Chronic subdural hematoma patients aged 90 years and older

Abstract

Objectives:

The American population above 65 years of age will double by 2050, and more nonagenarians will present to neurosurgeons for treatment for subdural hematomas (SDH), common in the elderly. Healthcare providers, and patients’ relatives, often choose treatment when there is little chance of recovery. Hospital mortality is 24% (n=5) in chronic subdural hematoma (cSDH) patients over 65 years, but there are no studies on cSDH outcomes in patients aged over 90 years. This retrospective study evaluates outcomes in this population.

Methods:

We reviewed all patients with cSDH between December 2005 and December 2011. We analyzed charts of patients aged 90 years and older. Patient demographics, Glagow Coma Scale (GCS) at presentation, medical co-morbidities, length of stay, disposition, treatment, and radiographic characteristics were abstracted.

Results:

Twenty-one patients aged 90 or older with 24 admissions for cSDH were identified. Median age was 92 (SD=2.5); 76% (n=16) underwent surgery. Median presentation GCS was 14. Disposition to home, rehabilitation facility, nursing home, hospice, or death were not significantly different between conservative and operative groups (P=0.10), nor was admission GCS (P=0.59). The size of SDH was significantly (P=0.02) larger in the operative group. Overall, only 24% (n=5) of patients were discharged home.

Conclusion:

Clinical presentation with cSDH is a sentinel event for patients aged 90 years or older; 67% have surgical intervention. Disposition does not vary with surgical or non-surgical treatment. Only 24% of patients of this age group presenting with cSDH return home despite a good admission GCS.

Introduction

Chronic subdural hematoma (cSDH) is a common neurosurgical disease in elderly patients¹ and represents a significant health concern for this population.² The role of cSDH as a sentinel event for overall decline in health in an elderly population has been investigated by Dumont et al.² He found that mortality after diagnosis increased for older patients at presentation. The American population over 65 years of age will double by 2050, and more nonagenarians will present to neurosurgeons for treatment of cSDH. The incidence of cSDH in the elderly is estimated to be 7.4/100 000.³ A more recent study by Karibe et al.⁴ reported an increase in the overall incidence of cSDH to 20.6/100 000/year, in the age group 70–79 years, and 127.1 in the age group over 80 years. This begs the question how elderly patients do with cSDH, especially in the light that health resource consumption for subdural hematoma (SDH) is increasing without clear evidence that management practices are leading to improved outcomes.⁵

The factors determining mortality in cSDH are age, GCS at presentation, and associated illnesses such as cardiac and renal problems.^6–10 Age is an important non-neurological variable that affects patients with cSDH, mostly the elderly. Diagnosis of cSDH must be considered a sentinel event exposing decreased function of important organ systems.¹¹ The phenomenon of excess mortality following apparently successful treatment of underlying conditions has been widely reported for another surgical disease, that of hip fractures. The same is true for cSDH, which may unmask underlying medical conditions and exacerbate them,¹² leading to worse outcomes and increased mortality in the elderly.²

Data on the morbidity and mortality of cSDH in the elderly has been published before.^6–10 However, there are no studies that report the outcome of cSDH in patients aged 90 years and older. The objective of our retrospective study is to evaluate outcomes in this growing population.

Methods

We retrospectively requested a summary of all patients who were presented with the International Classification of Disease code for cSDH (432.1; SDH, non-traumatic). Between December 2005 and December 2011, 274 patients were admitted for cSDH. Of these, 8% (n=21) of patients (Table 1) were aged ⩾90 years. Medical records and head computer tomography (CT) scans were reviewed retrospectively. Patient demographics, GCS at presentation, medical co-morbidities, length of stay, disposition, treatment, and radiographic characteristics were abstracted. Descriptive statistical analyses were done using the t-test, Fisher’s exact test, and logistic regression.

Table 1.

Clinical characteristics of 21 patients with chronic subdural hematoma (cSDH) according to age

Gender
Male n (%)	18 (86)
Female n (%)	3 (14)
Mean age years±SD	92±2.5
Median GCS	14
Maximal thickness of SDH (mm)	18.6±7.8
Hematoma location
Left n (%)	13 (62)
Right n (%)	5 (24)
Bilateral n (%)	3 (14)
Anticoagulant therapy n (%)	1 (4.7)
ASA n (%)	3 (14)

Results

Twenty-one patients aged 90 years or older with 24 admissions for cSDH were identified. Sixteen patients underwent surgery for cSDH in 18 surgical procedures. Overall, eight craniotomies and 10 burr hole craniostomies were performed. The mean age was 92 years (SD=2.5), the mean maximal thickness was 18.6 mm (SD=7.8), and the median admission GCS was 14. At the time of diagnosis of cSDH, anticoagulants or antiaggregants were used in 19% of patients (n=4). The size of the subdural was significantly higher in the surgery group with a mean of 20 mm±7.8, compared to 13 mm±5.2 in the conservative group (P=0.02). Overall mortality at the time of discharge was 24% (n=5/21). Mortality was not significantly higher in the surgical group (P=0.28).

Outcome was measured by disposition and Glasgow outcome score (GOS) and disposition at the time of discharge. The median GOS was 2. The discharge disposition was tracked for all patients (n=21). Overall, 38% (n=8) of patients were discharged to a skilled nursing facility (SNF); 14% (n=3) went to inpatient rehabilitation; and 24% (n=5) went home. Twenty-four per cent (n=5) died or went to hospice (Fig. 1).

Figure 1. Overall distribution of disposition.

Surgical intervention was performed in 76% of patients (n=16) with a total of 18 surgeries. Nearly 31% (n=5/16) in the surgical group were sent to hospice or died. None of the patients in the conservative group died, and none of them showed deterioration on the GOS at the time of discharge when compared to the GCS at admission (Fig 2). Overall treatment strategy did not significantly alter the outcome (P=0.10).

