You Cannot Go Home: Routine Concussion Evaluation Is Not Enough
Traditional care of mild traumatic brain injury (MTBI) is to discharge patients from the emergency department (ED) if they have a Glasgow Coma Score (GCS) of 15 and a normal head computed tomography (CT) scan. However, this does not address short-term neurocognitive deficits. Our hypothesis is that a notable percentage of patients will need outpatient neurocognitive therapy despite a reassuring initial presentation. This is a retrospective review of patients with MTBI at an urban Level I trauma center. Inclusion criteria were a diagnosis of MTBI in patients 14 years old or older, GCS 15, negative head CT scan, a completed neurocognitive evaluation, blunt mechanism, and no confounding psychiatric comorbidities. Six thousand thirty-two patients were admitted over 18 months. Three hundred ninety-five patients met inclusion criteria. Average age was 38 years (range, 14 to 93 years), 64 per cent were male, and mean Injury Severity Score (ISS) was 8.1. Forty-one per cent were cleared for discharge without follow-up. Twenty-seven per cent required ongoing neurocognitive therapy. Three per cent were deemed unsafe for discharge home. Of the patients cleared for discharge, 88 per cent had positive/questionable loss of consciousness (LOC), whereas 81 per cent who required additional therapy had positive/questionable LOC (P = 0.20). Age, gender, ISS, and alcohol use were compared between the groups and not found to be statistically different rendering them poor predictors for appropriate discharge from the ED. A surprisingly high percentage (27%) of patients who would have met traditional ED discharge criteria were found to have persistent deficits after neurocognitive testing and were referred for ongoing therapy. We provide evidence to suggest that we should take pause before discharging patients with MTBI without a cognitive evaluation.
TRAUMATIC BRAIN INJURY (TBI) affects almost two million people in
The majority of patient complaints concerning MTBI are physical. These symptoms are often subtle and described as headache, dizziness, nausea, and difficulties with memory and concentration. The symptoms occur early and are thought to resolve spontaneously.7-10 A recent guideline from the
A past study on hospitalization use for MTBI suggested that 80 per cent of patients evaluated for TBI do not need to be admitted to the hospital and the overtriage of patients with MTBI is a tremendous burden to healthcare costs.13 However, this study did not look at the outcomes after discharge and included no formal neurocognitive assessment. In our modern healthcare environment of resource conservation, we set out to evaluate our own management strategies concerning MTBI. In the current work, we seek to determine the safety of discharging patients with MTBI after neurocognitive evaluation and we propose a strategy for outpatient follow-up for patients with MTBI. Our hypothesis is that even the most minimally brain-injured patients are at risk for requiring further therapy when a formal neurocognitive evaluation is performed by a trained professional in the early postinjury period.
Methods
These patients had a formal neurocognitive evaluation performed by a certified trauma speech therapist within 48 hours of admission. The neurocognitive evaluations were performed by one of three full-time or six part-time certified speech therapists who collaborate to maintain consistency in the evaluation. After a complete patient history is taken, the therapists evaluate each patient in all of the following areas: orientation, awareness, biographical memory, shortterm memory, long-term memory, attention/concentration, organization, sequencing, expressive language, receptive language, problem-solving/reasoning, math management skills, and reading (Table 1). The recommendations of the therapists are developed based on the patient's overall performance during the interview in regard to the aforementioned areas as well as the cognitive requirements that would be demanded from the patient on return to home, work, or school, which is unique to each patient. Standardized tests used by the therapists to formulate the initial recommendations include the Cognitive-Linguistic Quick Test (CLQT) (Pearson,
Charts were reviewed and data were collected with regard to the following variables: patient age, gender, mechanism of injury, psychiatric diagnoses, results of cognitive evaluation/speech therapist recommendations including recommended disposition, LOC, GCS score, Injury Severity Score (ISS), CT head findings, additional injuries, and results of alcohol and drug screen, if available.
Descriptive statistics were used to describe numerical data relative to patient characteristics. For categorical variables (e.g., gender, injury mechanism), outcomes were compared across categories by x2 tests and for normally distributed, continuous variables, Student's t test was used to compare means between independent groups. Differences were considered statistically significant for P values < 0.05.
