Click here to listen to the podcast featuring an interview with the authors.

J Neurosurg 116:234245, 2012

Postoperative outcomes following closed head injury and craniotomy for evacuation of hematoma in patients older than 80 years
Clinical article
DARRYL LAU, B.A.,1 AbdULRAHMAN M. EL-SAYed, B.S., 24 JOHN E. ZIeWACz, M.D., M.P.H., 57 PRIYA JAYACHANdRAN, B.A.,1 FARHAN S. HUQ, B.S.E.,1 GReTTeL J. ZAMORA-BeRRIdI, M.H.S.,1 MATTHeW C. DAVIs, B.A.,1 ANd STepHeN E. SULLIVAN, M.D. 5
1 University of Michigan Medical School and 5Department of Neurosurgery, University of Michigan, Ann Arbor, Michigan; 2Department of Public Health, Oxford University, Oxford, United Kingdom; 3Department of Epidemiology and 4College of Physicians and Surgeons, Columbia University, New York, New York; 6Center for Surgery and Public Health, Harvard School of Public Health; and 7Department of Surgery, Brigham and Womens Hospital, Boston, Massachusetts

Object. Advances in the management of trauma-induced intracranial hematomas and hemorrhage (epidural, subdural, and intraparenchymal hemorrhage) have improved survival in these conditions over the last several decades. However, there is a paucity of research investigating the relation between patient age and outcomes of surgical treatment for these conditions. In this study, the authors examined the relation between patient age over 80 years and postoperative outcomes following closed head injury and craniotomy for intracranial hemorrhage. Methods. A consecutive population of patients undergoing emergent craniotomy for evacuation of intracranial hematoma following closed head trauma between 2006 and 2009 was identified. Using multivariable logistic regression models, the authors assessed the relation between age (> 80 vs 80 years) and postoperative complications, intensive care unit stay, hospital stay, morbidity, and mortality. Results. Of 103 patients, 27 were older than 80 years and 76 patients were 80 years of age or younger. Older age was associated with longer length of hospital stay (p = 0.014), a higher rate of complications (OR 5.74, 95% CI 1.2925.34), and a higher likelihood of requiring rehabilitation (OR 3.28, 95% CI 1.139.74). However, there were no statistically significant differences between the age groups in 30-day mortality or ability to recover to functional baseline status. Conclusions. The findings suggest that in comparison with younger patients, patients over 80 years of age may be similarly able to return to preinjury functional baselines but may require increased postoperative medical attention in the forms of rehabilitation and longer hospital stays. Prospective studies concerned with the relation between older age, perioperative parameters, and postoperative outcomes following craniotomy for intracranial hemorrhage are needed. Nonetheless, the findings of this study may allow for more informed decisions with respect to the care of elderly patients with intracranial hemorrhage. (DOI: 10.3171/2011.7.JNS11396)

KeY WORds age craniotomy hematoma intracranial hemorrhage morbidity trauma traumatic brain injury

hemorrhage and/or hematoma due to trauma are serious conditions that often require emergent surgical intervention; if immediate treatment is not provided, morbidity and mortality are high.31 Intracranial hemorrhage can lead to acute neurological
NTRACRANiAL

Abbreviations used in this paper: BMI = body mass index; GCS = Glasgow Coma Scale; ICU = intensive care unit.

deterioration, cardiovascular instability,25 localized mechanical damage to brain tissue,9,15,24 perilesional edema,39 increased intracranial pressure,31 brainstem compression, herniation,2 and death.8 When conservative interventions such as hemostatic therapy (reversal of anticoagulant therapy and coagulative induction) and pharmaceutical management of intracranial pressure27 fail to treat intracranial hemorrhage efficiently or if the hematoma is too
J Neurosurg / Volume 116 / January 2012

