J Neurosurg 77:901-907, 1992

Extracranial complications of severe head injury JURGEN PIEK, M.D., RANDAI.L M. CHESNUT, M.D., LAWRENCE F. MARSHALL, M.D., MARJAN VAN BERKUM-CLARK, R.N., MELVILLE R. KLAUBER, PH.D., BARBARA A. BLUNt', M.P.H., HOWARD M. EISENBERG, M.D., JOHN A. JANE, M.D., ANTHONY MARMAROU, PH.D., AND MARY A. FOULKES, PH.D.

Neurochirurgische Klinik. Universitgtt Di~sseldmL Dz~sseldor/, Germany, Division qf Neurological Surgeo, and Department of Community and Family Medicine, University of California, San Diego Medical Cenger, San Diego, Cal~fl)rnia; University o[ Tezas Medical Branch, Galveston, Texas; Universily of Virginia School ~71Medicine, Charlottesville, Virginia; and National Institute O[ Neurological Disorders and Stroke, Bethesda, Marl'land In order to define the role of intracranial and extracranial complications in determining outcome from severe head injury, 734 patients from the Traumatic Coma Data Bank were analyzed. Nine classes of intracranial and 13 classes of extracranial complications occurring within the first 14 days after admission were analyzed, while controlling for age, admission Glasgow Coma Scale motor score, early hypoxia or hypotension, and severe extracranial trauma. Outcome for survivors was based on the last recorded Glasgow Outcome Scale score, obtained a median of 521 days after injury, lntracranial complications did not significantly alter outcome for the study group. Of the exlmcranial complications, pulmonary, cardiovascular, coagulation, and electrolyte disorders occurred most frequently at 2 to 4 days. Infections developed later, peaking at 5 to 11 days. Gastrointestinal, renal, and hepatic complications followed no specific time course. Electrolyte abnormalities were the most frequent occurrence (59% of patients) but did not alter outcome. Pulmonary infections (41%), shock (29%, systemic blood pressure _< 90 mm Hg for 30 minutes or more), coagulopathy (19%), and septicemia (10%) were significant independent predictors of an unfavorable outcome. Backward-elimination, stepwise logistic regression modeling indicated that the estimated reduction of unfavorable outcome was 2.9% for the elimination of pneumonia, 3.1% for coagulation disturbances, 1.5% for septicemia, and 9.3% for shock. These data suggest that extracranial complications are highly influential in determining the outcome from severe head injury and that significant improvements in outcome in a sizeable proportion of patients could be accomplished by improving the ability to prevent or reverse pneumonia, hypotension, coagulopathy, and sepsis. K};Y W o ~ o s head injury extracranial complications

A

9

Traumatic Coma Data Bank

MAJOR objective of the Traumatic Coma Data Bank is to evaluate the determinants of recovery in patients with severe head injury. As the fate of many patients suffering from severe head injury is determined at the time of impact by the type of brain lesion, previous studies have generally focused on the type and severity of the anatomical brain injury as predictors of outcome when coupled with a neurological examination. Of course, many clinical,2'4'5'7 9,~J,J4, 20,2~,2~-3o,33,3~,42,49.,~2-54~7neurophysiological,23,44.4~.4s,5oneuv9 33 3f 40 41 47 60 rochemical] '~e and neuroradiological lg ..... ". . . . . . . . factors have been identified as predictors of outcome as well. It is widely known that severe head injury may be followed by numerous extracranial complications. 3'6

J. Neurosurg. / Vo&me 77/December, 1992

9

outcome

9

However, little attention has been paid to the influence of these complications on outcome 47 and whether specific subgroups of patients are at particular risk. In an attempt to answer these questions, we have studied a subgroup of 734 patients from the Traumatic Coma Data Bank. Clinical Material and Methods

Traumalic Coma Data Bank The Traumatic Coma Data Bank is a National Institute of Neurological Disorders and Stroke (NINDS) collaborative project involving four clinical centers: the Medical College of Virginia at Richmond, the University of California at San Diego, the University of Vir901

J. Piek, et al. ginia at Charlottesville, and the University of Iexas Medical Branch at Galveston, with a coordinating center within the Biometry and Field Study Branch of the NINDS. These clinical centers prospectively studied all severely head-injured patients admitted between April, 1983, and April, 1988.

