to the patient prior to injury, and what it now means as they experience the multitude of sequelae post-injury. I remind myself that perspectives may change during significant and life-altering events. My role grants the privilege and responsibility to assist, encourage, educate, and guide the patient and family’s decision-making process during their time of need. This requires interval self-evaluation of my own biases, expectations, transference, and countertransference throughout each instance of triage, counseling, and care.

Prognostication is commonly invoked during the evaluation of acute traumatic brain injury (TBI). The two most extensively validated prognostic calculators in TBI are those derived from the International Mission on Prognosis and Analysis of Clinical Trials in Traumatic Brain Injury (IMPACT) and the Corticosteroid Randomization After Significant Head Injury (CRASH) multicenter trials,2 which utilized data from TBI patients presenting with Glasgow Coma Scale (GCS) scores of 3-12 between 1984-1997,3 and GCS scores of 3-14 between 1999-2004,3,4 respectively. Using a parsimonious set of socio-demographic, clinical, radiologic, and laboratory predictors at the time of presentation to hospital, the IMPACT and CRASH calculators provide risk estimates of six-month mortality and unfavorable outcome (Glasgow Outcome Scale-Extended [GOSE] 1-4 vs. 5-8), and two-week mortality and six-month unfavorable outcome, respectively, within their indicated TBI cohorts. While numeric estimates for risk of death or “unfavorable” outcome may seem informative, we must remember that the current prognostic models use clinical data obtained at a snapshot in time (upon hospital presentation) to predict the likelihood of an outcome with multifactorial inputs at six-months postinjury, and do not integrate or adjust for the range of variables subsequent to the immediate time of hospital presentation.5 A single variation in secondary injury, neuroworsening,6,7 therapeutic intervention and treatment response, post-acute rehabilitation, caregiver support, and many other variables, could profoundly alter the patient’s trajectory of outcome. Secondly, the IMPACT and CRASH models use dichotomized outcomes (favorable vs. unfavorable), in contrast to true clinical outcomes which are multidimensional and dynamic. What is favorable may also differ based on the beliefs of the patient and family relative to injury severity, social support, recovery process, and other factors. Long-term outcomes no longer terminate at six months, as robust recent evidence has shown that subgroups of moderate to severe TBI patients continue to improve at two years and beyond 8,9 — some of whom may be clinically similar to patients who underwent early withdrawal of life-sustaining therapies.10 As explicitly stated by the IMPACT and CRASH study investigators,3,4 prognostic models provide estimates rather than certainty (the outcome has not yet come to pass), were not intended for use in isolation to direct clinical care, and certainly not for therapeutic nihilism or to justify withdrawal of care. As clinicians and neurosurgeons, we possess the ability and charge to interpret TBI prognostic model outputs with careful attention and nuance in light of their stated limitations.

The contemporary TBI casemix has changed over the past two decades, commensurate with our advancements in updating best practices for triage and treatment, neurocritical care, multimodal intracranial monitoring, neuroimaging modalities and blood-based biomarker options, multidisciplinary team-based care as the standard of care at trauma centers, understanding of outcome domains and trajectories, and community awareness and prevention. Therefore, prognostic models should be updated to remain applicable to their intended populations.11 The acute blood-based biomarkers glial fibrillary acidic protein (GFAP) and ubiquitin carboxyl-terminal hydrolase-L1 (UCH-L1) drawn within 12 hours of confirmed or suspected head trauma have received United States (US) Food and Drug Administration (FDA) approval to aid emergency and acute care clinicians in determining the need for a head CT,12 heralding the critical milestone of a new category of objective diagnostic tests for acute brain injury. Updated neuroimaging classification schemes for TBI should include lesion type, location, eloquence, size and volume,13 similar to the methodologies for grading intracranial neoplasms and arteriovenous malformations. We knew 10 years ago that 27% TBI patients without traumatic intracranial pathology on CT had detectable structural injury on MRI,14 and last month we determined that 46% of patients with isolated acute traumatic subarachnoid hemorrhage (tSAH) on CT had CT-occult intracranial injuries visible on MRI, changing their narrative from “isolated” tSAH to multifocal brain injury and bringing us closer to the pathoanatomic correlate to why tSAH is a risk factor for functional disability at three- and six-months.15,16 TBI-specific frailty,17 systemic and neuroinflammation,18 polytrauma and extracranial surgery,19 gender,19,20 and prior TBI history21 constitute additional evidence-based predictors for near-term integration into updated prognostic models of modern TBI care. Just this week, we observed evidence that earlier time to external ventricular drain placement and ICP monitoring may improve six-month functional outcomes.22 Standardized guidance for TBI follow-up care is available in the public domain at the US Centers for Disease Control and Prevention website,23,24 and we must continue building and expanding pathways and resources for community health referrals, similar to the provisions available at the UCSF Neurorecovery Clinic 25 and the UCSF Benioff Children’s Hospital Bay Area Concussion and Brain Injury Program.26

I am excited and fortunate to step toward the neurotrauma practice of today. We have our living best practices, diverse technological tools and strategies, machine learning, natural language processing, and artificial intelligence models, empowering us to become better informed on TBI diagnosis, classification, treatment, and prognosis than ever before. Most importantly, we have a cohesive, driven, and invested broader community across clinical, academic, advocacy, and policy domains and institutions. The future is bright for improving patient outcomes.