234

Age and craniotomy following intracranial hemorrhage

large for nonsurgical treatments,16 a craniotomy for evacuation of the hematoma or hemorrhage may be required. However, even with a craniotomy, reported postoperative morbidity and mortality rates for patients with intracranial hemorrhage remain high; recovery from such aggressive treatment can be prolonged, and failure to return to baseline is common. Studies from the 1980s report mortality rates ranging from 65% to 90% after craniotomy for evacuation of acute subdural hematomas.13 By the 1990s, mortality rates for the same procedure had dropped to between 25% and 58%.8,20,35 Previous neurosurgical studies have also demonstrated age-dependent morbidity and mortality rates, with younger patients faring better than older patients undergoing craniotomy for intracranial bleeding.13,35 However, current advances in imaging, rapid diagnostics, medical management of hemodynamic properties, and surgical techniques have improved patient outcomes,36 and a further decrease in mortality associated with treatment for intracranial bleeding has been demonstrated; in 2008, postoperative mortality associated with craniotomy for evacuation of acute subdural and epidural hematoma ranged from 22% to 41%.17,33,34 It is plausible that this overall decrease in mortality may also be reflected by improved survival among older patients. Given the breadth and scope of medical and surgical advancements, this raises an interesting and valuable question: Do older patients continue to suffer higher morbidity and mortality rates than younger patients following evacuation of intracranial hematoma and/or hemorrhage? Because older age is a risk factor for intracranial hemorrhage after head trauma, this question is of increasing importance, as the geriatric population grows in size and the costs and efficacy of end-of-life care continue to be a contentious topic of debate.7,12,26 An improved understanding of the potential risks of craniotomy for treatment of intracranial hemorrhage in patients over the age of 80 years may allow physicians to make better-informed decisions about treatment options for this population. As there is a paucity of research that has assessed outcomes and morbidity following neurosurgical interventions among these patients relative to younger patients, we assessed the relation between older age and 30-day postoperative outcomes following craniotomy for evacuation of intracranial hemorrhage and hematoma due to closed head trauma. The age criterion used in this study was patient age over 80 years at time of surgery; this criterion was chosen because there has been a significant rise in the number of patients over the age of 80 in the US in the past several decades.38 Also, end-of-life decisions may be more relevant in this population than in patients between 65 and 80 years of age, who would be included in our exposure group if traditional geriatric cutoffs were used. adult patients undergoing craniotomy for evacuation of hematoma or hemorrhage between the beginning of 2006 and the end of 2009. We identified all patients suffering closed head trauma resulting in intracranial hemorrhage and hematoma requiring evacuation. The indications for craniotomy for evacuation of hematoma were hemorrhage with mass effect and/or neurological decline. Of the 111 consecutive patients identified, 8 patients with additional comorbidities and diagnosis of HIV/AIDS, multiple cranial hematomas, brain tumors/lesions, or skull fractures requiring cranioplasty were excluded from this analysis since these comorbidities required additional medical and/or surgical intervention that might confound the analysis. In all cases, intraparenchymal hematoma, subdural hematoma, or epidural hematoma was diagnosed by imaging. Our covariate set for this study included demographic and baseline clinical variables: age at surgery, sex, BMI (calculated using the standard formula of weight in kg/ [height in meters]2), and prior diagnosis of diabetes mellitus, coronary artery disease, or hypertension. Cranial nerve deficits potentially resulting from intracranial hemorrhage were noted as well. In addition, presentation GCS scores were recorded. Computed tomography scans were analyzed to determine the type of hematoma, maximum thickness of the hematoma, and degree of midline shift of the brain. Thickness and midline shift were included in our data and reported in millimeters. Preoperative laboratory blood values were also recorded: including serum creatinine, hemoglobin, international normalized ratio of blood coagulation, platelet count, and partial thromboplastin time. Intraoperative parameters of interest included estimated blood loss (not including the volume of blood from the hematoma), operative time (defined as first incision to last surgical dressing), need of blood transfusion, and complications. Postoperative outcomes with which we were concerned included length of stay in the ICU until transfer to the general floor (ICU stay immediately after surgery was counted as Day 1), total length of stay in the hospital (date of admission to date of discharge), requirement for rehabilitation, and return to baseline status (defined as return of the patients normal physiological and mental status before the incident of intracranial bleeding). Return to baseline was measured subjectively by the associated neurosurgical care team based on patient self-report. Complications (which include any neurological deficit detected after surgery compared with findings on preoperative examination, whether transient or permanent, as well as any perioperative or postoperative events that required medical or surgical intervention up to 30 days after surgery) and 30-day postoperative mortality data were collected. Complications included cardiac complications (arrhythmias, myocardial infarct, and cardiac arrest), infection (urinary tract infection, pneumonia, wound infection, cellulitis, pseudomembranous colitis, and sepsis), reoperation, neurological complications (deficit, hydrocephalus, pseudomeningocele, seizure, and coma), other complications (renal failure, bleeding/anemia, pulmonary embolus, deep vein thrombosis, delirium, respiratory failure, ileus, and malignant hypertension), and death. This study was reviewed and approved by the Medical
235

Data

Methods and Materials

Electronic medical records, including patient records and intraoperative anesthesia records, at the University of Michigan Health System were queried to identify all

J Neurosurg / Volume 116 / January 2012

D. Lau et al.
Sciences Institutional Review Board of the University of Michigan.
Analysis

First, univariate statistics were used to describe our sample. Second, we employed bivariate chi-square tests to identify significant associations between covariates of interest and categorical outcomes (rehabilitation, return to baseline, complications, and mortality), along with ANOVA to assess significant associations between explanatory covariates of interest and continuous outcomes (ICU and total hospital stay). Third, multivariable logistic regression models of likelihood for rehabilitation, return to baseline, 30-day mortality, and complication risk were fit and were adjusted for potential confounders. We adjusted for covariates found to be significantly associated with postoperative outcomes in bivariate chi-square tests. Fourth, ANCOVA models of ICU stay and hospital stay by various covariates were fit and were adjusted for covariates found to be significantly associated with the outcomes of interest by ANOVA. Further analysis of the distribution of complications by type (cardiac, infection, reoperation, neurological, and other) and by age group (> 80 vs 80 years) was also performed. If a patient had a particular complication type, he or she was counted as 1 patient for that complication category. Patients with more than 1 type of complication were counted as 1 patient in each of the specific complication categories. Chi-square tests were used to assess the statistical significance of age-dependent differences in complication risk. Results with probability values less than 0.050 were considered significant. This study was powered to detect a minimum odds ratio of 1.92 for patients older than 80 years relative to those 80 years or younger with 80% power and a = 0.05. Posthoc power analysis was run using NCSS Power Analysis and Sample Size. All other statistical analyses were run using SAS 9.2 (SAS Institute).

Table 1 shows demographic characteristics and the results of bivariate chi-square tests between covariates and 30-day complications and mortality. Overall, 61 patients had complications (a complication rate of 59.2%). There were no reported intraoperative complications (all complications were postoperative). Difference in age was significantly associated with complications (p = 0.012). Patients older than 80 years had a complication rate of 81.5%, which was higher than those aged 5080 years and those younger than 50 years (58.3% and 45.0%, respectively). The overall mortality rate after craniotomy was 25.2%. Although patients older than 80 years had slightly higher absolute rates of mortality, there was no significant difference between the 3 age groups. Table 1 also describes covariates that are associated with return to baseline and need for rehabilitation. Age was not associated with significant differences in return to baseline (p = 0.816). Of 103 patients, 36.4% required rehabilitation.
236