Patient Popltlation The operational definition of severe head injury was a Glasgow Coma Scale s.' (GCS) score of 8 or less on admission (postresuscitation) or during the ensuing 48 hours. In addition to data for the acute-care period, prehospital information and rehabilitation follow-up results were collected. Patient outcome was determined by the last recorded Glasgow Outcome Scale 27 (GOS) score. The median time from injury to obtaining the last recorded GOS score for survivors was 521 days (range 5 days to 3.4 years). There were 1030 patients admitted to the Traumatic Coma Data Bank: of these, 284 were brain-dead on admission, did not survive resuscitation, or suffered a gunshot wound to the brain. An additional 23 patients had no completed complication forms. As some patients fulfilled more than one of these criteria, a total of 734 remained for analysis. Of these, 613 cases were available for logistical regression analysis because preresuscitation GCS scores were missing for 82 and data on hypoxia or hypotension at admission to the Traumatic Coma Data Bank hospital were missing for 39 others. Dala Colleclion The evolution and development of extracranial and intracranial complications were recorded daily for 14 days postinjury using the definitions listed in the Appendix. Although a patient may have suffered a particular complication more than once during the hospital course, it was only counted as one event. Previous investigators have used the postresuscitation GCS score to gauge the severity of head injury because it accounts better for differences in prehospital resuscitation. 27'3s' .~6.3~ Unfortunately, due to the widespread early use of muscle relaxants and sedation during triage in Traumatic Coma Data Bank patients, postresuscitation GCS scores are very often unavailable. We have, therefore, chosen to use preresuscitation values. Data Analysis As the GCS motor score is generally available for the preresuscitation period and since Marshall, et al.)9 have shown the motor score to be the third most powerful predictor of outcome in Traumatic Coma Data Bank patients, we used the maximum preresuscitation GCS motor score as our index of severity of head injury. This strategy lowered the number of missing observations from 193 to 82, thereby notably increasing our sample size. For patients whose GCS score and GCS motor score were available, the Pearson correlation coefficient was r = 0.89. Outcome was dichotomized 902

tbr statistical analysis and assessed as either "favorable" (last recorded GOS score was good or moderate disability) or "unfavorable" (last recorded GOS score was severe disability, vegetative state, or dead). In addition, patients were separated into two groups by age: less than 40 years old and 40 years or older. A backward-elimination, stepwise logistic regression model was used to identify the dependent variable (the log odds of unfavorable outcome). 25 We used the Hosmer-Lemeshow approach to test the goodness-of-fit of the logistic model. Within the model the following factors were included: 1) age on admission, divided into < 40 years or >_ 40 years; 2) maximum preresuscitation GCS motor score; 3) hypoxia upon arrival at the Traumatic Coma Data Bank hospital emergency room; 4) hypotension upon arrival at the Traumatic Coma Data Bank hospital emergency room; 5) multiple trauma to any extracranial system with an Abbreviated Injury Scale score of 3 or greater; 6) presence or absence of any intracranial complication as defined in the Appendix; and 7) presence or absence of the different extracranial complications as defined in the Appendix ("presence" was defined as a severity score for the complication of serious, severe, or life-threatening; "absence" was a score of none, mild, or moderate). The first four factors were included because previous Traumatic Coma Data Bank research has shown these factors to be related to outcome. 39 The 5% level was used for statistical significance. Results Of the 734 patients, 566 (77.1%) were male and 579 (78.9%) were less than 40 years old (range 1 to 93 years, median 24 years). There were 1709 individual extracraniat complications, ranging in severity from mild to life-threatening. A good outcome occurred in 199 patients (27.1%), moderate outcome in 121 (16.5%), severe outcome in 114 (15.5 %), vegetative outcome in 37 (5.0%), and death in 263 (35.8%). Therefore, 320 (43.6%) had a favorable outcome and 414 (56.4%) had an unfavorable outcome. Approximately 22% of all patients developed at least one intracranial complication, but extracranial complications were more frequent and appeared to be much more important. As can be seen from Fig. 1, electrolyte abnormality was the most common complication and occurred in more than onehalf (59.3%) of all patients; pneumonia was the next most common factor (40.6%). Shock of at least 30 minutes' duration occurred in 29.3%, disturbances of blood coagulation in 18.4%, and septicemia in 10%. Complications affecting the digestive tract, liver, or kidneys occurred rarely. Although disturbance of electrolytes occurred in nearly two-thirds of the patients, logistic regression analysis did not identify electrolyte imbalance (p = 0.35) as an independent predictor of unfavorable outcome. However, age (p < 0.001), maximum preresuscitation GCS motor score (p < 0.001), shock upon arrival at the Traumatic Coma Data Bank emergency

