The University of Chicago Journal Club. Feb-March 2014.

“Management of brain trauma”

Participating faculty; Awad I; Frim D, Ramos E, Hekmat-Panah J, Roitberg B.

Residents: Dey M, Ralston A, Stamates M, Hobbs J, Wong R, Polster S,  Bhansali A, Sun Mar M.

1) Polytrauma: Pathophysiology, Priorities and Management. Otmar Trentz. In H-J Oestern et al (eds.) General Trauma Care and Related Aspects. European Manual of Medicine. Springer-Verlag 2014.

Stamates: Otmar Trentz is a Swiss Professor of trauma surgery. This book chapter deals with patients who have polytrauma.  Although not a neurosurgical problem in the narrow sense, the presence of polytrauma can dramatically change our management.

The most important objective in properly managing trauma patients is to ensure their proper resuscitation, according to ATLS protocol. If a patient is unable to be resuscitated, Trentz discusses the idea of “bail-out” surgery to decompress body cavities or control life threatening hemorrhage.  After these immediate risks are controlled, new problems arise. Patients often respond to trauma with a systemic inflammatory response, noted in literature as SIRS, reacting to pain and stress. Therefore Trentz recommends to proceed with staged, definitive surgery in the “window of opportunity” 5 to 10 days post-trauma.  This window is defined as a period of immunosuppression with new cell recruitment that follows the acute phase of hyperinflammation.

In managing polytrauma patients with fractures, the article captures the most common fracture management concepts: control hemorrhage & contamination, debride dead & ischemic tissue, limit ischemia-reperfusion injury, manage stress & pain & lower the hindrance to ICU care. In long-bone fractures, the options of early intramedullary pinning v. external fixation are debated, each with unique advantages.

The article describes three common trauma examples: massive hemorrhage from pelvic injury, planning fracture surgery in the setting of TBI and early fracture fixation in patients with concomitant chest injury. In all trauma scenarios however, Trentz advocates utilizing algorithms to optimize all trauma patients, whether or not they require surgical treatment. Early fixation has proven a critical piece of the “stabilization” objective of all trauma centers & the surgeons who care for this population.

Hekmat-Panah: We are starting the journal club with discussion of the management of polytrauma because so many of the patients we see for neurosurgical trauma are victims of polytrauma. Around 70% of car accident victims suffer from polytrauma. The existence of injury elsewhere in the body may adversely affect the condition of the brain and spinal cord.

Awad: When a patient with polytrauma arrives in the ER, trauma surgeons want to be in charge first – leading the resuscitation.

Roitberg: This book chapter helps our awareness of the general trauma patient management strategy. We must find the correct balance.  During the immediate evaluation period we should let the general/trauma surgeons stabilize the patient, but move in rapidly with any patient who has a neurological deficit or a high likelihood of a neuro deficit based on the mechanism of injury. Later in management, even if the patient has been transferred to our service as a “pure” neurosurgical problem, we must be aware of potential complications related to polytrauma. I treated a patient with traumatic intracerebral hemorrhage who had sudden hypotension three days after the event, and one day after the evacuation of traumatic growing intracerebral hemorrhage. I was the only physician called for this problem, which turned out to be a delayed splenic hemorrhage.

2) Fakhry SM, Trask AL, Walter MA, Watts DD, for the ITRC Neurotrauma Task Force. Management of Brain-Injured Patients by an Evidence-Based Medicine Protocol Improves Outcomes and Decreases Hospital Charges.  The Journal of Trauma Injury, Infection, and Critical Care 2004 March; 56:492-500.

Mar: Every year, approximately 52,000 people die as a result of TBI and 70,000-90,000 persons suffer permanent neurologic impairment. Thus, the Brain Trauma Foundation (BTF), a collaboration of 261 trauma centers throughout the continental US, developed and published Guidelines for the Management of Severe Head Injury in 1995, as an effort to promote best practice guidelines to optimize care for TBI patients.

The purpose of this paper was to investigate whether adherence to a protocol based on the BTF guidelines may result in a statistically significant decrease in hospital days, overall cost, mortality, and disability outcome measures.  Inclusion criteria was based upon blunt head injury, age > 14years, Glasgow Coma Scale sore ≤ 8, and severe head injury with AIS (Abbreviated Injury Score) >2 and <6, with survival > 48h. Data was subsequently collected on mortality, intensive care unit days, total hospital days, total charges, Rancho Los Amigos Scores (RLAS), Glasgow Outcome Scale (GOS), and related demographic data.  Total of 830 patients were identified – the population was divided into 3 groups: prior to instatement of protocol (1991-1994), during low compliance (1995-1996) and lastly during high compliance (1997-2000).