Results

Age was significantly associated with requirement for rehabilitation (p = 0.037). In the groups of patients younger than 50 years and patients between 5079 years, 25.0% required rehabilitation; this percentage was significantly lower than in the group of patients who were older than 80 years (51.9%). Table 2 shows postoperative ICU and hospital stay, as well as the results of an ANOVA between explanatory covariates of interest and both outcomes of interest among patients in our sample. The mean postoperative ICU stay was 4.6 days in the total group (103 patients). Age was not associated with a significant difference in postoperative ICU stay (p = 0.137). The mean total hospital stay was 10.2 days in the total group. Older age was associated with longer hospital stay (p = 0.030). The mean duration of stay for patients younger than 50 years was 7.8 days, for those aged 5080 years it was 10.0 days, and for those over 80 years old it was 12.8 days. Table 3 shows multivariable models of postoperative outcomes risk by age, adjusted for potential confounders. Age greater than 80 years was significantly associated with increased risk of postoperative complications after adjusting for potential confounders. Compared with patients younger than 50 years, the odds ratio of complications among patients older than 80 years was 5.74 (95% CI 1.2925.34). Age was not, however, associated with significant risk for mortality. In addition, we found that patients older than 80 years were more likely to require rehabilitation compared with patients younger than 50 (OR 3.28, 95% CI 1.139.74). Both groups of patients 50 and older returned to baseline at a rate similar to the rate observed in patients younger than 50 years. Table 4 shows the results of ANCOVA models of postoperative ICU stay and hospital stay, adjusted for potential confounders. Both ICU and hospital stays were significantly longer in patients older than 80 years (p = 0.039 and p = 0.014, respectively, for comparison with patients younger than 50 years). Table 5 reports the specific postoperative complications in patients over 80 years old and patients 80 years or younger. The most common complication among the older group was infection; 37.0% of the 27 patients older than 80 years had at least 1 infection (Clostridium difficile causing pseudomembranous colitis, pneumonia, cellulitis, or sepsis). Age-dependent differences in risk for specific complications are shown in Fig. 1. Infection was the only category with statistically significant age-dependent differences in prevalence (p = 0.005). The prevalence of infection among the patients older than 80 years was 37.0% and the prevalence of infection among those 80 years of age or younger was 10.5%. There were no other statistically significant differences in complication rate by age: cardiac (p = 0.089), reoperation (p = 0.950), neurological (p = 0.979), and other (p = 0.829). In this study about the relation between age and postoperative outcomes following closed head injury and emergent craniotomy for evacuation of hematoma, we found that patients over 80 years of age had significantly
J Neurosurg / Volume 116 / January 2012

Discussion

Age and craniotomy following intracranial hemorrhage

TABLE 1: Descriptive statistics and results of bivariate chi-square tests between demographic and medical covariates and postoperative outcomes* Rehabilitation Pt Descriptives total age (yrs) <50 5079 >80 sex M F BMI <20 2025 >25 diabetes mellitus yes no coronary artery disease yes no hypertension yes no anticoagulation yes no cranial nerve deficit yes no GCS score 38 912 1315 type of hematoma intraparenchymal subdural epidural maximum hematoma thickness (mm) <10 1025 >25 midline shift of brain (mm) <5 510 >10 serum creatinine (mg/dl) <0.5 0.51.4 >1.4 No. of Pts 103 40 (38.8) 36 (35.0) 27 (26.2) 60 (58.3) 43 (41.8) 7 (8.0) 34 (38.6) 47 (53.4) 11 (10.7) 92 (89.3) 10 (9.7) 93 (90.3) 39 (37.9) 64 (62.1) 37 (35.9) 66 (64.1) 11 (10.7) 92 (89.3) 23 (22.3) 32 (31.1) 48 (46.6) 36 (35.0) 43 (41.8) 24 (23.3) 13 (14.3) 44 (48.4) 34 (37.4) 38 (40.0) 33 (34.7) 24 (25.3) 2 (2.2) 82 (89.1) 8 (8.7) No. of Pts 36 (35.0) 0.037 10 (25.0) 12 (33.3) 14 (51.9) 0.213 18 (30.0) 18 (41.9) 0.930 3 (42.9) 12 (35.3) 17 (36.2) 0.440 5 (45.5) 31 (33.7) 0.730 3 (30.0) 33 (35.5) 0.788 23 (35.9) 13 (33.3) 0.949 13 (35.1) 23 (34.8) 0.745 4 (36.4) 32 (34.8) 0.141 10 (43.5) 14 (43.8) 12 (25.0) 0.333 16 (44.4) 13 (30.2) 7 (29.2) 0.408 5 (38.5) 13 (29.6) 15 (44.1) 0.783 13 (34.2) 11 (33.3) 10 (41.7) 0.562 0 (0.0) 27 (32.9) 2 (25.0) 2 (100.0) 44 (53.7) 2 (25.0) 27 (71.1) 14 (42.4) 10 (41.7) 0.118 2 (100.0) 45 (54.9) 8 (100.0) 6 (46.2) 28 (63.6) 15 (44.1) 0.021 19 (50.0) 22 (66.7) 15 (62.5) 0.023 0 (0.0) 20 (24.4) 4 (50.0) (continued) 14 (38.9) 23 (53.5) 18 (75.0) 0.192 10 (76.9) 23 (52.3) 20 (58.8) 0.334 4 (10.5) 12 (36.4) 9 (37.5) 0.202 7 (30.4) 10 (31.3) 38 (79.2) 0.023 24 (66.7) 26 (60.5) 11 (45.8) 0.284 6 (46.2) 10 (22.7) 8 (23.5) 0.017 2 (18.2) 53 (57.6) <0.001 17 (73.9) 24 (75.0) 20 (41.7) 0.268 9 (25.0) 13 (30.2) 4 (16.7) 0.216 18 (48.7) 37 (56.1) 0.013 8 (72.7) 53 (57.6) 0.003 10 (43.5) 11 (34.4) 5 (10.4) 0.472 20 (51.3) 35 (54.7) 0.469 27 (73.0) 34 (51.5) 0.335 5 (45.5) 21 (22.8) 0.004 6 (60.0) 49 (52.7) 0.737 25 (64.1) 36 (56.3) 0.034 11 (29.7) 15 (22.7) 0.103 7 (63.6) 48 (52.2) 0.660 9 (90.0) 52 (55.9) 0.432 10 (25.6) 16 (25.0) 0.433 4 (57.1) 20 (58.8) 24 (51.1) 0.471 9 (81.8) 52 (56.5) 0.037 3 (30.0) 23 (24.7) 0.942 28 (46.7) 27 (62.8) 0.779 2 (28.6) 17 (50.0) 31 (66.0) 0.107 3 (27.3) 23 (25.0) 0.716 22 (55.0) 20 (55.6) 13 (48.2) 0.106 39 (65.0) 22 (51.2) 0.104 2 (28.6) 9 (26.5) 9 (19.2) 0.870 p Value Return to Baseline No. of Pts 55 (53.4) 0.816 18 (45.0) 21 (58.3) 22 (81.5) 0.159 20 (33.3) 6 (14.0) 0.687 p Value Complications No. of Pts 61 (59.2) 0.012 9 (22.5) 8 (22.2) 9 (33.3) 0.026 p Value Mortality No. of Pts 26 (25.2) 0.530 p Value

