Craniotomy for Intracranial Aneurysm and Subarachnoid Hemorrhage
Is Course, Cost, or Outcome Affected by Age?
Background and Purpose Age may influence cost or effectiveness of treatment for subarachnoid hemorrhage (SAH). This study examined the effect of age on both.
Methods Patients (n=219) who underwent craniotomy for intracranial aneurysm and SAH over 6 years at one tertiary care center were divided in two ways by age: single advanced age (<65 years and ≥65 years) and decade of age (23 to 39, 40 to 49, 50 to 59, 60 to 69, and 70 to 81 years). Data recorded for each patient included numbers of procedures and complications in the surgical intensive care unit (SICU), number of days in the SICU and the hospital, costs for SICU and ward care, total cost (SICU plus ward costs), and the Acute Physiology and Chronic Health Evaluation (APACHE) II score at admission and discharge, the Hunt-Hess grade at admission and immediately preoperatively, and quality of life score, a measure of outcome. Mortality rates by age group were calculated.
Results The only variable significantly affected by decade of age was mortality rate, which increased as decade of age increased (3% to 17%). With the 65-year comparison, mortality rate, cost, APACHE II score at admission and discharge, days before operation, and days in the SICU were significantly higher for age ≥65 years.
Conclusions Whereas mortality is higher for the older age group, quality of life scores appear acceptable for those who survive. Even though the hospital costs of treating elderly patients for SAH may be higher than those for younger patients, this should not be used to justify withholding care from the elderly.
The International Cooperative Study on the Timing of Aneurysm Surgery1 suggested that multiple factors contribute to mortality from SAH, including location of the aneurysm (eg, basilar aneurysm), thickness of the subarachnoid clot on CT, level of consciousness, high blood pressure, coexisting illness, and increased age. The importance of many of these factors has come under debate with recent advances in microneurosurgery and perioperative intensive care. In particular, disagreement exists as to the effect of age on ultimate outcome.1 2 3 4 5 6 Age is a substantial issue because the elderly represent an increasing proportion of the population and therefore may account for a greater proportion of total healthcare costs than do patients of other ages. Thus, we studied whether age significantly affects the course, cost, or final outcome of craniotomy for intracranial aneurysm and SAH.
Subjects and Methods
Patients (n=219) with SAH from a ruptured intracranial aneurysm who underwent craniotomy and clipping of the aneurysm between February 1989 and December 1994 were grouped by age in two ways: (1) a single older age, <65 years of age (AGE<65) and those ≥65 years (AGE≥65); and (2) decade of age, 23 to 39 years of age (AGE<40), 40 to 49 years (AGE40), 50 to 59 years (AGE50), 60 to 69 years (AGE60), and 70 to 81 years (AGE≥70). The study consisted of a combined prospective/retrospective analysis. Contemporaneously collected data were age, sex, race, diagnoses on admission to the hospital and SICU, sequential APACHE II scores, discharge status, and numbers of complications, procedures, and days in the SICU. Retrospectively collected data were location of aneurysm, incidence of ventriculotomy-dependent hydrocephalus, and costs of SICU and non-SICU care and combined cost. Each patient’s condition and outcome were scored in three ways. Two measures of morbidity were used. For severity of illness, APACHE II scores at admission and at discharge from the SICU were recorded.7 For severity of neurological insult, HH grade8 at admission and immediately before operation was recorded as follows: grade 0, no hemorrhage, asymptomatic aneurysm; grade 1, SAH with no neurological deficit, possible nuchal rigidity and headache; grade 2, moderate to severe headache, nuchal rigidity, no neurological deficit other than cranial nerve palsy; grade 3, drowsiness, confusion, or mild focal deficit; grade 4, stupor, moderate to severe hemiparesis, and possibly early decerebrate rigidity and vegetative disturbances; and grade 5, coma, decerebrate rigidity, and moribund appearance.
Outcome was assessed by QOL score as follows: grade 1, normal lifestyle or a return to that before SAH; grade 2, minor neurological dysfunction but activities of daily living can be performed without help; grade 3, needs help with some activities of daily living; grade 4, unable to perform activities of daily living and requires full-time care, eg, in a rehabilitation center; and grade 5, death.
Mortality rate was calculated for each age group.
With a computerized statistical analysis system (SigmaStat version 1.01, Jandel Corp), all data were evaluated for normality of distribution. Continuous data not normally distributed were analyzed by Mann-Whitney rank sum, Wilcoxon signed rank, and Kruskal-Wallis tests. Normally distributed data were analyzed by the unpaired t test. APACHE II and HH scores on admission, before operation, and at discharge were evaluated with ANOVA. All costs and durations of care were treated as non–repeated measures in each population and were analyzed with the Mann-Whitney rank sum test. Proportional data, such as QOL score, were analyzed with either the χ2 or the Fisher’s exact test. Multiple regression analysis was used to determine which variable or group of variables most closely correlated with QOL score. Statistical significance was set at P≤.05.