J. Neurosurg. / Volume 77/December, 1992

Extracranial complications of severe head injury

FIG. 2. Graph showing incidence of early complications. FIG, 1. Chart showing frequency of extracranial complications in severely head-injured palients. any one complication may have a favorable influence on others. room (p = 0.0077), shock of at least 30 minutes' duration during the hospital course (p < 0.001), pneumonia (p = 0.037), coagulopathy (p = 0.0036), and septicemia (p = 0.047) were significant independent predictors of an unfavorable outcome. All other complications, including presence of an intracranial complication, could not be identified as being significantly responsible for overall unfavorable outcome in patients with severe head injuries. Although it was not our aim to develop a statistical model for outcome prediction in these patients, the model did show acceptable sensitivity (75.4 % ) and specificity (75.8%). Approximately 20% of all patients had an estimated 40% to 60% probability of an unfavorable outcome. This indicates a large subgroup of patients whose prognosis was uncertain at the time of injury and whose outcome might have been significantly influenced by eliminating or minimizing complications. The time course of the different complications could be divided into three periods. Figure 2 shows that pulmonary insufficiency, cardiovascular complications (including shock), coagulation disorders, and electrolyte abnormalities were early complications, with peak occurrence during the first 2 to 4 days after trauma. Figure 3 shows that late complications were exclusively infections (pneumonia and septicemia), which reached peak occurrence from Days 5 through 11. A third group of complications (gastrointestinal, renal, hepatic) had no special time course and were observed with a fairly constant frequency from Days 1 to 14. We systematically removed the complications using the backward-elimination, stepwise logistic regression model described above in order to determine the estimated reduction in unfavorable outcome if the complications had not occurred. The estimated reduction in unfavorable outcome was 2.9% for the elimination of pneumonia, 3.1% for coagulation disturbances, 1.5% for septicemia, and 9.3% for either form of shock. Note that these estimates assume the elimination of each complication, holding all others constant. These estimates may be conservative, since the elimination of

J. Neurosurg. / Volume 77/December, 1992

Discussion Extracranial complications occur frequently in severely head-injured patients. 3'6"37"47Although the outcome of an individual patient may be adversely affected by a number of different complications, we were able to identify only a few that had an independent influence on the patients in the Traumatic Coma Data Bank. Hypotension, pneumonia, coagulopathy, and septicemia were significant factors in producing less favorable outcomes in this study group. Hypotension and coagulopathy occurred early during the postinjury course (Fig. 2), while pneumonia and septicemia were later complications (Fig. 3). The questions that arise from these data include: 1) are these complications preventable and, if so, how? and 2) how much improvement in outcome could be expected if they were prevented?

Hypotension In other analyses, we have shown that hypotension, if present, is a significant and powerful determinant of outcome after severe head injuryJ 3 The occurrence of a systolic blood pressure of 90 mm Hg or less during the time period between injury and arrival at the receiving hospital doubles the mortality rate independent of other factors. Such hypotension occurred in 34.6%

FIG. 3. Graph showing incidence of late complications. 903

J. Pick, et al. of Traumatic Coma Data Bank patients. Some percentage of hypotension occurring during the triage period may be preventable: however, the majority will only be treatable. Nevertheless, given their profound impact on outcome and their surprising frequency, the methods of resuscitating the head-injured patient need to be seriously re-examined, both in terms of efficacy and completeness. As shown above, hypotension continues to occur and exert its adverse influence on outcome even after the patient has been admitted to the intensive care unit (ICU). By our model, the effective elimination of shock would have resulted in an estimated 9.3% reduction of unfavorable outcome. The desirability of averting those instances that are preventable adds to the growing mass of evidence that head-injured patients should never be dehydrated but rather be kept euvolemic. Central venous and pulmonary artery pressures should be maintained within normal limits. For those hypotensive episodes that are not preventable, rapid and vigorous treatment appears particularly important for the patient with an injured brain. Volume replacement and, if necessary, pressors and/or inotropes should be readily employed to empirically correct the perfusion abnormality if the etiology of the pressure drop is not clearly apparent. In contrast to the situation in a patient with an extracranial injury, hypotension even of relatively low magnitude and short duration does not appear to be tolerated by the head-injury victim. Coa guh~pat h y The situation with coagulopathy is slightly different. Previous investigators have reported a positive correlation between the presence and severity of disseminated intravascular coagulation (DIC), assessed by plasma fibrinogen degradation product levels, and the degree of brain injury. 4~'4ss'~When DIC occurs due to extracranial trauma, mortality is most directly correlated with the severity of the underlying disease and not with the DIC itself. '~ Although the GCS score and laboratory markers of DIC did factor out as separate predictors of outcome in one study using stepwise logistic regression, 4: this result might well be due to the nonspecific nature of the GCS score in reflecting anatomical brain pathology. This suggests that some of the correlation of coagulopathy with poorer outcome may be a reflection of the severity of the underlying injury. The central nervous system is the most common site of diffuse intravascular thrombosis in patients with coagulopathy, frequently accompanied by necrosis. 32 Petechial hemorrhages, usually the result of thrombosis followed by clot lysis and hemorrhage, are commonly found in the brain and spinal cord of patients with diffuse coagulopathic processesY- Therefore, although most of the acute coagulol~athies associated with brain injury are not preventable, prompt treatment should be effective in reducing the morbidity associated with coagulopathy even in the strongly confounding situation noted above. 904

