Trauma Outcome Following Moderate Traumatic Brain Injury Todd W. Vitaz, M.D., Jennifer Jenks, B.A., George H. Raque, M.D., and Christopher B. Shields, M.D. Department of Neurological Surgery University of Louisville School of Medicine and University Health Care, Louisville, Kentucky Vitaz TW, Jenks J, Raque GH, Shields CB. Outcome following moderate traumatic brain injury. Surg Neurol 2003;60:285–91. BACKGROUND Little is known about the outcome following moderate traumatic brain injury (TBI) (GCS 9-12). Most patients regain consciousness; however, the full magnitude of long-term cognitive and functional deficits is unknown. METHODS We conducted a prospective observational study evaluating the outcome of patients suffering moderate TBI between October 1995 and March 1998. Long-term outcome was assessed by telephone interviews. RESULTS A total of 79 consecutive patients were included. Average length of ICU and total hospital stay was 9.1 and 15.8 days respectively. The median GCS at 24 hours was 10 with 67% improving to GCS 15 by time of discharge. The presence of multisystem trauma did not affect outcome; however, age ⱖ45, initiation of enteral feeding after postinjury day 4 and the presence of pneumonia were all associated with longer lengths of stay and increased complication rates. Fifty-six (71%) patients were contacted for follow-up at an average of 27.5 months. GOS scores were 5 in 44%, 4 in 41%, 3 in 9%, 1 in 6%. Seventy-four percent of patients employed premorbidly returned to full-time work. Questions regarding cognitive and functional status revealed significant problems in the majority of patients. CONCLUSIONS Pneumonia, age ⱖ45 years and a delay in initiation of enteral feeding all increased the duration of acute care hospital stay following moderate TBI. In addition, cognitive, emotional, and functional problems following such injuries are extensive and long lasting. Physicians must be knowledgeable of these long-term sequela so they can provide the appropriate support and treatment to these patients. © 2003 Elsevier Inc. All rights reserved. KEY WORDS Moderate traumatic brain injury, outcome, neurologic recovery, clinical pathway. Address reprint requests to: Dr. Todd W. Vitaz, 210 East Gray St., Suite 1102, Louisville, KY. Received July 17, 2001; accepted March 6, 2003. © 2003 Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010 –1710 T raumatic brain injury (TBI) remains a major health problem in the United States, accounting for 500,000 hospital admissions and 10,000 deaths each year. Among children and young adults, TBI is the leading cause of death and disability [5]. Between 30 and 50% of TBIs are of moderate Glascow Coma Scale (GCS 9-12) or greater severity (GCS 3-8) [6]. Patients with moderate closed head injuries pose a case management challenge. This group of patients represents a heterogeneous population with significant variability in terms of trauma severity, hospital course, neurologic recovery, and sequela. In addition there is a paucity of research concerning this level of head injury and subsequently there are no definitive treatment guidelines. The aim of this study was to evaluate the outcome of patients who had sustained moderate TBI with respect to both short- and long-term outcomes including neurologic recovery, length of hospital stay, and complication rates, as well as long-term cognitive and functional deficits. Methods A prospective database was compiled for all patients admitted to the University of Louisville Hospital, Louisville, Kentucky, with moderate closed head injury (GCS 9-12) between October 1995 and March 1998. Information regarding patient demographics, associated injuries, length of stay (LOS), and complications were collected. All patients were treated following criteria developed as part of our moderate TBI clinical pathway (Figure 1). This pathway was developed at our institution to facilitate management of patients with head injuries. This pathway provides general guidelines for patient management, and similar pathways have been shown to help standardize and improve the quality of patient care [2,7,10,16]. 