Figure 2. Distribution of disposition by treatment was not significantly different.

The surgical group had eight patients who underwent craniotomy and drainage and eight who underwent burr hole and drainage. Fifty per cent (n=4/8) of the craniotomy group were transferred to hospice or died, and 13% (n=1/8) of the burr hole group died. There was no statistical significant difference in the outcome (P=0.11). However the study was not designed to determine optimal surgical management.

Discussion

With the growing elderly population, more nonagenarians with cSDH are seen in a neurosurgery practice. Our study reports on the outcome associated with cSDH in this population. We choose to focus on this age group because we found the cSDH data for the very elderly population to be scarce. We analysed our institutional data focusing on disposition and mortality associated to cSDH in patients older than 90 years. A study by Miranda et al.¹² was a retrospective study comparing long-term survival after diagnosis of cSDH in elderly patients. This study included estimates of survival for 209 patients, with a mean age of 80.6 years, treated conservatively or surgically for cSDH. The authors described an inpatient mortality rate of 17%, which is similar to our present study (24%). The authors also observed that the long-term outcome of elderly patients with cSDH showed persistent excess mortality up to 1 year beyond diagnosis. Similar findings were reported by Dumont et al.² where the authors compared mortality rates of patients with cSDH to standardized mortality ratios based on the patient’s age. They concluded that cSDH is a sentinel health event for the elderly population. We also saw this in our data. Although the median GCS was 14, 76% (n=16) of patients did not return home. This shows that even with a minor neurological impairment, cSDH may unmask underlying medical conditions and exacerbate them.¹²

Previous studies have reported on outcomes measured by GOS. For example, Ramachandran and Hegde⁶ reported on 647 patients with cSDH of whom about one-third were 60 years of age or older. A favorable outcome was seen in two-thirds of patients older than 60 years. Gelabert-Gonzalez et al.¹³ reported a favorable outcome also measured by GOS in 86.1% of patients older than 70 years at hospital discharge. We chose disposition as the outcome, as done by others,¹² for our study because disposition is a clinically relevant factor important for the family and relatives of patients with cSDH to make decisions regarding treatment and end of life.¹⁴ Our society’s definition of ‘old age’ is shifting as the life expectancy increases. This may lead to a wrong understanding on what the recovery potential of elderly with a cSDH is. Especially as the growing aging population includes many elderly individuals who are functional and independent when they are diagnosed with a cSDH.

In our study, the in-hospital mortality rate was 24% (n=5) and 38% (n=8) of patients were transferred to a SNF, which is a negative predictor for the outcome, by itself according to Dumont et al.² He reported that median survival was shortest for patients discharged to a nursing home (1.5 years ±0.7), followed by acute rehabilitation (3.5 years ±0.7) and discharge to home (6.7 years ±1.4). With only 24% (n=5) being discharged to home the odds are not in favor for the patient aged 90 years and older with a cSDH to return to an independent life. This is the first time disposition data on this age group are reported and this data need to be shared with the family and the patient when discussing the outcome.

We report a hospital mortality rate of 24%, but there was no statistically significant difference in the mortality between the surgical and non-surgical treatment groups indicating that surgery does not aid survival. Other studies also report a high mortality in non-surgical groups. Asghar et al.¹⁵ reported 17% mortality in the surgical group and 44% mortality in the non-surgical group. In his study, most deaths were due either directly to the cSDH or to the overall frailty and poor mobility.¹⁵ A study in Taiwan¹⁶ compared post-operative complication in patients <40 and >75 years of age and found that the surgical complication rate was not statistically different between the two age groups. Miranda et al.¹² also found that patients aged 85 years and older had a lower Karnofsky performance status at discharge compared to that before suffering a cSDH, which was not true for the age group between 65 and 84 years. This emphasizes the theory that exacerbation of underlying medical conditions contributes to poor long-term outcomes in the very elderly.

In another study, Fukui¹⁷ reported 89% of patients with cSDH improved at the time of discharge, suggesting that surgical procedure is the best choice for cSDH even in extremely aged patients. However, the long-term outcome was not assessed in this study and was only measured by GOS and not disposition, reflecting the overall outcome to be more meaningful for the families and the patients.¹⁷ Other studies present data indicating better outcomes with surgery in elderly patients, but none present a group older than 90 years.^10,15–21

Surgical management of cSDH in our study consisted of burr hole drainage, percutaneous drainage, or craniotomy with surgeon preference playing a major role in treatment. However, outcomes reported are mixed^10,15–21 and do not focus on the age group of 90 years and older. We did not show a statistical significant difference in the outcome (P=0.11) between these two treatment options. However the study was not designed to determine optimal surgical management.

The retrospective nature of this study places several inherent limitations on the scope of our conclusions. Patients were not prospectively followed, and we concede some patients may have been lost to follow-up. Patients were not randomly assigned to surgical procedure or conservative management, and effects of surgery may represent selection bias because the size of the subdural was larger in patients who were offered surgical treatment (P=0.02).

Despite these limitations, we believe this study represents a valid description of the recovery potential, in-hospital mortality, and disposition faced particularly by patients aged 90 years and older regarding the outcome following diagnosis of cSDH.

Conclusions

With the aging population our society’s definition of ‘old age’ is shifting as the life expectancy increases, and an increasing number of nonagenarians with cSDH will present to neurosurgeons for treatment. Our data will help in the discussion with families about the expected outcome. We conclude that the outcome is poor independent of treatment strategies, i.e., surgical or observation. Presentation with cSDH in this age group must be considered a serious health event given the fact that only 24% of patients, despite a good presentation GCS, return home.