Pneutmmia Pneumonia is a complication thai was observed in the intermediate and late stages, primarily from Days 5 to 10 postinjury. Following electrolyte imbalance, pneumonia was the second most frequent complication in the study group and, in contrast to electrolyte abnormalities, had a significant influence on the overall outcome of the group. Improved strategies for the prevention or effective treatment of pulmonary infections would potentially have a positive effect on the outcome of severe head injury. The risk of pneumonia in patients in an ICU setting increases with impairment of airway reflexes (particularly with a known history of aspiration) and with more than 24 hours of ventilatory assistance, both situations common in head-injured patients. ~5.34.s6The presence of an intracranial pressure (1CP) monitor is itself a statistical predictor of increased risk of pneumonia,'S as is the use of barbiturates for the treatment of ICP. ~(j For these reasons, there will likely always be some pneumonias that are not preventable and can only be palliated by rapid diagnosis, prompt identification of the responsible organism(s), and vigorous treatment. Assiduous vigilance for signs of pulmonary infection, including the employment of highly selective methods of obtaining surveillance and diagnostic cultures (for example, a protected brush catheter, bronchial washings or brushings, and transthoracic needle aspiration) 31,55 should be combined with a low threshold for initiating antibiotic therapy for suspected pneunomia.~2 There are, however, a number of concepts regarding the prevention of pneumonia that may be very applicable to head-injured patients. The neutralization of gastric pH, whether by antacids, 17"~ H2 receptor blockers, '5'7 or enteral nutrition, 2~ has been associated with an increased risk of pneumonia. The use of sucralfate, which is equally protective against gastric bleeding but does not raise gastric pH to the same extent, appears less likely to be associated with pneumonia. ~v,Ss The potentiation of successive colonization of the stomach, oropharynx, and trachea with Gram-negative organisms by the presence of a neutral pH appears to be the mechanism.'7'26 It has been suggested that intravenous antibiotic prophylaxis combined with venous intestinal and oropharyngeal topical antibiotic decontamination can be effective in decreasing nosocomial pneumonias in general. S~Such treatment paradigms for head-injured patients need to be rigorously investigated. If pneumonia could be eliminated as a single complication, our statistical model suggests that an estimated 2,9% of all patients would shift from an unfavorable to a favorable outcome. Furthermore, because pulmonary infections are frequently associated with other complications (such as acute respiratory distress syndrome, hypoxia, septic shock, and subsequent multisystem organ failure), the actual number of patients who might benefit from prevention and treatment of pulmonary infections is potentially much higher. Further research must focus on infectious complications J. Neurosurg. / Volume 77/December, 1992

Extracranial complicalions of severe head injury and especially on mechanisms for preventive treatment. Pneumonia clearly is an ideal target for attack because of its high frequency and relatively late onset, &7),g.~ Sepsis is a significant complication in any patient and is a particular problem in the ICU due to the illness of the patient population and the invasive nature of many treatments and monitoring devices. In this study, elimination of sepsis would shift an estimated 1.5% of all patients from an unfavorable to a favorable outcome category. Due to its multifactorial nature and many venues of origin, sepsis is particularly difficult to prevent. Fundamental techniques for prevention include screening studies such as su:rveillance pulmonary' cultures, careful attention to invasive monitors and vascular access lines for early evidence of local infection, and possibly, systemic nutritional support in order to increase the patient's immunological competency. When prevention is unsuccessful, management of sepsis requires rapid diagnosis and treatment. It is important to keep m mind that some of the more occult sources for bacteremia, such as sinusitis, can easily be overlooked in the head-injured patient. > Early diagnosis may be particularly difficult in the patient treated with high-dose barbiturates for intracranial hypertension. The mild hypothermia and systemic anergy often seen in these patients makes it more critical to recognize the secondary indicators of sepsis, such as progressive thrombocytopenia, hyperglycemia, and unexplained cardiovascular hyperdynamism with low sequential vascular response and a falling arteriovenous oxygen content difference. ~) Unfortunately, the first recognized sign of sepsis in such patients is often septic shock. This adds the morbidity of hypotension to the patient's complication profile, further increasing the delrimental effects of sepsis.