J Neurosurg / Volume 116 / January 2012

237

D. Lau et al.
TABLE 1: Descriptive statistics and results of bivariate chi-square tests between demographic and medical covariates and postoperative outcomes* (continued) Rehabilitation Pt Descriptives hemoglobin (g/dl) <13.5 13.516.5 >16.5 INR <0.8 0.81.2 >1.2 platelets ( 103 per l) <100 100450 >450 PTT (sec) <18 1828 >28 blood loss (ml) <500 5001000 >1000 transfusion yes no operative time (hrs) <2 2 No. of Pts 62 (63.3) 33 (33.7) 3 (3.1) 0 (0.0) 76 (83.5) 15 (16.5) 36 (36.4) 61 (61.6) 2 (2.0) 0 (0.0) 65 (72.2) 25 (27.8) 76 (73.8) 15 (14.6) 12 (11.7) 33 (32.0) 70 (68.0) 53 (51.5) 50 (48.5) No. of Pts 25 (40.32) 10 (30.3) 0 (0.0) 0.742 0 (0.0) 27 (35.3) 6 (40.0) 8 (22.2) 27 (44.3) 0 (0.0) 0.041 0 (0.0) 25 (38.5) 4 (16.0) 0.135 26 (34.21) 8 (53.3) 2 (16.7) 0.497 10 (30.3) 26 (37.1) 0.306 21 (39.6) 15 (30.0) 30 (56.6) 25 (50.0) 14 (42.4) 41 (58.6) 0.502 31 (58.5) 30 (60.0) 42 (55.3) 10 (66.7) 3 (25.0) 0.125 26 (78.8) 35 (50.0) 0.876 9 (17.0) 17 (34.0) 0 (0.0) 35 (53.9) 11 (44.0) 0.080 43 (56.6) 8 (53.3) 10 (83.3) 0.006 15 (45.5) 11 (15.7) 0.047 0 (0.0) 39 (51.3) 8 (53.3) 15 (41.7) 38 (62.3) 0 (0.0) 0.403 0 (0.0) 39 (60.0) 16 (64.0) 0.190 15 (19.7) 2 (13.3) 9 (75.0) 0.001 p Value 0.264 29 (46.8) 23 (69.7) 1 (33.3) 0.886 0 (0.0) 47 (61.8) 9 (60.0) 27 (75.0) 31 (50.8) 1 (50.0) 0.727 0 (0.0) 13 (20.0) 11 (44.0) <0.001 Return to Baseline No. of Pts p Value 0.078 40 (64.5) 16 (48.5) 2 (66.7) 0.893 0 (0.0) 20 (26.3) 4 (26.7) 15 (41.7) 9 (14.8) 1 (50.0) 0.021 Complications No. of Pts p Value 0.307 18 (29.0) 6 (18.2) 0 (0.0) 0.978 Mortality No. of Pts p Value 0.305

0.052

0.045

0.062

0.009

* Values in parentheses are percentages. Abbreviations: INR = international normalized ratio; Pt = Patient; PTT = partial thromboplastin time. Patients with more than 1 type of complication were counted as 1 patient in each of the specific complication categories.

higher rates of postoperative complications, had significantly longer hospital stays, and required rehabilitation more frequently than patients 80 years old or younger. However, the rate of mortality and return to baseline among this population was similar to what was found in younger patients. Taken together, our findings suggest that, compared with younger patients, older patients may be similarly able to return to functional baseline, but may require added postoperative medical attention in the form of rehabilitation and longer hospital stays. Similar likelihoods of return to functional baseline and nonsignificant differences in the rate and risk of mortality were unexpected, as past studies have demonstrated worse outcomes and higher mortality rates following neurotrauma in older patients than in younger patients.3,12,13 Our findings suggest that advancements in the management of intracranial hematomas, particularly postoperative ICU care and rehabilitation, may mitigate the increased risk of death traditionally associated with older age in past studies. There is some literature that has
238

supported these findings. A recent study showed promising outcomes in older patients after neurosurgical intervention for intracranial hemorrhage and hematoma.1 Also, a cohort study done by Baechli et al. 5 showed that of 354 patients, those older than 65 years of age had rates of mortality similar to those of younger patients after undergoing neurosurgical interventions for subdural hematoma. Age may not be an adequate indicator for likelihood of postsurgical recovery; rather a clinical metric of frailty32 and other preoperative parameters22 have been shown to be effective in predicting survival rates postoperatively. Also, regarding the value of neurosurgical treatment among the elderly, a recent study showed that patients 70 years and older with subarachnoid hemorrhages had favorable clinical outcomes, and concluded that age should not preclude treatment.4 Age above 80 years was a significant predictor of complications after emergent craniotomy for evacuation of hematoma in our study. In particular, patients older than 80 years had significantly higher rates of infection.
J Neurosurg / Volume 116 / January 2012