0090-3019/03/$–see front matter doi:10.1016/S0090-3019(03)00378-1 286 Surg Neurol 2003;60:285–91 1 Vitaz et al University of Louisville moderate closed head injury clinical pathway. The data were analyzed for the entire group and then a subset analysis was performed looking for parameters that either prolonged hospital stay or decreased neurologic recovery. Statistical analysis was conducted using student’s t test, ␹2, correlation coefficients and multiple regression analysis with a p value of ⬍0.05 considered significant. For the ␹2 analysis admission, GCS was divided into two groups [9 –10,11–12], intensive care unit (ICU) LOS was divided into two groups (⬍9, or ⱖ9 days), and pneumonia was divided into two groups (presence or absence). Moderate CHI 1 Surg Neurol 287 2003;60:285–91 Telephone Interview Questions Regarding Activities of Daily Living What type of assistance (none, minimal, full) do you require to complete the following? 1. Getting into and out of bed 2. Using the bathroom 3. Bathing 4. Personal hygiene (shaving, brushing teeth, combing hair) 5. Dressing 6. Cooking meals 7. Cleaning home 8. Shopping for groceries Long-term follow-up was obtained by a telephone interview conducted in February and March 1999. Information included the patient’s overall condition, functional level, and work status. This information was then used to determine the Glasgow Outcome Scores (GOS) and the percentage of activities of daily living (ADLs) that each patient was able to perform (Tables 1 and 2) [8]. In addition patients were asked generalized questions regarding persistent symptoms or complications, as well as memory and cognitive functions (Table 3). Patients were also given the opportunity to make any additional comments regarding their recovery. Results OUTCOME AT TIME OF HOSPITAL DISCHARGE Seventy-nine consecutive patients with moderate TBI were included in this study. No deaths occurred during the acute in-patient hospitalization. The average age at the time of injury was 37.8 (⫾17.2) years with a median 24-hour GCS of 10 (Figure 2). Pneumonia occurred in 33% of the patients. The average (range; SD) length of hospital and ICU stay and the number of ventilator days were 15.8 (4 –58; ⫾9.9), 9.1 (2–36; ⫾6.6), and 5.5 (0 –31; ⫾4.4) days, respectively. At the time of discharge 67% had improved to GCS 15 and 27% to GCS 14 (Figure 2). Sixty-one 2 3 Telephone Interview Questions Regarding Mental Functioning Do you experience or suffer from any of the following on a regular basis, and if so how often (daily, weekly, monthly)? 1. Headaches 2. Concentration problems 3. Short or long-term memory problems, such as remembering names or telephone numbers or misplacing items on a regular basis 4. Difficulty focusing attention 5. Mood swings or unexpected changes in emotion 6. Dizziness 7. Fatigue 8. Decreased vision or blurred/double vision 9. Problems speaking 10. Problems walking 11. Decreased hearing 12. Weakness 13. Seizures If yes, are you currently taking antiepileptic medication? Did you have a seizure disorder prior to your head injury? percent of patients were discharged to an in-patient rehabilitation facility, 28% were discharged home, 9% to a nursing home, and 3% were transferred to another acute care facility. A subset analysis was performed using the student’s t test to evaluate the impact of multisystem trauma, injury severity score (ISS), occurrence of pneumonia, the patient’s age, and timing of initiation of enteral feeding. None of these parameters were found to affect neurologic recovery at time of discharge. However, pneumonia, age ⱖ45 years, and delay in enteral feeding past postinjury day (PID) 4 all led to significantly increased length of hospital and ICU stay and length of mechanical ventilator dependence (Tables 4 –7). Admission GCS and ISS were not found to be correlated with outcome at discharge (Table 8). There was a trend Glasgow Outcome Scoring SCORE DEFINITION 5 4 3 2 1 Normal daily function Moderate disability, disabled but independent Severe disability, unable to care for self Persistent vegetative state Dead 2 GCS at admission. 