Conclusions This study demonstrates that the prevention or effective early treatment of major extracranial complications would have a salutary role in improving the outcome from head injury. This area of study represents fertile ground for innovative strategies to improve the overall outcome of severe head injury. APPENDIX

Definitions of Intracranial and Extraeranial Complications* lntracranial Complications Intracrania/Hemorrhage: Reaceumulation of intracerebral hemorrhage in the operative site. Any new hemorrhage remote * All definitions are available via the TCDB Manual of Operations and data forms from the National Technical Information Service (NTIS), United States Department of Commerce, 5285 Port Royal Road, Springfield, Virginia 22161 (NTIS Accession No. PB87 228060/AS).

J. Neurosurg. / Volume 77/December, 1992

tiom the surgical field is not considered a complication.

hllravc~tricular lfemorrha~e. Bleeding in the ventricular system [bllowing an intracranial procedure, either definitive surger~ or diagnostic procedure. Diagnosis is made by computerizcd tornography scan. Subdtoa! ltemamma. Development of recurrent subdural hematoma after primary operation for delinitive treatment. Epidura/ lh,matoma: Development of recurrem epidural hematoma following operative procedure. Cerebro,v~ma/ kTuid Leak. Leak of" CSF from a surgical incision. Otorrhea or rhinorrhea is not a surgical complication unless the procedure is performed on the base of the skull. Intracrania/l're.~sure Trace Loss: Loss of pulse wave on the oscilloscope if it is not corrected before the next scheduled reading. This information is obtained from ICU nurses" notes or continuous ICP trace recordings. t'osL~urgical Ventriculitis/Mening, ilis." Positive CSF culture or, in the absence of a positive culture, one of the following: 1) > 50% polymorphonuclear cells on CSF cell count (minimum 50 cells counted), or 2) CSF glucose < 15 rag/100 ml. Abscess: Abscess documented by needle aspiration, surgical exploration, or autopsy. The location of the abscess must be anatomically related to the previous cranial procedure. 14"ound ln/eclion: This includes such entities as ostcomyeIitis of a bone flap, subgaleal infection, and epidural or subdural empyema secondary to cranial procedure.

Extracranial Complications Ptdmonarr Complications: Includes such diagnosis as acute respirato@ distress syndrome, atelectasis, Neural effusion, pulmonary embolus, and respiratory failure. The essentials of diagnosis are positive radiographic findings and abnormal arterial blood gas levels, including PaOe < 60 mm Hg, PaCO: > 45 mm Hg. It includes lung abscess and empyema. Pneumonia is a separate category and not coded here unless one of the above conditions also applies. Cardiovascular Complications." Includes cardiac arrhylhmias, congestive heart failure, myocardial ischemia, and systolic hypotension < 90 mm Hg or systolic hypertension > 160 mm Hg if it persists for at least 30 minutes and requires treatment. This applied to normotensive patients. In known hypertensive patients, variations of blood pressure > 40 mm Hg requiring adjustment of treatment is considered a complication. Periptwra/ Vascular O~mplications: Deep venous thrombosis of an extremity and/or pelvic organ requiring therapeutic intervention (bed rest, heat, anti-inflammatory' agents, anticoagulant therapy, etc.). Gastrointeslinal Complications: Gastrointestinal hemorrhage requiring transfusion, gastric perforation, and secondary pancreatitis are included within this category. Renal Complications: Includes acute renal failure which is determined by creatinine levels > 2.0 rag/100 ml. Hepatic Complications. Includes liver failure, hepatitis, cholangitis, and hepatic renal syndrome. The diagnosis can be established by determination of serum glutamic-pyruvic transaminase levels > 30 IU and total serum bilirubin levels > 2.5 rag/100 ml. EleclroO,te Abnormalities. Include any electrolyte imbalance that results in clinical signs or requires specific treatment. Omgulopathy: Determined by a platelet count < 50,000/cu ram, prothrombin time > 16 seconds, partial thromboplastin time > 50 seconds.