Results
The trauma registry produced 6032 trauma patients who were admitted at our trauma center over an 18month period from
We summarized the speech therapy recommendations into two major groups: Group 1,patients who could go home without ongoing therapy-the ''no therapy'' group; and Group 2, patients who could either be discharged with ongoing therapy or who were deemed unsafe for discharge-the ''further therapy/ unsafe for home'' group. The ''no therapy'' group includes all patients who had a recommendation from the therapist to be discharged without ongoing therapy, with or without 24-hour companion supervision. The ''further therapy/unsafe'' group includes patients who required ongoing outpatient neurocognitive therapy (they would be discharged with a prescription for outpatient speech therapy in addition to office followup), required inpatient rehabilitation, or were unsafe for discharge at the time of evaluation.
Average age was 38.0 years (range, 14 to 93 years; standard deviation [SD] 18.7) and 64 per cent were male with a mean ISS of 8.1 (SD 5.8). Patients sustained a variety of blunt mechanisms of trauma with motor vehicle collision being the most common (51%) (Fig. 2). Alcohol levels were measured on 147 patients with the mean value of 0.083 g/dL (SD 0.110 g/dL; range, 0 to 0.423 g/dL). Neurocognitive evaluations were completed when patients were clinically sober, because repeat alcohol levels were not routinely measured. The formal neurocognitive evaluation was completed on hospital Day 1 in 54.2 per cent of all included patients with the remainder completed by 48 hours after admission.
After a formal neurocognitive evaluation, we found that 162 patients (41%) were cleared for discharge without need for follow-up. An additional 123 patients (31%) received the recommendation to be discharged to home with 24-hour supervision by a responsible adult until they are seen in follow-up. Two patients (0.5%) were specifically advised to have supervision but no specific recommendation for follow-up. These three sets of patients constitute Group 1-the ''no therapy'' group. An additional 73 patients (18%) of patients were given a prescription for outpatient speech therapy, 24 patients (6.1%) were advised to have outpatient therapy and supervision, and finally 11 patients (2.8%) were deemed ''unsafe for discharge.'' The latter three sets constitute Group 2-the ''further therapy/ unsafe'' group (Table 2).
Of the patients cleared for discharge home, 88 per cent had positive or questionable LOC, whereas 81 per cent who required referral for additional therapy had positive or questionable LOC (P 4 0.20). Comparisons were made between the ''no therapy'' and ''further therapy/unsafe'' groups with respect to gender, LOC, alcohol level, age, and ISS scores. No statistically significant differences were found between the two groups with respect to any of these variables (Table 3). This analysis yielded no statistically significant independent predictors of ''unsafe discharge.''
We individually reviewed and summarized the characteristics of the patients who were not cleared for discharge. When this group of 11 patients was analyzed against all other patients, we saw a trend toward older age in the group not safe for discharge (NSFD), but this did not reach statistical significance (Table 4). Mean age in NSFD was 54.2 years (SD 28.9); mean age in all others was 37.5 years (SD 18.2) (P 4 0.09).
Discussion
We found that more than one in four (27%) of our patients, after initial evaluation, would be referred for ongoing neurocognitive therapy. This overall concerning analysis was the first indication that none of the usual parameters used for evaluation of a trauma patient in the ED or trauma bay such as negative head CT and GCS 15 are able to accurately predict those patients with MTBI who are safe for discharge without need for neurocognitive evaluation and follow-up. An additional 31 per cent of our patients were discharged with instructions to remain with a responsible adult 24 hours a day until seen in follow-up and although this supervision was not considered a formal ''intervention'' in our study, the principle remains that unqualified discharge of these patients does not acknowledge the subtle deficits they suffer. We sought to find a predictor that could be used in the ED such as positive LOC or older age that would predict the need for neurocognitive evaluation resulting in the need for ongoing therapy. However, we could find no such predictor. We conclude that any patient diagnosed with MTBI, with or without LOC, will benefit from immediate (within 48 hours of admission) formal neurocognitive evaluation by a trained professional who can identify the need for ongoing therapy or follow-up.