288 Surg Neurol 2003;60:285–91 4 Vitaz et al Isolated TBI Versus Multisystem Trauma ISOLATED (N ⴝ 37) MULTITRAUMA (N ⴝ 42) 38.6 ⫾ 18.9 19.8 ⫾ 6.6 10 15 15.7 ⫾ 12.6 days 8.6 ⫾ 7.4 5.5 ⫾ 5.8 29.7% 4 37.1 ⫾ 15.8 25.6 ⫾ 7.3 10 15 15.9 ⫾ 6.9 days 9.5 ⫾ 5.8 5.5 ⫾ 2.8 35.7% 4 Age ISS Median GCS at 24 hours Median GCS at discharge Hospital LOS ICU LOS Ventilator days Pneumonia Median GOS* p ⫽ 0.005 *Measured at time of last follow-up. in 24 (44%), 4 in 22 (41%), 3 in 5 (9%), and 1 in 3 (6%) patients. Thirty-eight of the 51 patients worked full-time before their injuries. Of these 28 (74%) returned to full-time employment and an additional 2 (5%) returned to part-time work. The remaining 8 (21%) patients had either retired or were unable to return to work. In addition, there were 5 patients who did not work full-time before their injuries but were working at the same level at the time of last followup. Most patients could function independently with 48 (94%) being able to perform greater than 75% of their ADLs without assistance. However, a significant number of patients experienced persistent cognitive or functional problems (Table 9). toward an increased incidence of pneumonia in patients who received enteral feeding past PID 4; however, this did not reach statistical significance (p ⫽ 0.9). In addition, Pearson correlation and multiple regression analysis were conducted to determine the impact of these variables on the length of ICU stay (Figure 3). Univariate analysis showed a significant relationship between patient age, timing of enteral feeding, and number of days of mechanical ventilation with length of ICU stay (r ⫽ 0.31, 0.24, 0.83, respectively; p ⬍ 0.05); ISS was not found to be correlated (r ⫽ 0.24, p ⬎ 0.05). Multivariate regression analysis also confirmed these results. Once again age, timing of feeding, and ventilator days reached significance (R2 ⫽ 0.745, p ⬍ 0.05), while ISS was not found to be correlated with ICU length of stay. Discussion LONG-TERM OUTCOME Fifty-six of the 79 (71%) patients were contacted; the remainder of the patients could not be contacted because of a change in address since the time of last follow-up or a lack of a telephone. Two patients refused to participate and another three patients died since the time of hospital discharge (two of related causes and one of an unrelated cause) leaving 51 patients to participate in the questionnaire. Average follow-up was 27.5 months postinjury (range 16.5– 41.7 months). GOSs were 5 The short-term outcome following moderate closed head injury is quite unpredictable. In our group of 79 patients, the average length of acute hospital stay was 15.8 days. These patients experienced progressive neurologic improvement throughout their hospital stay as evidenced by their increase in GCS. The use of the clinical pathway helped to provide patients with the highest standard of care and facilitated communication between the numerous health care providers. This pathway also ensured that the ancillary services such as physical and 5 Effect of Pneumonia on Outcome PTS Age ICU LOS Hosp LOS Ventilator days Median GOS* *Measured at time of last follow-up. WITHOUT (N ⴝ 53) 35.3 ⫾ 16.8 years 6.8 ⫾ 3.8 days 12 ⫾ 5.2 4.1 ⫾ 2.5 5 PTS WITH (N ⴝ 26) 43 ⫾ 17.3 years 14.1 ⫾ 8.4 days 23.8 ⫾ 12.4 8.5 ⫾ 5.9 4 p ⬍ 0.0003 p ⬍ 0.0001 p ⬍ 0.001 Moderate CHI 6 Surg Neurol 289 2003;60:285–91 Effect of Age on Outcome <45 Age ICU LOS Hosp LOS Ventilator days Pneumonia Median GOS* YEARS (N ⴝ 55) 28.5 ⫾ 8.4 years 7.5 ⫾ 5.1 days 13.7 ⫾ 7.6 4.5 ⫾ 2.7 27% 5 >45 YEARS (N ⴝ 24) 59 ⫾ 12.9 years 13 ⫾ 8.0 days 20.8 ⫾ 12.7 7.9 ⫾ 6.4 46% 4 p ⬍ 0.0001 p ⬍ 0.005 p ⬍ 0.02 p ⬍ 0.02 *Measured at time of last follow-up. occupational