Self-Fulfilling Prophecies Through Withdrawal of Care: Do They Exist in Traumatic Brain Injury, Too?

Abstract

Objective

We examined factors associated with withdrawal of care (WOC) in moderate-severe traumatic brain injury (msTBI) patients, and how WOC may affect short-term mortality and receipt of neurosurgery. Variability in msTBI-related outcome prognostication by clinicians from different specialties was also assessed.

Methods

Rates of WOC, factors associated with WOC, and the relation between WOC and in-hospital case-fatality rate (CFR) and neurosurgery were determined in 232 prospectively enrolled msTBI patients in the ongoing OPTIMISM Study at a level-1 trauma center. In a concomitant web-based survey with clinical vignettes, outcome prognostication comfort, treatment aggressiveness, and WOC recommendations were examined among 106 respondents from neurology, neurocritical care, neurosurgery, trauma and anesthesia/critical care.

Results

The average age of the study sample was 53 years, with a median Glasgow Coma Scale of 6. The in-hospital CFR was 36 and 68 % of patients had WOC. Factors independently associated with WOC were advanced age, pupillary reactivity, lower intensive care unit-length-of-stay, pre- and in-hospital cardiac arrest, herniation, intracranial pressure crisis, and pre-existing endocrine disease. Inclusion of WOC in our multivariable regression model predicting in-hospital CFRs negated all other variables. Survey results suggested that in younger patients, some clinicians prognosticated overly pessimistically based upon data available at the time of presentation.

Conclusion

In our msTBI cohort, WOC was the most important predictor of in-hospital mortality. We identified several important independent predictors of WOC. Large within-center variability in msTBI outcome prognostication with varying levels of possible clinical nihilism exists, which may form the basis of self-fulfilling prophecies.

Appendix: Survey Questions

UMASS Brain TOPiC: Brain Trauma Outcome Prediction in Critical Care

We are conducting an anonymous survey of clinicians who provide care for patients with moderate-severe traumatic brain injury (TBI) and/or may be asked to prognosticate the outcome of these patients. This survey should take no more than 8 min to complete. Your responses will be kept anonymous and will be part of a larger national survey of professional subspecialty societies of critical care medicine, neurology and neurosurgery. This survey is approved by the University of Massachusetts Medical School IRB (docket #H-14732). Risks and benefits: The only risk associated with completing this survey is the inconvenience of the time you spend filling it out. There is no direct benefit to you from completing this survey. However, your participation may improve the way in which clinicians provide outcome prognostication to patients with traumatic brain injury and to their families. Confidentiality: Your privacy is important to us; therefore, no identifying information will be collected in this survey. Your participation in this project is entirely voluntary. You may withdraw from the study at any time. You may skip questions. You may stop the survey at any time by exiting the browser. We greatly appreciate your time and participation.

Case 1

An 80-year-old man with a history of mild mental retardation was an unrestrained passenger in a high velocity roll-over motor vehicle accident. He had loss of consciousness on the scene. His initial GCS was 13. In the field he was normotensive, but required a non-rebreather mask. At the Level I Trauma Center he was found to have an L SDH and SAH with 5 mm midline shift, and a small SDH adjacent to the midbrain. Other injuries included multiple L rib fractures, bilateral pulmonary contusions and L arm fractures. In the emergency room trauma bay, his GCS declined to 8, and he was intubated. Repeat head CT showed an increase in the size of the L SDH with slightly increased midline shift, a diffuse axonal injury in the midbrain and new R traumatic SAH. In the trauma bay he had a fever of 38.7 °C and hypoxia requiring a PEEP of 7.5.

Case 2

A 65-year-old woman with a remote history of breast cancer was found unresponsive for an unknown period of time at the bottom of her basement stairs with an initial GCS of 3 and with “low” sats documented by EMS, but no actual value recorded. The initial BP in the field was 90/palp. She was intubated by EMS. In the emergency room of the Level I Trauma Center she was normotensive but tachycardic (HR 121) with O₂ sats of 92 % and a GCS of 7T with reactive pupils bilaterally. The initial trauma workup revealed bilateral acute temporal–parietal SDH, measuring 7 mm on the R and 10 mm on the L, bilateral diffuse traumatic SAH, without other injuries. Her INR was 1.6 without known coumadin use or liver disease. She was reversed with 4 units of FFP. She had acute renal failure [creatinine = 2.5 (baseline 0.8)], a urinary tract infection and a lactic acid level of 9.2 mmol/L. Her CXR revealed bilateral consolidations.

Case 3

A 22-year-old woman was the unrestrained driver in a roll-over motor vehicle accident during which she was ejected, and the car rolled on top of her. Her initial GCS was 3 with non-reactive pupils bilaterally. In the field she was hypotensive (BP 84/52, HR 74), with no O₂ saturation documented. She was intubated and transported to a Level I Trauma Center. Here, her GCS was 3T with non-reactive pupils bilaterally. The initial trauma workup revealed an acute L parietal subdural hematoma with 12 mm midline shift and uncal herniation, diffuse axonal injury in the L thalamus and bilateral atlanto-occipital dislocation, with associated prevertebral swelling extending to C4. In addition, she had multiple facial fractures, bilateral rib fractures with a left hemo-pneumothorax, bilateral pulmonary contusions, shock bowel and several extremity fractures. Toxicology screen was positive for alcohol and marijuana.