In reviewing the 11 cases of patients who received the preliminary recommendation of ''unsafe for discharge,'' we note a few patients who initially were not recognized to have dementia. These patients presented to the trauma team alert and oriented to basic questioning and went on to have a negative CT of the head. However, as a more thorough evaluation went on, the patients were recognized to have baseline deficits. Whether their deficits were truly a result of MTBI or the result of a longstanding neurocognitive disorder cannot be definitively determined from this retrospective chart review. In at least two of these 11 cases, an underlying disorder was brought to light. Patients who have baseline dementia and then sustain MTBI may be at higher risk for complications such as falls or errors in taking medication than patients without dementia. We acknowledge that this question cannot be answered by the present study because, from a practical standpoint, many of our trauma patients with dementia are not seen within 48 hours of admission, but rather just before discharge, often later than 48 hours as a result of comorbid conditions, to aid in decision-making for disposition. Thus, many patients with dementia would have been excluded from our study by the 48-hour criteria. However, our recommendation stands that even a patient with a normal GCS who has sustained head trauma benefits from a neurocognitive evaluation, which can reveal subtle underlying disorders and aid in decision-making for safe discharge.
The extent of neurocognitive deficits found in the 11 patients not safe for discharge is noted in Table 4 and is reflected in the proposed disposition by the early speech therapy evaluation. However, the actual disposition of the patient was determined on discharge and took into account the recommendations of the speech therapist, the physical and occupational therapists, consulting physicians, patient and family preferences, insurance coverage options, and the final decision by the trauma attending. The disposition listed in Table 4 references the independent recommendations of the speech therapist and captures only the extent of care needed as a result of early (within 48 hour of injury) neurocognitive deficits. The early recommendation by the speech therapist may not be the patient's actual disposition, but the take home point is that in the absence of concomitant injuries requiring prolonged hospitalization, the neurocognitive deficits alone would have precluded safe discharge to home in the early period.
Most patients who sustain MTBI will recover within three months,6-11 but up to 15 to 40 per cent of patients will experience symptoms and report persistent deficits for a year or longer.6, 14, 15 Previous studies have identified that even very early in the post-MTBI period, subtle but discernible neurocognitive deficits exist compared with control patients and that early identification may prompt early treatment.16 Early intervention, in the first 48 hours postinjury, provides an opportunity to evaluate and document baseline symptoms and search for treatable causes of symptoms such as dizziness with medications and vestibular therapy.14 The 48-hour cutoff is an arbitrary decision by the authors but in practice reflects the typical timeline of our patients who are admitted for MTBI, with or without additional injuries, who are evaluated early and released home rapidly on hospital Day 1 or 2, depending on the hour of day on admission. We chose 48 hours because we wanted to study the deficits found very early in the MTBI course that could affect decisions for return to work, school or driving, or disposition immediately after injury. We believe that our results, 27 per cent of our patients with MTBI needed some type of intervention, substantiate our claim that even the very early deficits of MTBI deserve attention and should affect disposition. From a very practical standpoint, the benefits of early evaluation and intervention may not be easily measured, but using the ''teachable moment'' immediately postinjury to educate the patient and family about MTBI, what to expect, what to watch for, and providing resources for follow-up allow us to re-evaluate the patient, make recommendations for safe return to work, school and driving, and connect the patients to resources such as a neurologist, neuropsychiatrist, or psychotherapists as indicated.14 Early intervention gets patients with MTBI ''hooked into the system.'' Since we have completed this study, we have a heightened awareness of persistent MTBI symptoms and we have formalized our approach to early screening and capture of patients with MTBI in the acute setting.
We recognize several limitations of our study. First, because this is a retrospective study, there were not strict inclusion criteria at the time of admission for the patients and thus we realize that although attempts were made to gather data on a homogenous group of patients, there may be variability resulting from individual surgeon practice in admitting patients with MTBI. Patients were admitted for a neurocognitive evaluation but specific data regarding previous MTBI, sleep disturbances, litigation, or extenuating social issues were not routinely obtained and documented in the history, yet these factors are known to contribute to the severity of MTBI and ability to recover from the current injury. Without these circumstantial data, we note we must be careful in making assumptions from our data that would broadly apply to all patients.
