|
Article 1
Prospective derivation of a clinical decision rule for thoracolumbar spine evaluation after blunt trauma: An American Association for the Surgery of Trauma Multi-Institutional Trials Group Study. Inaba K, Nosanov L, Menaker J, Bosarge P, Williams L, Turay D, Cachecho R, de Moya M, Bukur M, Carl J, Kobayashi L, Kaminski S, Beekley A, Gomez M, Skiada D; AAST TL-Spine Multicenter Study Group. J Trauma Acute Care Surg. 2015 Mar;78(3):459-65; discussion 465-7.
Unlike in cervical spine injury, where NEXUS and the Canadian C-spine rules can be used to rule out a significant injury, there exists no evidence-based guidelines to determine a clinically significant thoracolumbar spine injury. Therefore, Inaba et al. sought to develop a clinical decision rule to aid in the diagnosis of injuries to the thoracolumbar spine. The authors performed a prospective multi-institutional observation trial as part of the American Association for the Surgery of Trauma multi-institutional trials group study. Blunt trauma patients were prospectively enrolled at 13 US trauma centers over a two-year period. Exclusion criteria included patients with either a cervical spine injury with neurologic deficit, pre-existing tetraplegia/quadriplegia or did not undergo any thoracolumbar imaging. Patients were also excluded if they were < 15 yrs. of age or deemed unevaluable (had a GCS score < 15, were intoxicated or had a distracting injury). The primary endpoint for the study was the presence of a clinically significant fracture of the thoracolumbar spine. Clinically significant fractures were determined based on the need for either thoracolumbar spine orthosis or surgical stabilization. Components of the clinical decision rule were determined based on a single prospective study as well as univariate analysis of the current study. Based on these criteria, the clinical decision rule was composed of 3 components which included 1)a positive physical examination ( pain, tenderness to palpation, deformity or neurologic deficit) 2) a high risk mechanism of injury (fall, crush injury, MVC with rollover and/or ejection, MCC and MVC vs PED) and age greater than 60. Using receiver operator characteristics and multi-variant regression analysis, the clinical decision rule was then assessed to determine its predictive capacity.
The authors screened 12,479 patients of which 3,065 were eligible for inclusion in the study. Thoracolumbar spine injury was diagnosed in 16.3% of the study cohort (n=499). Of those sustaining a thoracolumbar spine injury, 264 were deemed clinically significant with 77 patient’s requiring surgical stabilization and the remaining 187 treated with a thoracolumbar spine orthosis. The majority of the study cohort underwent CT evaluation of their thoracolumbar spine (93.3%), with only 6.3% being evaluated by plain films and 0.2% by MRI exclusively. In patients with a clinically significant T/L spine injury, a positive physical exam finding resulted in only a 78% sensitivity and a 73% specificity. When high risk mechanism of injury and/or age greater than 60 were included the sensitivity of predicting a clinically significant T/L spine injury rose to 98.9% with a specificity of 29%. Based on these results, the authors recommended any patient who is either unevaluable, has a positive physical exam, sustained a high risk mechanism of injury or is older than 60 years of age should undergo imaging to rule out an injury. All other patients may be considered low risk and can undergo clearance of their T/L spine by physical exam alone.
This is a very well designed study. By doing a power analysis up front, they ensured that the study would be adequately powered. In addition, they ensured that all imaging studies were independently reviewed for clinical significance as this can be somewhat of an objective standard. However, given that the study was observational, a validation study would need to be done to confirm these results and the authors acknowledge this in their limitation section. Despite the lack of long term follow up after discharge and the possibility that an injury could have been missed, this most likely would not have affected their results significantly. Finally, it is unclear from this study how many imaging studies could have been avoided using this clinical decision rule as the patients also underwent CT imaging of their torso as well and it was this imaging study that was used most often as the imaging study for the spine.
Article 2
Intracranial pressure monitoring and inpatient mortality in severe traumatic brain injury: A propensity score-matched analysis. Dawes AJ, Sacks GD, Cryer HG, Gruen JP, Preston C, Gorospe D, Cohen M, McArthur DL, Russell MM, Maggard-Gibbons M, Ko CY; Los Angeles County Trauma Consortium. J Trauma Acute Care Surg. 2015 Mar;78(3):492-501; discussion 501-2.