Age and craniotomy following intracranial hemorrhage

TABLE 2: Analysis of variance between demographic and medical covariates and outcomes* ICU Stay Pt Descriptives total age (yrs) <50 5080 >80 sex M F BMI <20 2025 >25 diabetes mellitus yes no coronary artery disease yes no hypertension yes no anticoagulation yes no cranial nerve deficit yes no GCS score 38 912 1315 type of hematoma intraparenchymal subdural epidural maximal hematoma thickness (mm) <10 1025 >25 midline shift of brain (mm) <5 510 >10 serum creatinine (mg/dl) <0.5 0.51.4 >1.4 Mean (days) 4.6 4.2 0.137 3.9 4.5 3.9 3.5 5.9 4.8 0.964 4.5 4.7 4.4 3.7 0.573 6.1 5.1 4.2 4.8 4.6 4.2 <0.001 8.2 7.2 4.0 3.6 0.017 7.5 6.2 4.1 3.9 0.428 4.9 5.6 4.2 3.3 0.202 4.0 4.0 5.2 4.7 0.650 5.0 6.4 4.4 4.0 0.719 4.6 3.6 4.9 4.8 4.1 4.3 0.734 4.6 4.7 4.6 3.9 3.8 4.5 0.356 5.8 5.4 4.1 3.8 4.0 3.6 0.210 5.4 5.4 4.2 3.4 3.5 3.5 0.242 9.5 0.7 4.3 4.4 5.0 4.8 13.5 0.7 9.7 7.9 10.4 9.9 (continued) 11.1 8.7 9.3 6.6 8.1 7.0 0.793 10.7 10.4 9.5 6.8 8.8 7.3 0.326 10.5 7.8 10.0 7.4 8.6 8.3 0.752 8.3 5.6 9.8 6.8 10.7 9.0 0.630 9.1 7.2 9.9 7.8 0.489 11.8 9.2 8.7 6.5 0.730 12.6 9.7 8.2 5.6 0.048 15.8 11.4 9.2 7.0 0.005 17.3 10.9 8.9 6.8 <0.001 8.9 6.6 8.1 7.0 12.1 8.7 <0.001 9.5 7.7 10.3 7.7 0.083 7.8 5.7 10.0 7.0 12.8 10.0 0.596 p Value Total Hospital Stay Mean (days) 10.2 7.6 0.030 p Value

J Neurosurg / Volume 116 / January 2012

239

D. Lau et al.
TABLE 2: Analysis of variance between demographic and medical covariates and outcomes* (continued) ICU Stay Pt Descriptives hemoglobin (g/dl) <13.5 13.516.5 >16.5 INR <0.8 0.81.2 >1.2 platelets ( 103 per l) <100 100450 >450 PTT (sec) <18 1828 >28 blood loss (ml) <500 5001000 >1000 transfusion yes no operative time (hrs) <2 2 * Mean values are given with SDs. Mean (days) 5.6 4.8 3.0 2.7 1.0 1.0 0.751 4.5 4.5 4.9 4.1 4.6 3.8 4.5 4.7 3.0 2.8 0.446 4.4 4.5 5.2 4.2 0.652 4.4 4.5 5.1 4.2 3.6 3.3 0.810 4.4 4.6 4.6 3.6 0.084 3.7 4.0 5.2 4.5 9.5 8.2 10.2 7.2 9.2 7.5 10.1 7.8 0.676 10.2 8.0 11.1 5.0 6.1 7.9 0.576 10.7 8.6 8.5 5.6 0.183 10.4 8.4 9.0 5.1 9.5 7.1 10.6 8.2 5.0 5.7 0.268 p Value 0.006 11.7 8.5 7.5 5.5 6.5 4.9 0.538 Total Hospital Stay Mean (days) p Value 0.029

0.876

0.532

This was expected and may be attributed to a less robust immune system and consequent higher risk for infection in older patients.18,29 Moreover, the natural symbiotic flora may be less able to defend against pathogenic bacteria that colonize the intestines as the gastrointestinal tract changes with age,6 making older patients more susceptible to nosocomial infections. The higher rate of postoperative complications we found may also contribute to significantly longer hospital stays as well,19 which has been suggested in past studies.23,28 However, even in the presence of increased risk of complications, rates of mortality and functional recovery were similar between all age groups. An article by Ghaferi et al.11 discussed the concept of failure to rescue which essentially describes mortality after postoperative complicationsour findings suggest that the management of postoperative complications in older patients has been effective in steering them through their postoperative course. There are several limitations concerning the findings of the presented research. First, this study was retrospective; therefore, it is plausible that the associations we noted may have been confounded by underlying factors
240

not explicitly studied, but associated with both exposures and outcomes. Second, the study included analysis of data from patients in one context and may limit the generalizability of the findings. Third, although we based our delineation of age groups on previous research in the area, because of our treatment of age as a categorical rather than continuous variable, our findings do not suggest any linearity or highlight trends with regard to the relation between age and outcomes following craniotomy for intracranial hemorrhage. Despite these limitations, our findings have important implications for clinical management as well as future research. As previously established, intracranial hemorrhage and hematoma can lead to neurological deterioration, death, or both. Even with emergent medical and surgical interventions, morbidity and mortality rates are high. Our work may be of interest to clinicians and healthcare policymakers when considering the surgical management of intracranial bleeding in older patients. In complement to prompt diagnosis, which can prevent morbidity,14 our findings suggest that rehabilitation and longer hospital stays among older patients may yield outcomes
J Neurosurg / Volume 116 / January 2012