References

  1. NCI Dictionary of Cancer Terms. Published February 2, 2011. Accessed May 28, 2024. https://www.cancer.gov/publications/dictionaries/cancer-terms
  2. Dijkland SA, Foks KA, Polinder S, et al. Prognosis in Moderate and Severe Traumatic Brain Injury: A Systematic Review of Contemporary Models and Validation Studies. J Neurotrauma. 2020;37(1):1-13.
  3. Steyerberg EW, Mushkudiani N, Perel P, et al. Predicting outcome after traumatic brain injury: development and international validation of prognostic scores based on admission characteristics. PLoS Med. 2008;5(8):e165; discussion e165.
  4. MRC CRASH Trial Collaborators, Perel P, Arango M, et al. Predicting outcome after traumatic brain injury: practical prognostic models based on large cohort of international patients. BMJ. 2008;336(7641):425-429.
  5. Hawryluk GWJ. Editorial. Prognostication in traumatic brain injury: now what? J Neurosurg. Published online March 15, 2024:1-2.
  6. Yue JK, Krishnan N, Kanter JH, et al. Neuroworsening in the Emergency Department Is a Predictor of Traumatic Brain Injury Intervention and Outcome: A TRACK-TBI Pilot Study. J Clin Med Res. 2023;12(5). doi:10.3390/jcm12052024
  7. Hawryluk GWJ, Aguilera S, Buki A, et al. A management algorithm for patients with intracranial pressure monitoring: the Seattle International Severe Traumatic Brain Injury Consensus Conference (SIBICC). Intensive Care Med. 2019;45(12):1783-1794.
  8. McCrea MA, Giacino JT, Barber J, et al. Functional Outcomes Over the First Year After Moderate to Severe Traumatic Brain Injury in the Prospective, Longitudinal TRACK-TBI Study. JAMA Neurol. 2021;78(8):982-992.
  9. Deng H, Nwachuku EL, Wilkins TE, et al. Time to Follow Commands in Severe Traumatic Brain Injury Survivors With Favorable Recovery at 2 Years. Neurosurgery. 2022;91(4):633-640.
  10. Sanders WR, Barber JK, Temkin NR, et al. Recovery Potential in Patients Who Died After Withdrawal of Life-Sustaining Treatment: A TRACK-TBI Propensity Score Analysis. J Neurotrauma. Published online May 13, 2024. doi:10.1089/neu.2024.0014
  11. Yue JK, Lee YM, Sun X, et al. Performance of the IMPACT and CRASH prognostic models for traumatic brain injury in a contemporary multicenter cohort: a TRACK-TBI study. J Neurosurg. Published online March 15, 2024:1-13.
  12. 510(k) Premarket Notification. Accessed May 29, 2024. https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfpmn/pmn.cfm?ID=K223602
  13. Yue JK, Deng H. Traumatic Brain Injury: Contemporary Challenges and the Path to Progress. J Clin Med Res. 2023;12(9). doi:10.3390/jcm12093283
  14. Yuh EL, Mukherjee P, Lingsma HF, et al. Magnetic resonance imaging improves 3-month outcome prediction in mild traumatic brain injury. Ann Neurol. 2013;73(2):224-235.
  15. Yuh EL, Jain S, Sun X, et al. Pathological Computed Tomography Features Associated With Adverse Outcomes After Mild Traumatic Brain Injury: A TRACK-TBI Study With External Validation in CENTER-TBI. JAMA Neurol. 2021;78(9):1137-1148.
  16. Yue JK, Yuh EL, Elguindy MM, et al. Isolated Traumatic Subarachnoid Hemorrhage on Head Computed Tomography Scan May Not Be Isolated: A TRACK-TBI Study. J Neurotrauma. Published online March 7, 2024. doi:10.1089/neu.2023.0253
  17. Galimberti S, Graziano F, Maas AIR, et al. Effect of frailty on 6-month outcome after traumatic brain injury: a multicentre cohort study with external validation. Lancet Neurol. 2022;21(2):153-162.
  18. Yue JK, Kobeissy FH, Jain S, et al. Neuroinflammatory Biomarkers for Traumatic Brain Injury Diagnosis and Prognosis: A TRACK-TBI Pilot Study. Neurotrauma Rep. 2023;4(1):171-183.
  19. Roberts CJ, Barber J, Temkin NR, et al. Clinical Outcomes After Traumatic Brain Injury and Exposure to Extracranial Surgery: A TRACK-TBI Study. JAMA Surg. 2024;159(3):248-259.
  20. Levin HS, Temkin NR, Barber J, et al. Association of Sex and Age With Mild Traumatic Brain Injury-Related Symptoms: A TRACK-TBI Study. JAMA Netw Open. 2021;4(4):e213046.
  21. Etemad LL, Yue JK, Barber J, et al. Longitudinal Recovery Following Repetitive Traumatic Brain Injury. JAMA Netw Open. 2023;6(9):e2335804.
  22. Taylor JD, Bailey M, Cooper DJ, et al. Association Between Early External Ventricular Drain Insertion and Functional Outcomes 6 Months Following Moderate-to-Severe Traumatic Brain Injury. J Neurotrauma. Published online February 16, 2024. doi:10.1089/neu.2023.0493
  23. U.S. Centers for Disease Control and Prevention. Tips to Feel Better After a Mild Traumatic Brain Injury or Concussion. Accessed May 28, 2024. www.cdc.gov/recovery_tips_ENG-508
  24. CDC. Managing Return to Activities. HEADS UP. Published May 16, 2024. Accessed May 29, 2024. https://www.cdc.gov/heads-up/hcp/clinical-guidance/index.html
  25. University of California, San Francisco. Neurorecovery Clinic. ucsfhealth.org. Accessed May 29, 2024. https://www.ucsfhealth.org/clinics/neurorecovery-clinic
  26. Concussion Guidelines. ucsfbenioffchildrens.org. Published August 4, 2020. Accessed May 29, 2024. https://www.ucsfbenioffchildrens.org/education/concussion-guidelines