In Table 1 the comparison of demographics within the 3 populations, overall show no significant difference in age, gender, AIS.  The most common causes of injury were motor vehicle accidences, followed by falls and the mean age was 34.7years.  The findings of the paper were remarkably positive with statistically significant improvement in all studied parameters with the application of guideline based practices.  The first finding was a decline in mortality - found particularly in patients’ with lower AIS score - indicating a population for which improved guideline-based management may have a drastic effect on outcome.  The second was a decrease in length of ICU stay as well as hospital stay.  To respond to potential criticism that overall ICU stay changed during the period of the study, the data were compared with non-TBI patients, which actually showed an increase in ICU stay and no significant difference in hospital days to account for any hospital-wide changes, including social work personnel or rehab/home placement procedures, which may have biased the result.  Finally, there was an overall decrease in mean hospital charges after the protocol was implemented – this change failed to account for inflation, however, which we would imagine would skew results in favor of contradicting the paper’s outcome.  The authors also included a comparison between total charges per patient in non-TBI groups, which demonstrates a modest increase across the years in cost.  This comparison reinforces the strength of the findings, particularly in eliminating the effect of other hospital-wide changes that may have confounded the authors’ overall results.

The first identifiable concern within this paper lay in its use of a retrospective design.

Of note, there was a significant difference in alcohol level, with the high compliance population having a mean alcohol level of 77  (below the level of mean intoxication >80ml/dl), as well as a significant difference in initial GCS, with the non-compliance group having the highest GCS, the opposite of would be expected given the outcome findings.

In overall discussion, the findings and results published by this paper are significant to note in terms of the effect of instating hospital-wide policies and guidelines and its potential to profoundly affect outcome.  There is much emphasis to be placed on the extensive educational program that was created following implementation of the BTF guidelines.  The increased emphasis and awareness of critical neurologic changes that may occur in TBI patients in the ICU, including increased intracranial pressure, decreased systemic blood pressure, hypoxemia, and decline in clinical exam..  It is likely the effect of careful attention and immediate response to neurological changes that led to the overall improved outcome in the post-guideline TBI patient population. This is in stark comparison with the less timely and less efficient management by different physician protocols that was previously the standard of care. Although there are multiple flaws including its retrospective control group and analysis and the need for a multi-institutional comparison and adjustment for cost inflation, this paper provides evidence  in favor of guideline based protocol in the management of  TBI patients.

Dey: The study period is quite long, and general ICU care may have improved during that time, affecting the results. The authors tried to control for these factors, comparing the TBI population to non-TBI patients. However, this is not an ideal comparison.

Awad: This paper is an opportunity to discuss ways in which historical controls can fail you.  First, other events, not the alleged intervention could be the real cause of the changed outcome. Second, there is lack of accounting for risk factors. You have to match the groups for known factors at the very least. In this article, the authors made a serious attempt to control for biases, so this work is a worthy contribution to the literature. Another point we need to consider is the general effect of creating management guidelines: they promote a culture of attention to a particular problem. The establishment of stroke centers improved care for all stroke patients.

Hekmat-Panah: This article supports the idea that following guidelines leads to better outcomes. It is not an intuitive or self-evident result.

Roitberg: For the record, Abbreviated Injury Score was introduced in 1969, and rates injury from 1 (minor) to 6 (unsurvivable). The BTF guidelines for in-hospital management of patients with traumatic brain injury include 15 chapters covering brain specific and general management of patients. They include a broad range of topics like monitoring intracranial pressure, use of sedatives, prevention of seizures and nutrition.

This study raises an intriguing question. Are the specific rules adopted as part of this particular protocol the reason for improved outcomes? Would different (yet reasonable) rules have worked just as well? Is the mere action of implementing a protocol enough to introduce a culture of attention and better care for the condition covered by the guideline? I believe that establishing guidelines helps to diminish ignorance and silly errors. It can take a poorly functioning ICU, with frequent substandard practices, and improve the practice in a systematic way. With guidelines, a person without much experience can still make a reasonable decision.  However, guidelines cannot replace judgment based on extensive experience and tailoring care to the specific patient. In other words, rigid guidelines cannot improve upon excellent and very attentive care  that is also properly individualized.

3) Schutzman SA, Barnes P, Duhaime AC, Greenes D, Homer C, Jaffe D, Lewis RJ, Luerssen TG, Schunk J. Evaluation and Management of Children Younger than 2 years old with Apparently Minor Head Trauma: Proposed Guidelines. Pediatrics. 2001 May;107(5):983-93. Ralston: This article is a guideline creation by consensus among 9 specialists, 4 pediatric  ER, 1 ER with pediatric specialization, 2 pediatric neurosurgeons, 1 neuroradiologist and 1 general pediatrician. They reviewed 404 articles including those with children less than 2 years old with minor head trauma. They excluded those with inflicted injuries and suspected abuse. The goal was to minimize radiologic imaging while maximizing detection of intracranial injury and preventing secondary injury.