TABLE 3: Logistic regression models of postoperative outcomes risks by risk factor adjusted for potential confounders* Rehabilitation 95% CI ref 0.514.18 1.139.74 8.00 ref 4.31 ref 0.82 ref 0.24 ref 0.09 0.12 ref 0.21 0.32 ref ref 0.67 0.35 ref 0.22.14 0.091.37 ref 0.494 0.139 1.20 ref NA 0.25 ref NA 0.070.91 ref NA 0.028 ref NA 0.413.66 ref NA 0.756 ref NA 2.83 ref 5.23 0.5216.83 ref 0.13253.71 0.252 ref 0.403 (continued) 0.041.01 0.071.53 ref 0.066 0.146 ref ref 1.14 3.52 ref 0.226.25 1.4131.18 ref 0.877 0.026 0.030.38 0.030.54 ref <0.001 0.006 ref 7.48 3.14 ref 2.4523.11 0.8211.44 ref 0.041.41 ref 0.117 ref <0.001 0.092 ref 23.64 2.85 ref 2.40228.34 0.330.6 ref 0.006 0.387 ref 0.242.91 ref 0.755 ref 0.3254.37 ref 0.225 ref 2.3441.32 ref ref 0.452 0.033 ref 1.01 1.12 ref 0.333.43 0.294.34 ref 0.987 0.870 ref 2.03 5.74 ref 0.716.32 1.2925.34 ref 0.225 0.028 ref 1.19 1.16 ref 0.5221.77 0.197.91 ref 0.238 0.881 0.006 ref p Value OR 95% CI p Value OR 95% CI p Value OR 95% CI p Value Return to Baseline Complications Mortality

Pt Descriptives

OR

J Neurosurg / Volume 116 / January 2012

Age and craniotomy following intracranial hemorrhage

age (yrs) <50 5080 >80 sex M F coronary artery disease yes no anticoagulation yes no cranial nerve deficit yes no GCS score 38 912 1315 type of hematoma intraparenchymal subdural epidural midline shift of brain (mm) <5 510 >10 serum creatinine (mg/dl) <0.5 0.51.4 >1.4 platelets ( 103 per l) <100 100450 >450

ref 1.49 3.28

241

D. Lau et al.
p Value 0.157 ref 0.004 ref 0.274 0.030 0.022 ref ref 0.164 TABLE 4: Analysis of covariance models of hospital stay by various risk factors and predictors adjusted for potential confounders ICU Stay Mortality 0.54121.12 ref 3.0195.80 ref 0.013.74 1.3290.55 1.4199.84 ref ref 0.6214.74 95% CI Pt Descriptives age (yrs) <50 5079 >80 diabetes mellitus yes no coronary artery disease yes no hypertension yes no anticoagulation yes no hemoglobin (g/dl) <13.5 13.516.5 >16.5 Mean (days) p Value 3.9 4.5 3.9 3.5 5.9 4.8 8.2 7.2 4.0 3.6 ref 0.985 0.039 0.007 ref Total Hospital Stay Mean (days) 7.8 5.7 10.0 7.0 12.8 10.0 17.3 10.9 8.9 6.8 p Value ref 0.177 0.014 0.008 ref

11.96 ref 0.140 ref

p Value

OR

ref 3.06

7.47 ref 9.55

ref 0.2 9.65

TABLE 3: Logistic regression models of postoperative outcomes risks by risk factor adjusted for potential confounders* (continued)

Complications

0.837.30 ref

95% CI

7.5 6.2 4.1 3.9

0.100 ref

15.8 11.4 9.2 7.0 12.6 9.7 8.2 5.6 11.8 9.2 8.7 6.5

0.102 ref 0.445 ref 0.567 ref 0.210 ref 0.934

2.35 ref

OR

Return to Baseline

5.6 4.8 3.0 2.7 1.0 1.0

0.043 ref 0.456

11.7 8.5 7.5 5.5 6.5 4.9

p Value Rehabilitation 95% CI

OR 0.232.86 ref 0.080.89 0.692 ref 0.032

95% CI

p Value

* NA = not applicable; ref = reference.

PTT (sec) <18 1828 >28 blood loss (ml) <500 5001000 >1000 transfusion yes no operative time (hrs) <2 2

Pt Descriptives

0.76 ref 0.27

OR

FIG. 1. Comparison of postoperative complication rate by category between patients > 80 versus 80 years old with closed head injury who underwent craniotomy for evacuation of hematoma. *Infection was the only category with a statistically significant difference in rate of complications (p = 0.005). The difference in rate for cardiac complications was close to being significant (p = 0.089); the rates for reoperation (p = 0.950), neurological complications (p = 0.979), and other complications (p = 0.829) were not.

242

J Neurosurg / Volume 116 / January 2012

Age and craniotomy following intracranial hemorrhage

TABLE 5: Comparison of postoperative complications in patients stratified by age* Age > 80 Yrs (27 pts) 1 ptUTI 1 pt MI 1 ptpneumonia 1 ptND & cardiac arrhythmia 1 ptPE & DVT 1 ptatrial flutter, pneumonia, & delirium 1 ptND, reop, & pneumonia 1 ptpneumonia, respiratory failure, ventilation, & deep wound infection 1 ptND, bacterial pneumonia, pseudomeningocele, & car diac atrial arrhythmia 1 ptND, pneumonia, pseudomeningocele, & reop 1 ptleft upper-extremity DVT, right lower-extremity DVT, C. diff. infection, MI, & cardiac arrhythmia 1 ptND & bacterial sepsis; required postop intubation & mechanical ventilation, & reop 1 ptileus, C. diff. infection, hydrocephalus, cellulitis, pneu monia, & reop 9 ptsdeath Age 80 Yrs (76 pts) 1 ptUTI (Pseudomonas) 1 ptseizure 1 ptpneumothorax 1 ptwound infection 1 ptmalignant hypertension & reop 1 ptDVT & pneumonia 1 ptpneumonia & reop 1 ptchronic recurrent hematoma & reop 1 ptmalignant hypertension & reop 1 ptventilation & reop 1 ptND & reop 1 ptcoma & ventilation 1 ptND, cardiac arrhythmia, & reop 1 ptND, cardiac arrhythmia, & reop 1 ptanemia requiring transfusion, seizure, & deep wound infection 1 ptcoma, deep infection, intubation, ventilation, & reop 1 ptND, pneumonia, bacterial sepsis, intubation, & ventilation 1 ptND, Stenotrophomonas pneumonia, complete heart block, seizures, hypernatremia, ARF, & mechanical ventilation 1 ptARF, bleed requiring > 4 units of blood, reop, cardiac ar rest, & cardiac arrhythmia 3 ptsND 17 ptsdeath