Despite recent studies showing the benefits of intracranial pressure monitoring, its use continues to remain controversial. Therefore Dr. Daws and colleague’s s sought to determine the impact of ICP monitor placement on inpatient mortality in patients sustaining severe traumatic brain injury. The authors performed a multi-institutional registry study involving 14 trauma centers in Los Angeles County as part of the Los Angeles County trauma Consortium. Over a 2 year period (2009-2010), 14 hospitals collected similar demographic and injury related variables standard in most registries. In addition, all hospital collected 9 additional comorbid conditions as well as 11 TBI specific variables (pupil reactivity, INR and 9 separate intracranial findings on CT). The primary outcome of interest was death during hospitalization for TBI and their primary variable of interest was ICP monitor placement within 72 hours of arrival to the emergency room. Inclusion criteria included blunt injury, GCS < 8 and abnormalities on initial head CT. The study cohort was stratified into patients that received an ICP monitor and those that did not. Binary statistics were preformed comparing these two groups. In addition, two separate multivariate logistic regression models were performed to identify predictors for ICP monitor placement and to identify predictors of mortality. Finally, to address the issue of selection bias, a propensity score-matched model was developed to determine the impact of ICP monitor placement on mortality in three specific subgroups (ISS > 25, GCS of 3 and both ISS >25 and GCS of 3).
Over the two year study period, 843 patients sustained a severe TBI with 2.5% (n=21) dying upon arrival to the emergency room. The remaining 843 patients comprised the study cohort. The mean age was 43 years, the median ISS was 26, median GCS was 3, 75% were male and 42% were Hispanic. The most common mechanism was fall (30.9%), most common HCT finding was subdural (65.7%), the majority had at least one reactive pupil (65%), almost half of the cohort had signs of increased ICP (47%) but only 46% of the study cohort underwent ICP monitoring despite all patients meeting brain injury foundation criteria. On univariate analysis, patients undergoing ICP monitor placement had significantly lower mortality (31% vs 46%, p < 0.001) compared to non-monitored patients. In addition, monitored patients were more likely to have an increased ISS, at least one reactive pupil or an intraparenchymal contusion on head CT but less likely to be older, coagulopathic or diagnosed with hypertension or alcoholism compared to those that were not monitored. On multivariate analysis, factors associated with monitor placement were presence of a subdural hematoma, intraparenchymal contusion or mass effect seen on CT. Factors negatively associated with ICP monitor placement included age > 65, alcoholism and an elevated INR. Results of the propensity score-matched model revealed that ICP monitor placement resulted in an 8.3% reduction in overall risk adjusted mortality. The authors concluded that their data support the use of ICP monitor placement in pts with severe traumatic brain injury.
This study adds to the growing debate regarding benefit of ICP monitor placement in patients sustaining a severe TBI. While the authors showed a protective effect on mortality by placing an ICP monitor, it is safe to say that the benefit obtained was not due from placement of the monitor itself but possibly the way in which the data from the monitor was used to change the patient’s care. Unfortunately, that data was not provided (i.e. use of hyperosmolar therapy or hyperventilation). In addition, no data was provided on functional outcomes of patients. Therefore, it is not possible to determine if the mortality benefit shown extends to the overall functional status of the patient in the long-term. Finally, the real question that needs to be addressed is in which population of patients is an ICP monitor helpful and how can the data obtained from this lead to improvements in overall patient care.
Article 3
A controlled resuscitation strategy is feasible and safe in hypotensive trauma patients: Results of prospective randomized pilot trial. Schreiber MA, Meier EN, Tisherman SA, Kerby JD, Newgard CD, Brasel K, Egan D, Witham W, Williams C, Daya M, Beeson J, McCully BH, Wheeler S, Kannas D, May S, McKnight B, Hoyt DB; ROC Investigators. J Trauma Acute Care Surg. 2015 Apr;78(4):687-97.
This study sought to assess the feasibility of and safety of controlled fluid resuscitation (CR) compared to standard resuscitation (SR) for early resuscitation of patients with traumatic shock. The authors hypothesized that CR would result in a significant reduction in overall quantity of fluid administered compared to SR without an increase in morbidity or mortality. They performed a prospective, randomized, out-of – hospital pilot trial via the Resuscitation and Outcomes Consortium (ROC) with patients being randomized in the field by EMS personnel. Inclusion criteria included both blunt and penetrating trauma, age > 15, SBP < 90 and absence of severe head injury. Patients in the CR group (n=97) only received a 250 ml bolus of NS if they had no radial pulse or a SBP < 70. Patients received an additional 250 ml bolus of NS to keep their SBP > 70 or to maintain a radial pulse. Treatment was continued through their hospitalization for up to 2 hours or until surgical control of hemorrhage could be obtained. Patients in the SR group (n=95) received an initial bolus of 2 L of NS and additional fluid was administered to maintain SBP > 110. Blood products were administered to both groups at any point in the treatment period at the discretion of the providers. The primary endpoint was early crystalloid volume and the primary safety endpoint was 24-hour mortality. Secondary outcomes included 24-hr fluid volumes, in hospital mortality, renal performance and resource utilization. The study was adequately powered to detect a true difference in fluid volumes. Both univariate and multivariate regression analysis was performed.