The authors created risk stratification based on history and physical examination: High risk - symptoms of intracranial injury (altered mental status (AMS), irritability, focal neurologic deficit, bulging anterior fontanelle.  These patients should undergo head CT. Intermediate risk with clinical indicators (vomiting, loss of consciousness (LOC), not at baseline per their parents, or subacute fracture).  Head CT or observation may be chosen. Intermediate risk with increased mechanism of injury (Motor vehicle collision, fall onto hard floor, unwitnessed or unclear injury). Head CT or observation. Low risk - low energy mechanism or no symptoms - no imaging needed. Indications for head CT:  Intracranial injury (ICI) predictors: skull fracture, AMS, focal deficits, scalp swelling, younger age, inflicted injury or unclear history. Among patients with intracranial injury, 60-100% have a skull fracture; whereas 15-35% of skull fractures have associated intracranial injury. If CT is normal or only shows skull fracture with no underlying ICI, the child may be observed for 4-6 hours and then discharged home based on the following criteria: 1) No extracranial injuries. 2) Easily alerted. 3) No suspicion of abuse or neglect. 4) Has access to health care quickly if needed.

Although, as we discussed above, the existence of guidelines can be beneficial, these guidelines depend on many variables, such as age, mechanism or available resources.  They are still based mainly on history and physical exam. These are preliminary guidelines that have not been validated.  Their adoption of efficacy may be limited by the fact that they were determined by specialists and not general pediatricians or practitioners from the community. There has also not been long term follow up to determine if the guidelines have changed practice. Other proposed guidelines including criteria for children less than 2 years old include PECARN with over 10,000 infants (ref 1,2). They found 1 ICI (intracranial injury – SDH/EDH/ICH/fracture) per 50 scans, and had 1 neurosurgical intervention per 200 scans. In this study, approximately 40% of minor head trauma children underwent CT scan. Another study, using CHALICE-NICE criteria did not separate out infants from older children. This study had 22,000 children, had a 30% CT scan rate and found 1 ICI per 18 scans and had 1 neurosurgical intervention per 24 scans.

Overall, no validated guidelines have been put out in regards to minor head trauma in children less than two years, but these expert defined guidelines provide a good starting point.

Awad:  How does this article fit within the methodology of guideline development? Guidelines can be evidence based, or represent expert consensus. When good evidence exists, it trumps consensus. It this article, the authors tried to provide evidence from a large but inconclusive body of literature, so we are back to expert opinion. Guidelines have their own benefits, but with expert opinion we must always be cautious. It may not represent the full range of opinions in the community. An individual or a small group may sway the opinion in a particular direction, with the rest following a leader.

Hekmat:  Who knows about the Delphi process?

Bhansali: it is a method to approach consensus. In this article they describe a process where before the meeting each member of the panel was provided with the draft management algorithm, clinical questions, evidence tables, references.  The members tried to answers clinical questions. Then they held a closed meeting of the panel, where they attempted to reach a consensus regarding appropriate management strategies. The problems with this set of guidelines are arbitrariness of cut off points and associated lack of clarity. For example - a patient is considered intermediate risk if they vomited 3-4 times, but what if the child vomited twice?

Awad: How would this guideline change your practice? Maybe it helps increase awareness of the significance of skull fractures? Fractures can serve as an indicator of the force of trauma. However, we have little evidence for our practices and even guidelines. For example – how do we know how long to keep patients for observation?

Ramos: Are we considering radiation doses for children in guidelines?

Frim: Radiation doses are not yet part of the discussion, but change is coming, especially as MRI methods advance. Our practice at the University of Chicago is to observe patients with LOC overnight, but it is not the standard everywhere. For example, it is common to keep patients for 6 hours after injury that qualified them for head CT, and it was normal.

Hekmat:  We raised many issues here, and I would like to comment on three of them.  First, there is indeed uncertainty about the significance of vomiting. Vomiting is concerning in head trauma because it can be related to brainstem injury, but it can have many causes.  Also, please note that children who are suspected victims of abuse were excluded in the paper; they have different guidelines. Children are often symptomatic neurologically at the end of an abuse process rather than a singular event.  Finally, we need to always focus on the clinical picture, and not trust the first CT scan.  In a study from UCLA (ref 3) progressive hemorrhagic injury (PHI) was found in 42.3% of patients overall and in 48.6% of patients who underwent scanning within 2 hours of injury.

Roitberg: My comment this time is not specific to a particular article, but a general thought after this journal club. I appreciate the effort everyone put into preparing this discussion. So, what is the take-home message? The literature for brain trauma remains vague and incomplete. Most of our practice is based on limited scientific evidence, and relies on personal experience with the help of guidelines derived not from scientific research but expert opinion. The good news – knowledge, even partial, is better than ignorance. Guidelines, even imperfect, help improve outcome. Even messy and controversial literature when approached by an informed and rational reader, will edify and help patient care.

References:

1) Easter JS. Comparison of PECARN, CATCH, and CHALICE Rules for Children With Minor Head Injury: A Prospective Cohort Study. Ann Emerg Med 2014 Mar 10

 2) Schonfeld D. Pediatric Emergency Care Applied Research Network head injury clinical prediction rules are reliable in practice. Arch Dis Child. 2014 Jan 15

3) Oertel M, et al. Progressive hemorrhage after head trauma: predictors and consequences of the evolving injury” J Neurosurg  2002 Jan/ Vol. 96 / 1 / 109-116”