* Complications occurred in 22 (81%) of the 27 patients in the > 80 age group and 39 (51%) of the 76 patients in the 80 age group. Abbreviations: ARF = acute renal failure; C. diff. = Clostridium difficile; DVT = deep vein thrombosis; MI = myocardial infarct; ND = neurological deficit; PE = pulmonary embolus; UTI = urinary tract infection.

similar to those of younger patients after craniotomy for evacuation of intracranial hemorrhage and hematoma. In addition, our findings may be important for endof-life care guidelines, more broadly. Craniotomy remains a valid and safe technique for the management of older patients with emergent intracranial bleeds14 and is particularly important37 and successful in treating elderly patients,7 but is an expensive procedure.21 It has been reported that the average craniotomy procedure itself costs $15,867 at high-volume centers.21 Nevertheless, taken with findings that have demonstrated that quality-of-life scores remain high for elderly survivors after undergoing craniotomy for intracranial hemorrhage,30 our findings suggest that with some added care, patients over the age of 80 may recover as successfully as younger patients and that these costs may be warranted among this population. Our findings also demonstrated that age greater than 80 years predicted the need for rehabilitation after undergoing craniotomy. A study by Cope and Hall10 has
J Neurosurg / Volume 116 / January 2012

suggested that implementing rehabilitation programs after head injury is effective, and in light of our findings, rehabilitation programs that cater specifically to older patients may further improve postoperative recovery and may be warranted. Investigators interested in postneurosurgical sequelae in patients over the age of 80 might consider the relation between age and similar outcomes among patients undergoing different interventions such as spinal fusions, aneurysms requiring craniotomy, and craniotomy for tumor resection. In addition, large prospective studies concerned with the relation between age and intra- and postoperative outcomes following craniotomy for intracranial hemorrhage are needed. Compared with younger patients, patients over 80 years of age may have a higher risk for postoperative
243