The groups were similar in respect to age, ISS, mechanism of injury and transport characteristics. Mean transport time was 15 minutes and mean total out of hospital time was 45 min. CR patients received significantly less crystalloid in the out-of-hospital setting and in the first 2 hours after hospital arrival compared to the SR patients. No blood products were administered in the prehospital setting to either group. The CR group received significantly more PRBC and total blood products during the time from hospital arrival to 2 hours compared to the SR group. However there was no significant difference between the groups with regards to total crystalloid, PRBC or total blood products at 24 hr. Despite the SR group receiving significantly more fluid compare to the CR group, there was no significant difference in admission vital signs or laboratory parameters. In patients with blunt trauma, the 24 hours mortality was lower in the CR group compared to the SR group (3% vs 18%). This difference was not seen in the penetrating group (9% vs 9%) or in overall in-hospital mortality. The authors concluded that a controlled resuscitation strategy is both feasible and safe for initial resuscitation of hypotensive trauma patients.
This was a well-designed and well executed study proving that randomized prehospital trials, although admittedly difficult to perform are in fact possible. In addition, the study alludes to the fact that in hospital mortality and possibly even 24 hour mortality may not be the ideal endpoints. Earlier time points such as 6 hour or even 2 hour mortality may be more appropriate. Finally, given that this was a pilot study, the authors acknowledge that only limited information can be obtained from this study. None the less, this study is in agreement with a growing body of literature showing less crystalloid is not harmful and in fact beneficial to patients who are in shock.
Article 4
Decreased mortality in traumatic brain injury following regionalization across hospital systems. Kelly ML, Banerjee A, Nowak M, Steinmetz M, Claridge JA. J Trauma Acute Care Surg. 2015 Apr;78(4):715-20.
The study sought to determine if regionalization of trauma care across hospital systems in Northern Ohio reduced mortality in traumatic brain injury. The authors performed a retrospective registry based study from 2008 -2012 looking at two distinct time points, care prior to regionalization (2008-2010) and care after regionalization (2100 – 2012) using a Northern Ohio Trauma system database. This database was constructed in 2010 and prospectively populated through 2012 and respectively populated back to 2008. The database included baseline demographic and outcome variables as well as all data elements required for the National Trauma Data Bank. Inclusion criteria consisted of age > 14 years old with a TBI based on standard ICD 9 codes. In addition to standard demographic and outcome variables, all neurosurgical procedures and included craniotomy, ventriculostomy, intracranial pressure monitoring and any other intracranial procedure performed in the operating room. The primary outcome was in-hospital mortality and secondary outcomes included length of stay, overall neurosurgical procedural rate and craniotomy rate. Univariate and multivariate regression analyses were performed comparing pre-regionalization to post-regionalization TBI outcomes.
During the study period, 11,200 patients with TBI were identified with 40% (n=4,507) pre-regionalization and 60% (n=6,713) post-regionalization. Among all TBI patients, there was no difference between the groups in median ISS, median GCS, and % of patients with head AIS scores > 3, length of stay or overall number of neurosurgical procedures performed. However, the number of TBI air transports increased from 18% to 20% and TBI admissions to Level I trauma centers increased from 36% to 46%. The number of craniotomies increased from 2% to 3% and the mortality decreased from 6.2% to 4.9%. At the level I center, the craniotomy rate increased from 4% to 6% and the mortality rate decreased from 12% to 8% between the two time periods. Overall, patients who sustained severe TBI’s (head AIS > 3) had an increased craniotomy rate from 6% to 8% and a decreased mortality rate from 19% to 14%. For those patients admitted to a Level I trauma center, the craniotomy rate increased from 7% to 10% and the mortality rate decreased from 21% to 14%. Logistic regression analysis identified an independent effect on survival for both overall TBI patients and severe TBI patients during the post-regionalization period. Patients were 24% and 28% more likely to survive in the post regionalization period compared to the pre regionalization period respectively. This study is consistent with other studies showing an improvement in survival with regionalization of trauma care (i.e. getting the right patient to the right place in the right time). Though most of those studies looked at overall trauma patients and not specifically brain injured patients. However, given the retrospective nature of the study and the 4 year time period, there remains a possibility that improvements in care over time and not regionalization may have played some role in these improvements.
|