Syndrome of Inappropriale Secretion of Antidiuretic Hormone: Essentials of diagnosis include hyponatremia not secondary to overhydration (Na < 130 mEq/liter); urine osmolality to exceed serum osmolality. 905

J. Piek, et al. Diabetes lnsipidus: Urine output over 200 ml/hr for 24 hours not responding to fluid restriction or if patient requires a treatment with pitressin. Urine specific gravity < 1.005, urine osmolality < half that of plasma. Nonsurgical VentricuHtis/Meningitis: Positive CSF culture or, in the absence of a positive culture, one of the following: > 50% polymorphonuclear cells on CSF cell count (minimum 50 cells counted), or CSF glucose < 15 mg/100 ml. Pneumonia: Diagnosis is made by the presence of infiltration on chest x-ray film and sputum positive for organism on the Gram stain specimen or culture. Lung abscesses and empyemas are included under "Pulmonary Complications." Septicemia: Documented by positive blood culture associated with clinical evidence of sepsis such as hyperthermia, hypotension, etc.

References 1. The Abbreviated Injury Scale, 1985 Revision. Arlington Heights, 11l:American Association of Automotive Medicine, 1985 2. Alberico AM, Ward JD, Choi SC, et al: Outcome after severe head injury. Relationship to mass lesions, diffuse injury, and ICP course in pediatric and adult patients. J Neurosurg 67:648-656, 1987 3. Baigelman W, O'Brien JC: Pulmonary effects of head trauma. Neurosurgery 9:729-740, 1981 4. Becket DP, Miller JD, Ward JD, et al: The outcome from severe head injury with early diagnosis and intensive management. J Neurosurg 47:491-502, 1977 5. Berger MS, Pitts LH, Lovely M, et al: Outcome from severe head injury in children and adolescents. J Nenrosurg 62:194-199, 1985 6. Bloomfield EL: Extracerebral complications of head injury. Crit Care Clin 5:881-892, 1989 7. Born JD, Albert A, Hans P, et al: Relative prognostic value of best motor response and brain stem reflexes in patients with severe head injury. Neurosurgery 16: 595-601, 1985 8. Bowers SA, Marshall LF: Outcome in 200 consecutive cases of severe head injury treated in San Diego County: a prospective analysis. Neurosnrgery 6:237-242, 1980 9. Braakman R, Gelpke G J, Habbema JDF, et at: Systemic selection of prognostic features in patients with severe head injury. Neurosurgery 6:362-370, 1980 10. Braun SR, Levin AB, Clark KL: Role of corticosteroids in the development of pneumonia in mechanically ventilated head-trauma victims. Crit Care Med 14:198-2131, 1986 11. Carlsson CA, yon Essen C, Lrfgren J: Factors affecting the clinical course of patients with severe head injuries. Part 1: Influence of biological factors. Part 2: Significance of posttraumatic coma, J Neurosurg 29:242-251, 1968 12. Chesnut RM: Medical management of the head injured patient, in Cooper PR (ed): Head Injury, ed 3. Baltimore: Williams & Wilkins, 1992, pp 459-501 13. Chesnut RM, Marshall LF, Bowers-Marshall S: Medical management of elevated intracranial pressure, in Cooper PR (ed): Head Injury, ed 3. Baltimore: Williams & Wilkins, 1992, pp 225-246 14. Choi SC, Ward JD, Becker DP: Chart for outcome prediction in severe head injury. J Neurosurg 59:294-297, 1983 15. Craven DE, Kunches LM, Killinsky V, et at: Risk factors for pneumonia and fatality in patients receiving continuous mechanical ventilation. Am Rev Respir Dis 133: 792-796, 1986 16. DeSalles AAF, Kontos HA, Becker DP, et al: Prognostic 906

17.

18.

19. 20. 21.

22.

23.

24. 25. 26.

27. 28. 29. 30. 31. 32.

33. 34. 35.