Conclusions

D. Lau et al.
complications and significantly longer ICU and hospital stays after undergoing craniotomy for evacuation of hemorrhage and hematoma. However, with rehabilitation, older patients are able to return to their functional baseline at rates similar to those of younger patients within a 30day period and have no added 30-day mortality risk. Our findings may allow for more informed decisions when approaching the care of older patients with intracranial hemorrhage.
Disclosure This study was funded in part by grants from the Rhodes Trust (to A.M.E.-S.) and the NIH (HSO 18537-01 to J.E.Z.). The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper. Author contributions to the study and manuscript preparation include the following. Conception and design: Lau, Ziewacz. Acquisition of data: Lau, Jayachandran, Huq, Zamora-Berridi, Da vis. Analysis and interpretation of data: Lau. Drafting the article: Lau. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: El-Sayed. Statistical analysis: El-Sayed, Lau. Administrative/technical/material support: El-Sayed, Lau. Study supervision: Ziewacz, El-Sayed. References 1.Adhiyaman V, Asghar M, Ganeshram KN, Bhowmick BK: Chronic subdural haematoma in the elderly. Postgrad Med J 78:7175, 2002 2. Andrews BT, Chiles BW III, Olsen WL, Pitts LH: The effect of intracerebral hematoma location on the risk of brain-stem compression and on clinical outcome. J Neurosurg 69:518 522, 1988 3.Asghar M, Adhiyaman V, Greenway MW, Bhowmick BK, Bates A: Chronic subdural haematoma in the elderlya North Wales experience. J R Soc Med 95:290292, 2002 4. Awe OO, Gonzalez LF, Hasan D, Maltenfort M, Rossenwasser R, Jabbour P: Treatment outcome of aneurysmal subarachnoid hemorrhage in patients aged 70 years and older. Neurosurgery 68:753758, 2011 5. Baechli H, Nordmann A, Bucher HC, Gratzl O: Demographics and prevalent risk factors of chronic subdural haematoma: results of a large single-center cohort study. Neurosurg Rev 27: 263266, 2004 6. Biagi E, Nylund L, Candela M, Ostan R, Bucci L, Pini E, et al: Through ageing, and beyond: gut microbiota and inflammatory status in seniors and centenarians. PLoS ONE 5:e10667, 2010 7.Broderick JP: Intracerebral hemorrhage, in Gorelick PB, Alter M (eds): Handbook of Neuroepidemiology. New York: Marcel Dekker, 1994, pp 141167 8.Broderick JP, Brott TG, Duldner JE, Tomsick T, Huster G: Volume of intracerebral hemorrhage. A powerful and easy-touse predictor of 30-day mortality. Stroke 24:987993, 1993 9. Brott T, Broderick J, Kothari R, Barsan W, Tomsick T, Sauerbeck L, et al: Early hemorrhage growth in patients with intracerebral hemorrhage. Stroke 28:15, 1997 10. Cope DN, Hall K: Head injury rehabilitation: benefit of early intervention. Arch Phys Med Rehabil 63:433437, 1982 11. Ghaferi AA, Birkmeyer JD, Dimick JB: Variation in hospital mortality associated with inpatient surgery. N Engl J Med 361:13681375, 2009 12. Hanif S, Abodunde O, Ali Z, Pidgeon C: Age related outcome in acute subdural haematoma following traumatic head injury. Ir Med J 102:255257, 2009 13. Howard MA III, Gross AS, Dacey RG Jr, Winn HR: Acute subdural hematomas: an age-dependent clinical entity. J Neurosurg 71:858863, 1989 14. Iantosca MR, Simon RH: Chronic subdural hematoma in adult and elderly patients. Neurosurg Clin N Am 11:447454, 2000 15. Janny P, Papo I, Chazal J, Colnet G, Barretto LC: Intracranial hypertension and prognosis of spontaneous intracerebral haematomas. A correlative study of 60 patients. Acta Neurochir (Wien) 61:181186, 1982 16. Kase CS, Crowell RM: Prognosis and treatment of patients with intracerebral hemorrhage, in Kase CS, Caplan LR (eds): Intracerebral Hemorrhage. Boston: Butterworth-Heine mann, 1994, pp 467489 17. Kim KH: Predictors for functional recovery and mortality of surgically treated traumatic acute subdural hematomas in 256 patients. J Korean Neurosurg Soc 45:143150, 2009 18. Kovacs EJ, Palmer JL, Fortin CF, Flp T Jr, Goldstein DR, Linton PJ: Aging and innate immunity in the mouse: impact of intrinsic and extrinsic factors. Trends Immunol 30:319324, 2009 19. Lawrence VA, Hilsenbeck SG, Mulrow CD, Dhanda R, Sapp J, Page CP: Incidence and hospital stay for cardiac and pulmonary complications after abdominal surgery. J Gen Intern Med 10:671678, 1995 20.Lisk DR, Pasteur W, Rhoades H, Putnam RD, Grotta JC: Early presentation of hemispheric intracerebral hemorrhage: prediction of outcome and guidelines for treatment allocation. Neurology 44:133139, 1994 21.Long DM, Gordon T, Bowman H, Etzel A, Burleyson G, Betchen S, et al: Outcome and cost of craniotomy performed to treat tumors in regional academic referral centers. Neurosurgery 52:10561065, 2003 22. Petridis AK, Drner L, Doukas A, Eifrig S, Barth H, Mehdorn M: Acute subdural hematoma in the elderly; clinical and CT factors influencing the surgical treatment decision. Cen Eur Neurosurg 70:7378, 2009 23. Polanczyk CA, Marcantonio E, Goldman L, Rohde LE, Orav J, Mangione CM, et al: Impact of age on perioperative complications and length of stay in patients undergoing noncardiac surgery. Ann Intern Med 134:637643, 2001 24. Qureshi AI, Mohammad YM, Yahia AM, Suarez JI, Siddiqui AM, Kirmani JF, et al: A prospective multicenter study to evaluate the feasibility and safety of aggressive antihypertensive treatment in patients with acute intracerebral hemorrhage. J Intensive Care Med 20:3442, 2005 25. Qureshi AI, Tuhrim S, Broderick JP, Batjer HH, Hondo H, Hanley DF: Spontaneous intracerebral hemorrhage. N Engl J Med 344:14501460, 2001 26. Rathlev NK, Medzon R, Lowery D, Pollack C, Bracken M, Barest G, et al: Intracranial pathology in elders with blunt head trauma. Acad Emerg Med 13:302307, 2006 27. Sahni R, Weinberger J: Management of intracerebral hemorrhage. Vasc Health Risk Manag 3:701709, 2007 28. Sawada T, Kita J, Rokkaku K, Kato M, Shimoda M, Kubota K: Outcome of surgical resection for hilar cholangiocarcinoma in elderly patients. Hepatogastroenterology 55:19711974, 2008 29. Shaw AC, Joshi S, Greenwood H, Panda A, Lord JM: Aging of the innate immune system. Curr Opin Immunol 22:507 513, 2010 30. Stachniak JB, Layon AJ, Day AL, Gallagher TJ: Craniotomy for intracranial aneurysm and subarachnoid hemorrhage. Is course, cost, or outcome affected by age? Stroke 27:276281, 1996 31.Stieg PE, Kase CS: Intracranial hemorrhage: diagnosis and emergency management. Neurol Clin 16:373390, 1998 32.Sndermann S, Dademasch A, Praetorius J, Kempfert J, Dewey T, Falk V, et al: Comprehensive assessment of frailty for elderly high-risk patients undergoing cardiac surgery. Eur J Cardiothorac Surg 39:3337, 2011

244

J Neurosurg / Volume 116 / January 2012

Age and craniotomy following intracranial hemorrhage

33. Taussky P, Widmer HR, Takala J, Fandino J: Outcome after acute traumatic subdural and epidural haematoma in Switzerland: a single-centre experience. Swiss Med Wkly 138:281285, 2008 34. Tian HL, Chen SW, Xu T, Hu J, Rong BY, Wang G, Gao WW, et al: Risk factors related to hospital mortality in patients with isolated traumatic acute subdural haematoma: analysis of 308 patients undergone surgery. China Med J (Engl) 12:1080 1084, 2008 35. Tuhrim S, Horowitz DR, Sacher M, Godbold JH: Validation and comparison of models predicting survival following intracerebral hemorrhage. Crit Care Med 23:950954, 1995 36. Vyas NA, Chicoine MR: Extended survival after evacuation of subdural hematoma in a 102-year-old patient: case report and review of the literature. Surg Neurol 67:314317, 2007 37. White M, Mathieson CS, Campbell E, Lindsay KW, Murray L: Treatment of chronic subdural haematomas - a retrospective comparison of minicraniectomy versus burrhole drainage. Br J Neurosurg 24:257260, 2010 38. Wright JD: The graying of America: implications for health professionals. Care Manag J 6:178184, 2005 39.Yang GY, Betz AL, Chenevert TL, Brunberg JA, Hoff JT: Experimental intracerebral hemorrhage: relationship between brain edema, blood flow, and blood-brain barrier permeability in rats. J Neurosurg 81:93102, 1994

Manuscript submitted March 7, 2011. Accepted July 18, 2011. Please include this information when citing this paper: published online August 26, 2011; DOI: 10.3171/2011.7.JNS11396. Address correspondence to: Abdulrahman M. El-Sayed, B.S., De partment of Epidemiology, Columbia University, 722 West 168th Street, 15th Floor, New York, New York 10031. email: ame2145@ columbia.edu.

J Neurosurg / Volume 116 / January 2012

245