We recognize that there could be bias toward the exclusion of patients who appeared to have normal mental status and had a negative head CT and were discharged home before a speech therapy evaluation, although our accepted general practice was to admit patients with LOC or a diagnosis of concussion during the time reviewed in our study and to order the speech therapy evaluation. It was not our practice to require patients to leave against medical advice from the ED after an MTBI when they declined admission, so therefore patients who elected to leave before evaluation would be excluded from the study. We recognize that of the 901 patients who were diagnosed as having MTBI, less than 50 per cent went on to have a formal neurocognitive evaluation for the variety of reasons mentioned previously. This provides a significant bias that perhaps the patients who were admitted and received an evaluation (the 395 in the study) were more symptomatic or were a ''captive audience'' as a result of concomitant injuries and were biased to actually have the evaluation completed. We take caution in the interpretation of our results for this reason. However, the 413 patients (413 no evaluation/901 with MTBI [46%]) who did not have a neurocognitive evaluation matched our study cohort with a diagnosis of MTBI, GCS 15, and negative CT of the head and we would therefore presume a similar outcome in this group had they received the evaluation. Indeed, we believe that it is likely we may have missed the opportunity to engage those 413 patients in neurocognitive evaluation and follow-up. This presumption will be tested in the next phase of our study, which will include a prospective MTBI registry including documentation of immediate symptoms and results of the first neurocognitive evaluation (within 48 hours) as well as longitudinal follow-up with subsequent neurocognitive evaluations, duration of symptoms, how long the patients required therapy, and when and to what degree they are able to resume their normal daily activities.
We acknowledge that the exact methods used in the neurocognitive evaluation were not prescribed because this is a retrospective review and were not precisely the same from patient to patient. Our speech therapists most often use the M°CA test but will vary the testing based on the needs of the patient and their current abilities. For example, a patient who is right-handed but has a right-sided fracture that is splinted will not be asked to participate in testing that requires writing tasks. Additionally, a patient with complex work responsibilities such as an airline pilot or surgeon would be tested for attention, concentration, and processing deficits in a different manner than the patient whose daily tasks require less intense problem-solving or risk. This testing was left at the professional discretion of the speech therapist performing the evaluation. Although the therapists as a group routinely review their practice and strive for consistency, there is inherent interrater variability in the testing. Variability in neurocognitive testing is a weakness in the MTBI literature as a whole because most institutions do not perform the testing in the same manner. One study using a national survey found a variety of professionals completing the testing from speech language pathologists and occupational therapists to nurses, case managers, and attending physicians17 who surely perform neurocognitive testing in a variety of manners, not to mention variability between examiners.
The present study has prompted a change in our overall practice management of patients with MTBI. The study demonstrates that 27 per cent of our patients who were evaluated by a specialist were found to have symptoms and subtle markers of MTBI that would prompt ongoing therapy. We realize that now, more than ever, reimbursement for hospital admission for patients with isolated MTBI is not feasible and strains the resources of an already busy trauma service. To admit every patient with MTBI is not possible, but the results of our study have heightened our awareness of subtle MTBI symptoms. We continue to admit patients with any complaints of MTBI such as headache, photophobia, dizziness, or nausea. We strive to provide immediate neurocognitive evaluation as soon as possible. In fact, during the daytime, our speech therapists perform the evaluation in the ED to aid in decisionmaking about the need for admission versus discharge. For the patients who refuse admission, are unable to have a neurocognitive evaluation before discharge, or whose symptoms are controlled and otherwise do not requireadmission,wehavemadeamoreconcertedeffort to engage these patients in follow-up (Fig. 3). We have instituted a ''concussion clinic'' as part of our outpatient follow-up for trauma patients. The clinic is available every Friday and is staffed by our trauma nurse practitioner, a nurse, and a speech therapist (Fig. 4). Our study has shed light on the fact that many patients who we would have simply ''cleared'' will need supervision or therapy and we are providing a safety net for those patients with our growing ''concussion clinic.''
We acknowledge there has been much recent interest in sports-related concussions and a great deal has been published in the sports medicine literature on this topic. However, in our patient population, we did not see patients with sports-related injuries (Fig. 2) and therefore we hesitate to extrapolate our results to this subset of patients with MTBI. With an average age of 38 years, our patients need to return to work, to school, to driving, and to parenting or caregiving to aging family. Return to play places patients back into immediate risk of repeat concussion (football tackle, soccer head ball, etc.), whereas return to work/home often does not place patients back into a risk for repeat concussion and we feel that the recommendations for return to activity should not be ''one size fits all.'' We question the translation of the sports medicine experience with MTBI to our patient population, and vice versa, and believe there remains a need to study MTBI in the general population and across the spectrum of ages.