36.

significance of ventricular CSF lactic acidosis in severe head injury, d Neurosurg 65:615-624, 1986 Driks MR, Craven DE, Celli BR, et al: Nosocomial pneumonia in intubated patients given sucralfate as compared with antacids or histamine type 2 blockers. The role of gastric colonization. N Engl J Med 317:1376-1382, 1987 Eisenberg HM, Gary HE Jr, Aldrich EF, et al: Initial CT findings in 753 patients with severe head injury. A report from the NIH Traumatic Coma Data Bank. J Neurosurg 73:688-698, 1990 Feinstein DI: Diagnosis and management of disseminated intravascular coagulation: the role of heparin therapy. Blood 60:284-287, 1982 Frowein RA, auf der Haar K, Terhaag D: Assessment of coma reliability of prognosis. Neurosurg Rev 3:67-74, 1980 Gaab MR, Haubitz I: lntracranial pressure, primary/ secondary brain stem injury and prognosis in cerebral trauma, in Ishii S, Nagai H, Brock M (eds): Intracranial Pressure V. Berlin: Springer-Verlag, 1983, pp 501-506 Gennarelli TA, Spielman GM, Langfitt TW: Influence of the type of intracranial lesion on outcome from severe head injury. A multicenter study using a new classification study. J Neurosurg 56:26-32, 1982 Greenberg RP, Newlon PG, Hyatt MS, et al: Prognostic implications of early multimodality evoked potentials in severely head-injured patients. A prospective study. J Neurosurg 55:227-236, 1981 Grindlinger GA, Niehoff J, Hughes SL, et at: Acute paranasal sinusitis related to nasotracheal intubation of head-injured patients. Crit Care Med 15:214-217, 1987 Hosmer DW, Lemeshow S: Applied Logistic Regression. New York: John Wiley & Sons, 1989 Jacobs S, Chang RW, Lee B, et al: Continuous enteral feeding: a major cause of pneumonia among ventilated intensive care unit patients. J Parenter Enteral Nutr 14: 353-356, 1990 Jennett B, Bond M: Assessment of outcome after severe brain damage. A practical scale. Lancet 1:480-484, 1975 Jennett B, Teasdale G, Braakman R, et at: Predicting outcome in individual patients after severe head injury. Lancet 1:1031-1034, 1976 Jennett B, Teasdale G, Braakman R, et al: Prognosis of patients with severe head injury. Neurosurgery 4: 283-289, 1979 Jennett B, Teasdale G, Galbraith S, et al: Severe head injuries in three countries. J Neurol Neurosurg Psychiatry 40:291-298, 1977 Jimrnez P, Torres A, Rodriguez-Roisin R, et at: Incidence and etiology of pneumonia acquired during mechanical ventilation. Crit Care Med 17:882-885, 1989 Kaufman HH, Hui KS, Mattson JC, et al: Clinicopathological correlations of disseminated intravascular coagulation in patients with head injury. Neurosurgery 15: 34-42, 1984 Klauber MR, Marshall LF, Luerssen TG, et at: Determinants of head injury mortality: importance of the low risk patient. Neurosurgery 24:31-36, 1989 Langer M, Cigada M, Mandelli M, et al: Early onset pneumonia: a multicenter study in intensive care units. Intens Care Med 13:342-346, 1987 Lobato RD, Cordobes F, Rivas JJ, et al: Outcome from severe head injury related to the type of intracranial lesion. A computerized tomography study. J Neurosurg 59:762-774, 1983 Luerssen TG, Klauber MR, Marshall LF: Outcome from head injury related to patient's age. A longitudinal pro-

J. Neurosurg. / Volume 77/December, 1992

Extracranial complications of severe head injury

37. 38. 39. 40. 41. 42. 43.

44. 45. 46. 47. 48. 49. 50. 51.

52. 53.