Of the 219 patients studied, women outnumbered men and the predominate races were white and African American (Table 1⇓), but QOL score did not differ by sex or race. The slightly smaller number of posterior communicating artery aneurysms and the larger number of basilar aneurysms were likely due to the referral pattern for our tertiary care center. QOL score also did not differ significantly by location of aneurysm (Table 2⇓). There was a trend, however, suggesting that posterior circulation and ophthalmic segment artery aneurysms (many of which were giant) were associated with a slightly higher QOL score. For all ages, the most common complication was hypertension, followed by cerebral vasospasm (Table 3⇓). Patients with ventriculostomy-dependent hydrocephalus had a slightly higher median QOL score9 than those without hydrocephalus,10 but the difference was not significant.
Duration of SICU care, which ranged from 4 to 10.5 days, did not differ by decade of age; however, with the 65-year-based division, AGE≥65 had significantly more SICU days than AGE<65 (Table 4⇓). Number of days in the SICU before surgery was significantly higher in AGE60 than the two youngest groups and than the oldest group (5 versus 2 days). This difference in SICU days before surgery also occurred in the 65-year-based groups (2 versus 3 days, P=.017). Number of procedures and complications in the SICU did not differ by age group (Table 4⇓). Cost was significantly greater for AGE≥65 (Table 4⇓); costs tended to increase as decade of age increased, but this did not achieve statistical significance.
Morbidity and Outcome
APACHE II scores were significantly higher in most older groups compared with younger groups at admission and discharge (Table 4⇑); this was true for both the decade and 65-year comparisons. On admission and immediately before surgery, HH grade did not differ significantly by age (Table 4⇑). Compared with AGE<65, the AGE≥65 group had a significantly higher mortality (13% versus 4.6%) and QOL score (2 versus 1) (Table 4⇑). Mortality by decade ranged from 3% to 17%, the only significant difference occurring between AGE50 and AGE≥70 (Table 4⇑).
Power analysis suggested that each group contained a sufficient number of patients to show differences in outcome if they existed. Although analysis showed that mortality rate increased with AGE≥70, both mean and median morbidity statistics failed to differ between any age groups except AGE50 and AGE≥70 (Table 4⇑). The final HH grade–weighted QOL showed that higher HH grade at admission, regardless of age, was significantly associated with higher QOL score (P≤.05).
Multiple linear regression was performed on the following variables to evaluate their impact on QOL outcome and thus their impact on outcome: age, number of days between hemorrhage and surgery, HH grade on admission and just before surgery, APACHE II score on admission and discharge, non-SICU and SICU costs, and days and number of complications and procedures in the SICU. Only about 35% of the variation in outcome could be accounted for by these variables (r=.592, r2=.35).
SAH, most commonly secondary to aneurysmal rupture, is a subcategory of cerebrovascular accident, the third leading cause of death in the United States.8 Epidemiological studies suggest that SAH causes 8% to 10% of cerebrovascular accidents and 4% to 5% of related deaths.5 Of approximately 28 000 people who suffer SAH yearly in North America, 18 000 will die or be disabled11 ; 8% to 15% die before receiving medical care, 35% within the first 8 hours, 20% to 37% within 48 hours, and 42% to 61% within the first month.12 Overall mortality ranges from 6% to 52%.1 4 9 12 13 Furthermore, autopsy studies suggest that at least 1% of adults harbor unruptured aneurysms and approximately 0.01% to 0.02% suffer SAH14 ; thus, unruptured aneurysms outnumber ruptured aneurysms. Which unruptured saccular aneurysms should be surgically treated is a matter of debate among neurosurgeons and neurologists.14 15 16 17 The International Cooperative Study on aneurysms suggested that the following factors relate significantly to mortality: thickness of the subarachnoid clot on CT, basilar aneurysm, coexisting medical illness, level of consciousness, increased blood pressure, and advanced age.1 Progress in microneurosurgery and perioperative intensive care may have mitigated some of these factors, in particular age.3 10 12 15 18 19
Studies have reported that the incidence of SAH increases with age.6 20 Even before those studies, however, it seemed intuitive that older patients must necessarily have a worse outcome secondary to comorbidity. If this were true, then what age group has the greatest surgical risk or at what age does risk begin to become “greater”? Several large studies have compared patients by age in an attempt to answer this question (Table 5⇓). In general, patients with higher HH grade and significant underlying medical problems do worse than those without these risk factors. Even though early studies1 4 5 15 19 suggested that age was an independent risk factor for worse outcome, better data now available2 6 18 21 indicate that surgical outcome for elderly patients is not necessarily poor. Functional status at 6 months to 1 year does appear to worsen with HH grade and age.