therapy were involved with patient care as early as possible. Finally, the use of this type of clinical pathway reduces the disparity in patient care, thus providing more consistent care to a large group of patients and allows more accurate analysis of subgroups [2,10,16]. The GCS at 24 hours was used to prevent inconsistencies secondary to inadequate resuscitation or alcohol and other drugs. Length of ICU stay was chosen as the dependant variable for the linear and multiple regression analysis because this parameter is the most reliable indicator of the patient’s recovery. The total hospital LOS can be artificially prolonged by various factors, especially the availability of rehabilitation beds. The presence of multisystem trauma did not significantly affect patient outcome in terms of length of stay or short-term neurologic recovery. ISS was not correlated with outcome; however, the data for this parameter was unavailable for approximately 20% of the patients, which may have influenced the results of this analysis. The percentage of patients that required mechanical ventilation (96%) and the overall length of ventilation (avg. 5.5 days) may be somewhat higher than expected for patients with this severity of injury for several reasons. The first is an extremely low threshold of our emergency medical service personnel to routinely intubate patients with altered mental status in the field or before air transport from an outside facility to ensure an adequate 7 airway and prevent hypoxia. Second is the extremely high rate of alcohol-related injuries that may act to further depress the patient’s level of consciousness and may increase the risk of aspiration in the field. The third factor is the high rate of tobacco abuse in our community that decreases the patient’s pulmonary reserve and probably predisposes them to pneumonia. Finally, the mechanism of injury in our group may differ significantly from other large inner city trauma populations where the primary mechanism tends to be related to falls and assaults. Our trauma center serves a moderatesized metropolitan area with a large surrounding rural community and therefore high-speed motor vehicle crashes with the associated incidence of multisystem injuries predominates. As expected, the occurrence of pneumonia significantly prolonged all aspects of hospital stay, especially the length of time for mechanical ventilation. Aggressive pulmonary toilet and early mobilization have been shown to help decrease the occurrence of pneumonia. In addition, we recommend early tracheostomy in any mechanically ventilated patient who does not show signs of weaning in the first 4 hospital days. This regimen improves pulmonary function, facilitates pulmonary toilet, patient mobilization, and ventilator weaning with subsequent decreases in episodes of pneumonia in other groups of trauma patients [1,9,13]. A delay in the initiation of enteral feeding also in- Effect of Timing of Initiation of Enteral Feeding Age ICU LOS Hosp LOS Ventilator days Pneumonia Avg. PID fed Median GOS* *Measured at time of last follow-up. <PID 4 (N ⴝ 59) > PID 4 (N ⴝ 20) 37.4 ⫾ 16.4 years 7.9 ⫾ 5.7 days 14.2 ⫾ 9.2 4.9 ⫾ 4.6 29% 2.6 ⫾ 1.0 4 39.1 ⫾ 19.8 years 12.9 ⫾ 10.5 days 20.9 ⫾ 10.5 7.2 ⫾ 3.4 45% 6.3 ⫾ 1.8 4 p ⬍ 0.02 p ⬍ 0.02 p ⬍ 0.02 p ⬍ 0.0001 290 Surg Neurol 2003;60:285–91 8 Vitaz et al 9 Effect of GCS on Outcome GCS* (#) ICU LOS MEDIAN DISCHARGE GCS (RANGE) 9 10 11 12 14.6 days 8.7 11.5 9.5 15 (13–15) 15 (11–15) 15 (14–15) 15 (14–15) (30) (31) (14) (4) MEDIAN GOS (RANGE) 4 (1–5) 4 (1–5) 5 (1–5) 4 (3–5) *GCS 24 hours postadmission. creased length of stay and the incidence of pneumonia, presumably related to the development of malnutrition and associated immunosuppression [3,4,11,14]. We recommend early feeding either orally or through supplemental feeding devices (overall average PID 3.5 ⫾ 2.0); however, this may not always be possible because of associated injuries. In terms of long-term outcome the