spective study of adult and pediatric head injury. J Neurosurg 68:409-416, 1988 Marmarou A. Anderson RL, Ward JD, etal: lmpact of ICP instability and hypotension on outcome in patients with severe head trauma. J Neurosurg 75:$59-$66, 1991 Marshall LF, Becker DP, Bowers SA, etal: The National Traumatic Coma Data Bank. Part I: Design, purpose, goals and results. J Neurosurg 59:276-284, 1983 Marshall LF, Gautille T, Klauber MR, etal: The outcome of severe closed head injury'. J Neurosurg 75:$28-$36~ 1991 Marshall LF. Marshall SB. Klauber MR, et al: A new classification of head injury' based on computerized tomography. J Neurosurg 75:S 14-$20, 1991 Miller JD: ICP monitoring - - current status and future directions. Acta Neurochir 85:80-86, 1987 Miller JD, Becker DP, Ward JD, etal: Significance of intracranial hypertension in severe head injury'. J Neurosurg 47:503-516. 1977 Miner ME, Kaufman HH, Graham SH, el al: Disseminated intravaseular coagulation fibrinolytic syndrome following head injury in children: frequency and prognostic implications. J Pediatr 100:687-691, 1982 Newlon PG, Greenberg RP: Evoked potentials in severe head injury. J Trauma 24:61-66, 1984 Olson JD, Kaufman HH, Moake J, et al: The incidence and significance of hemostatic abnormalities in patients with head injuries. Neurosurgery 24:825-832, 1989 Papanicolaou AC, Loring DW, Eisenberg HM, etal: Auditory brain stem evoked responses in comatose headinjured patients. Neurosurgery 18:173-175, 1986 Price dE, Murray A: The influence of hypoxia and hypotension on recovery from head injury'. Br J Aeeid Surg 3: 218-224, 1972 Riffel B, St6hr M, Trost E, et al: Friihzeitige prognostische Aussage mittels evozierter Potentiale beim schweren Schfidel-Hirn-Trauma. EEG EMG 18:192-199, 1987 Saul TG, Ducker TB: Effect ofintracranial pressure monitoring and aggressive treatment on mortality in severe head injury. J Neurosurg 56:498-503, 1982 Seales DM, Rossiter VS, Weinstein ME: Brainstem auditory evoked responses in patients comatose as a result of blunt head trauma. J Trauma 19:347-353, 1979 Stoutenbeek CP, van Saene HKF, Miranda DR, et al: The effect of oropharyngeal decontamination using topical nonabsorbable antibiotics on the incidence of nosocomial respirator,:' tract infections in multiple trauma patients. J Trauma 27:357-364, 1987 Teasdale G, Jennett B: Assessment and prognosis of coma after head injury. Acta Neuroehir 34:45-55, 1976 TeasdaleG, Jennett B: Assessment of coma and impaired

J. Neurosurg. / Volume 77/December, 1992

consciousness. A practical scale. Lancet 2:81-84, 1974 54. Teasdale G, Skene A, Parker E, etal: Age and outcome of severe head injury. Acta Neuroehir Suppl 28: 140-143, 1979 55. Tobin MJ, Grenvik A: Nosocomial lung infection and its diagnosis. Crit Care Med 12:191-199, 1984 56. Torres A, Aznar R, Gatell JM, et al: Incidence, risk, and prognosis factors of nosocomial pneumonia in mechanically ventilated patients. Am Rev Respir Dis 142: 523-528, 1990 57. Toutant SM, Klauber MR, Marshall LF, et al: Absent or compressed basal cisterns on first CT scan: ominous predictors of outcome in severe head injury, d Neurosurg 61:691-694, 1984 58. Tryba M: Risk of acute stress bleeding and nosocomial pneumonia in ventilated intensive care unit patients: sucralfate versus antacids. Am J Med 83 (3B):117-124, 1987 59. Ueda S, Fujitsu K, Fujino H, et al: Correlation between plasma fibrin-fibrinogen degradation product values and CT findings in head injury, d Neurol Neurosurg Psychiatry 48:58-60, 1985 60. van Dongen KJ, Braakman R, Gelpke G: The prognostic value of computerized tomography in comatose headinjured patients. J Neurosurg 59:951-957, 1983

Manuscript received November 19, 1991. Accepted in final form April 10, 1992. This work was supported by the Traumatic Coma Data Bank (TCDB) under Contracts NO I-NS-3-2339, NO 1-NS-32340, NO1-NS-3-2341, NO1-NS-3-2342, and NOI-NS-62305 from the National Institute of Neurological Disorders and Stroke (NINDS) and by the "Wilhelm-T6nnis-Foundations" of the German Society for Neurosurgery. The TCDB Manual of Operations, which includes the TCDB data forms, is available from the National Technical Information Service (NTIS), United States Department of Commerce, 5285 Port Royal Road, Springfield, Virginia 22161 (NTIS Accession No. PB87 228060/AS). Address for Dr. Piek: Neurochirurgische Klinik, Universitfit Diisseldorf, Dusseldorf, Germany. Address for Dr. Eisenberg: University of Texas Medical Branch, Galveston, Texas. Address for Dr. Jane: University of Virginia School of Medicine, Charlottesville, Virginia, Address for Dr. Foulkes: National Institute of Neurological Disorders and Stroke, Bethesda, Maryland. Address reprint requests to: Randall M. Chesnut, M.D., Division of Neurological Surgery, University of California San Diego Medical Center, 200 West Arbor Drive, San Diego, California 92103-8893.

907