Conclusion
A surprisingly high percentage (27%) of our patients who would have met traditional discharge criteria from the ED were found to have persistent deficits after neurocognitive testing and were referred for ongoing therapy. In our review of all patients who had the diagnosis of MTBI, we found that nearly half (46%) were discharged without formal evaluation and we missed the opportunity to engage them in follow-up. We provide evidence to suggest that we should assess these patients with MTBI before discharging them including either a formal cognitive evaluation or planned follow-up. Premature discharge places patients with MTBI at risk for ongoing unrecognized and untreated neurocognitive deficits.
Finally, the current study is focused on the immediate needs of the patient with MTBI and appropriate initial triage and follow-up recommendations. Several questions remain unanswered by the present study including an understanding of the demographics of patients who follow up, what percentage of them continue to have symptoms, what the psychosocial impact of the MTBI has on them, the results of ongoing therapy, and the duration of therapy required. All of these questions are beyond the scope of this retrospective study but form the basis for the nature of our clinic follow-up and subsequent prospective studies.
Acknowledgments
We thank
Speech therapists: Rachael Crnarich, MA, CCC/SLP;
REFERENCES
1. Kraus JF, McArthur DL, Silverman TA, et al. Epidemiology of brain injury. In: Narayan RK, Wilberger JE, Povlishock JT, eds. Neurotrauma.
2. Thornhill S, Teasdale GM, Murray GD, et al. Disability in young people and adults one year after head injury: prospective cohort study. BMJ 2000;320:1631-5.
3. Yang C, Tu Y, Hua M, et al. The association between the postconcussion symptoms and clinical outcomes for patients with mild traumatic brain injury. J Trauma 2007;62:657-63.
4. Livingston DH, Lavery RF, Passannante MR, et al. Emergency department discharge of patients with a negative cranial computed tomography scan after minimal head injury. Ann Surg 2000;232:126-32.
5. Taheri PA, Karamanoukian H, Gibbons K, et al. Can patients with minor head injuries be safely discharge home? Arch Surg 1993;128:289-92.
6. Barbosa RR, Jawa R, Watters JM, et al. Evaluation and management of mild traumatic brain injury: an
7. Dikmen S, McLean A, Temkin N. Neuropsychological and psychosocial consequences of minor head injury. J Neurol Neurosurg Psychiatry 1986;49:1227-32.
8. Warrier EM, Rourke BP, Velikonja D, et al. Subtypes of emotional and behavioral sequelae in patients with traumatic brain injury. J Clin Exp Neuropsychol 2003;25:904-17.
9. Haboubi NHJ, Long J, Koshy M, et al. Short-term sequelae of minor head injury (6 years experience of minor head injury clinic). Disabil Rehabil 2001;23:635-8.
10. Ponsford J, Cameron P, Fitzgerald M, et al. Long-term outcomes after uncomplicated mild traumatic brain injury: a comparison with trauma controls. J Neurotrauma 2011;28: 937-46.
11. Peterson SE, Stull MJ, Collins MW, et al. Neurocognitive function of emergency department patients with mild traumatic brain injury.
12. Dischinger PC, Ryb GE, Kufera JA, et al. Early predictors of postconcussive syndrome in a population f trauma patients with mild traumatic brain injury. J Trauma 2008;66:289-97.
13. Thurman D, Guerrero J. Trends in hospitalization associated with traumatic brain injury. JAMA 1999;10:954-7.
14. Kushner D. Mild traumatic brain injury. Arch Intern Med 1998;158:1617-24.
15. Carroll LJ, Cassidy JD, Peloso PM, et al. Prognosis for mild traumatic brain injury: results of the
16. Peterson SE, Stull JM, Collins MW, et al. Neurocognitive function of emergency department patients with mild traumatic brain injury.
17. Blostein P, Jones SJ. Identification and evaluation of patients with mild traumatic brain injury: results of a national survey of Level 1 trauma centers. J Trauma 2003;55:450-3.
From *
Presented as a poster at the 27th Annual
Address correspondence and reprint requests to
Reasons for 30-day Postoperative Readmissions for Medicare Patients at a Community-based Teaching Hospital
Advisor News
Annuity News
Health/Employee Benefits News
Life Insurance News