majority of patients made a good functional recovery with almost 90% having GOS of 4 or 5. However, such scoring systems are crude and tend to underestimate the impact that such injuries may have on a patient’s life. Only 74% of patients working full-time before their injury returned to such levels of employment at the time of last follow-up. In addition we discovered a high incidence of subjective, cognitive, and mental complaints in this patient population. Although the GCS in most of these patients returned to nearly normal, this does not equate to normal function. In fact even the GOS is limited in its ability to portray the extent of deficits that affect these patients. The design of our study omitted any formal psychological testing because of the length of time and costs necessary to administer such tests. Psychological tests may have helped to quantify some of these deficits; however, all the patients felt their abnormalities were significant enough to impair Cognitive and Functional Deficits (n ⫽ 51) DEFICIT N (%) Memory problems Concentration problems Headache Fatigue Ambulatory problems Chronic pain Personality changes Dizziness Emotional problems Decreased school performance Decreased vision Hemiparesis Speech dysfunction Decreased hearing 25 18 15 16 11 9 5 7 6 3 3 3 2 1 (49) (35) (29) (31) (22) (18) (10) (14) (12) (6) (6) (6) (4) (2) daily activities. Spatt et al [15] compared patient self-reporting with family reporting and formal neuropsychological evaluations in a group of patients who were recovering from severe TBI. They noted that patients tended to minimize or underreport their cognitive and emotional deficits in contrast to family members who tended to overstate these deficits but found an overall high rate of correlation between these methods of reporting and results from formal testing [15]. Other studies have underscored the deficits in psychological and cognitive tests in patients following TBI. Rimel et al [12] looked at 3-month outcome following moderate TBI and noted the presence of headaches in 93% and memory problems in 90% with a 44% return-to-work rate. Our long-term results were somewhat better, and the difference is most likely secondary to the slowly progressive recovery following such injuries. Conclusion 3 GCS at discharge. Following moderate closed head injury most patients experienced significant neurologic improvement. Age ⱖ45 years, development of pneumonia, and delay in enteral feeding have all been shown to prolong the length of hospital stay. In addition, the subjective, cognitive, emotional, and functional deficits following this type of injury can be quite extensive and prolonged. Physicians should be aware of these long-term sequela so they can provide the appropriate support and treatment to patients following moderate TBI. Moderate CHI REFERENCES 1. Armstrong PA, McCarthy MC, Peoples JB. Reduced use of resources by early tracheostomy in ventilatordependant patients with blunt trauma. Surgery 1998; 124:763–6. 2. Aronson B, Maljanian R. Critical path education: necessary components and effective strategies. J Cont Edu Nurs 1996;27:215–9. 3. Borzotta AP, Pennings J, Papasadero B, et al. Enteral versus parenteral nutrition following severe closed head injury. J Trauma 1994;37:459 –65. 4. D’Amelia LF, Hammond JS, Spain DA, Sutyak JP. Tracheostomy and percutaneous endoscopic gastrostomy in the management of the head-injured patient. Am Surg 1994;60:180 –5. 5. Elovic E, Antoinette T. Epidemiology and primary prevention of traumatic brain injury. In: Horn LJ, Zasler ND, eds. Medical Rehabilitation of Traumatic Brain Injury. Philadelphia: Hanley and Belfus, 1996;133– 48. 6. Frankowski, RF, Annegers, T, Whitman, S. Epidemiological and descriptive studies: Part I. The descriptive epidemiology of head trauma in the United States. Central Nervous System Data Report, 1985:33-43. 7. Ibarra VA. Spine update: clinical pathways. Spine 1997;22:353–7. 