As several other investigators have done,1 2 4 6 15 18 21 we also examined the effect of clinical grade in the present study. Although scoring systems have flaws11 when used for predictive purposes, as has occurred with the APACHE II system, these systems can provide an objective barometer of the degree of comorbidity. Clinical grade as represented by HH score did not differ by age, which suggests that the severity of the SAH was a more important determinant than age. APACHE II scores, however, did vary by age: they were higher both at admission and at discharge with AGE≥65 and were higher for the two older decade groups compared with the two younger decade groups. Therefore, it was surprising that number of complications or duration of hospitalization did not differ by age. This emphasizes that, while risk related to comorbidity can be greater for the elderly, this is not a given.
With the onset of managed care, concern with cost-effectiveness in health care has increased.12 We wanted to determine whether our elderly patients were in fact more expensive to treat and whether outcome was significantly worse. Cost was, indeed, significantly higher in every area of care for AGE≥65 than AGE<65 (Table 4⇑); when evaluated by decade, however, there were no significant cost differences.
In the comparison by decade of age, mortality was increased in a clinically significant manner only with the oldest age group and in the overall comparison (Table 4⇑). Outcome also worsened with age; however, this difference is unlikely to be clinically significant. All of the older patients remained independent and able to continue to perform the activities of daily living.
Although mortality in our study was low and outcome generally excellent compared with the results reported in many previous studies, neurosurgeons may nonetheless need to reassess their understanding of a “good” outcome. Previous studies have primarily based evaluation of surgical results on mortality and some determinant of functional status, such as the Glasgow Coma Scale or QOL. Recently, more sophisticated neuropsychological testing has shown that patients previously considered to have had a good outcome in fact have significant impairment of cognitive function.22 23 For example, nonverbal memory deficits can be frequent, severe, and persistent, with 47% of patients having moderate-to-severe impairment 1 year after craniotomy for aneurysmal SAH.23 Problems with speed of response and visuospatial and frontal lobe abilities occurred and were more profound in the elderly, which suggests that they suffered greater long-term effects from SAH. These data, combined with the fact that fewer than 50% of those who suffer SAH return to work, suggest that tremendous work remains to be done in the treatment and rehabilitation of persons who suffer SAH.
We think that the results from our study combined with other recent findings18 suggest that elderly patients, certainly those older than 65 years, can benefit from aggressive treatment. Although mortality is worse for older patients, outcome of patients who survive is acceptable (data on cognitive impairment23 notwithstanding). Many of these patients will require rehabilitation but can, with aggressive postoperative care, continue to lead useful and productive lives. Thus, the analysis of data from our institution indicates that aggressive treatment of the elderly is appropriate unless and until the patient or family/surrogate determines that, for nonmedical reasons, nonsurgical treatment is preferred. Finally, even though the hospital costs for treating SAH may be higher in elderly patients, whether access to care should be based on the cost of care provision is a moral and ethical question and is not answerable strictly by accounting principles.
Selected Abbreviations and Acronyms
|APACHE||=||Acute Physiology and Chronic Health Evaluation|
|QOL||=||quality of life|
|SICU||=||surgical intensive care unit|
Special thanks to B.E. George and Brenda Hamby for their help in data collection and Lynn Dirk for editorial assistance.
Presented in part at the Congress of Neurological Surgeons, San Francisco, Calif, October 14-19, 1995.
- Received September 1, 1995.
- Revision received November 1, 1995.
- Accepted November 1, 1995.
- Copyright © 1996 by American Heart Association
Martindale B, Garfield J. Subarachnoid haemorrhage above the age of 59: are intracranial investigations justified? Br Med J. 1978;25:465-466.
Sacco RL, Wolf PA, Bharucha NE, Meeks SL, Kannel WB, Charette LJ, McNamara PM, Palmer EP, D’Agostino R. Subarachnoid and intracerebral hemorrhage: natural history, prognosis, and precursive factors in the Framingham study. Neurology. 1984;34:847-854.
Civetta J, Kirby R. Prediction and definition of outcome. In: Stoelting RK, Barash P, Gallagher TJ, eds. Advances in Anesthesia. St Louis, Mo: Mosby-Year Book; 1992:137-194.
Kilner J. Age as a basis for allocating lifesaving medical resources: an ethical analysis. J Health Polit Policy Law. 1988;13:405-423.
Hijdra A, Braackman R, van Gijn J, Vermeulen M, van Crevel H. Aneurysmal subarachnoid hemorrhage complications and outcome in a hospital population. Stroke. 1987;18:1061-1067.
Inagawa T, Takahashi M, Aoki H, Ishikawa S, Yoshimoto H. Aneurysmal subarachnoid hemorrhage in Izumo City and Shimane prefecture of Japan outcome. Stroke. 1987;19:176-180.
Ingall TJ, Whisnant JP, Weibers DO, O’Fallon WM. Has there been a decline in subarachnoid hemorrhage mortality? Stroke. 1989;20:718-724.
McKissock W, Paine KWE, Walsh LS. An analysis of the results of treatment of ruptured intracranial aneurysms: report of 772 consecutive cases. J Neurosurg. 1960;17:762-776.