8. Jennett B, Bond M. Assessment of outcome after severe brain damage: a practical scale. Lancet 1975;1: 480 –4. 9. Johnson SB, Kearney PA, Barker DE. Early criteria predictive of prolonged mechanical ventilation. J Trauma 1992;33:95–100. 10. Kitchiner D, Bundred P. Integrated care pathways. Arch Dis Child 1996;75:166 –8. 11. Moore FA, Haenel JB, Moore EE, Read RA. Percutaneous tracheostomy/gastrostomy in brain-injured patients: a minimally invasive alternative. J Trauma 1992;33:435–9. 12. Rimel RW, Giordani B, Barth JT, Jane JA. Moderate head injury: completing the spectrum of brain trauma. Neurosurgery 1982;11:344 –51. 13. Ross BJ, Barker DE, Russell WL, Burns RP. Prediction of long-term ventilatory support in trauma patients. Am Surg 1996;62:19 –25. 14. Spain DA, DeWeese RC, Reynolds MA, Richardson JD. Transpyloric passage of feeding tubes in patients with head injuries does not decrease complications. J Trauma 1995;39:1100 –8. 15. Spatt J, Zebenholzer K, Oder W. Psychological longterm outcome of severe head injury as perceived by patients, relatives, and professionals. Acta Neurol Scand 1997;95:173–9. 16. Zevola DR, Raffa M, Brown K, Hourihan EC, Maier B. Clinical pathways and coronary artery bypass surgery. Critical Care Nurse 1997;17:20 –33. COMMENTARY Few studies have attempted to prospectively identify factors within the initial hospitalization which affect outcome in moderate head injury. This effort by Vitaz et al is commendable. The authors identified factors that had a negative impact on duration of both hospital stay and mechanical ventilation, even though overall outcome was unaffected. These factors includ- Surg Neurol 291 2003;60:285–91 ed: age greater than 45 years, pneumonia, and delayed enteral feeding. Of particular use to the practicing neurosurgeon is the authors’ clinical pathway for moderate head injury. Such pathways can result in length-of-stay reductions, improved discharge dispositions, and consequent decreases in health care costs. Many studies have correlated Glasgow Coma Score (GCS) with functional recovery in severe head injury (GCS 3-8). It is logical to assume that functional recovery should be further increased in patients with moderate head injury (GCS 9-12). Vitaz et al noted no correlation for indiviual GCS scores to immediate outcome in this cohort. In fact, the majority of the cohort achieved a GCS score of 14 or 15 by discharge (94%). While 85% of the patients had favorable long-term outcomes according to the Glasgow Outcome Score, it is clear, through patient interviews, that there were lingering, significant cognitive issues that continued to plague them. These issues are often underestimated by clinicians and can greatly effect day to day living. The authors have highlighted this important point admirably. Jamie S. Ullman, M.D. Assistant Professor of Neurosurgery Mount Sinai School of Medicine New York, New York This prospective observational study of outcome following moderate traumatic brain injury suffers from having a relatively small number of patients, and an even smaller number that were contacted and chose to participate in follow-up by telephone interview. Additionally the final outcomes were entirely subjective, based upon the telephone interview and labeled under cognitive and functional deficits. These included memory problems, concentration problems, headache, fatigue, etc., none of which were evaluated in any sort of objective form. Having said this, it is a significant and useful study in that it highlights the significant subjective problems that patients surviving moderate traumatic brain injury suffer. It serves to emphasize to the practicing neurosurgeon that patients who seem to have an outwardly good initial recovery (all patients were discharged from the hospital with a Glasgow Coma Score of 14 –15), are substantially impaired and often cannot return to normal life or work. Brian T. Andrews, M.D., F.A.C.S. Department of Neurosurgery University of California, San Francisco San Francisco, California