Mortality, functional outcome and quality of life after aSDH surgery in the elderly
Sixty-seven percent of the patients aged 80 years or above surgically treated for traumatic aSDH in our study survived the in-hospital stay and the mortality rate for octa- and nonagenarians thus was only 33%. In contrast to our findings, mortality rates of patients aged 60 or above reported by Jennet et al. in 1976 [16
] or by Howard et al. in 1989 [20
] were substantially higher, with up to 88% and 74%, respectively. Moreover, the mortality rate of younger aSDH patients (18–40 years) reported by Howard et al. was significantly lower with 18% compared to the older aSDH patients (≥ 65 years) [20
]. Since then, increasing age is considered to be an independent predictor for mortality after aSDH, which has also been confirmed in more recent publications [10
]. While only 11% of the younger aSDH patients died in direct consequence of the hemorrhage in our study, the difference to the oldest patients did not reach statistical significance. Our results could, therefore, be interpreted as a trend towards decreased early mortality of elderly aSDH patients, similar to recent findings by Won et al. who reported a mortality rate of octogenarians of only 28%, [8
] These improvements in early survival of older patients after traumatic aSDH might reflect the advancements of surgical treatments and ICU care in the past decades.
Unfortunately, mortality after aSDH is not restricted to the in-hospital stay but rather continuing during the months after the injury. In a study from the 1990s, mortality during a 6 month follow-up period after aSDH was reported to be as high as 50% [19
]. In our study with a long follow-up period, a total of 15 additional aSDH patients had died (62.5%) 30 (21–40) months after the injury. Of those, only four were of the younger age group (27%) and 11 (73%) had been ≥ 80 years old when the aSDH had occurred. Higher mortality rates during follow-up in patients aged ≥ 80 years could be regarded as a direct consequence of the advanced age, often accompanied by an increasing number of comorbidities, a higher incidence of medical complications or a greater usage of antithrombotic agents as seen in our study. In addition, decreasing numbers of intact neurons and greater exposure to repetitive insults over a lifetime [29
] could be responsible for impaired repair capacity of brain damage in the elderly, leading to a stepwise increase in probability of poor outcome with increasing age [2
In previous studies, functional recovery of aSDH patients aged ≥ 65 years has been reported in only 9% of cases, [20
] questioning the benefit of surgical evacuation of aSDH in the elderly. Non-age specific favorable outcomes were, however, seen in 19–32%, [10
] increasing up to 91% in patients with admission GCS scores of 13–15 [17
]. In our current study, we found a favorable outcome at discharge in 22% of the oldest aSDH patients, suggesting improved functional recovery in octo- and nonagenarians compared to older reports in the literature [8
]. More importantly, outcome at discharge was not significantly different between the young and the oldest aSDH patients. This might in part be related to the fact, that aSDH patients in our study were mostly treated on our highly specialized neuro-intensive care unit with substantial experience in handling more severe and complex neurosurgical cases.
Although the rate of successful prospective follow-up achieved in our study was low (17%), it revealed exemplary cases of long-term recovery after aSDH surgery in the elderly: two patients aged ≥ 80 years who had both been discharged from the hospital in a severely disabled state (GOS 3) had GCS scores of 13 and 14 and showed good recovery (GOSE 6 and 8) at follow-up. On the other hand, quality of life measured by the QOLIBRI score was low or impaired (< 65%) in all aSDH patients who were available for follow-up in our study, including cases with good functional recovery (GOSE 7 and 8), although a significant correlation between QOLIBRI and GOSE scores for patients with traumatic brain injury has been reported in the literature [26
]. It, therefore, seems necessary to evaluate health-related quality of life with corresponding measurements such as the QOLIBRI separately from functional recovery to better understand the overall outcome after surgical treatment in aSDH patients.
Outcome prediction after aSDH surgery in octo- and nonagenarians
Although comorbidities such as arterial hypertension or atrial fibrillation, correlating with a frequent usage of antithrombotic medication, were significantly more common in the oldest than in the younger aSDH patients, they did not individually predict an unfavorable outcome at discharge in our study. We assume that pre-surgical reversal of antithrombotic medication, interdisciplinary treatment on a specialized ICU and standardized treatment protocols may have reduced the risk for complications and (re)bleedings, and thus might play important roles in the improvement of survival of elderly aSDH patients [8
]. However, the presence of fewer comorbidities was associated with a favorable outcome at discharge in patients aged 80 years or above whereas preexistence of ≥ 5 comorbidities was linked to an unfavorable outcome in many patients in both age groups, suggesting that the sum of comorbidities and the resulting general health status are indeed able to influence outcome at discharge after surgical aSDH treatment.
The neurological status after injury, mostly reported as the GCS score at admission, has frequently been associated with functional outcome after aSDH surgery [3
]. Correspondingly, we were able to confirm a comatose status at admission, defined as a GCS score of 3–8, as a significant predictor for an unfavorable outcome at discharge in the young, but more importantly also in the oldest aSDH patients in our study. This is in accordance with findings of Won et al., who also confirmed the predictive value of the GCS for functional outcome at discharge in octo-and nonagenarians [8
Interestingly, anisocoria and radiological signs of herniation were only found to be significant predictors for an unfavorable outcome at discharge in the group of younger aSDH patients in our study. In the elderly patients, less severe brain injury without clinical or radiological signs of herniation might have been sufficient to worsen results of the relatively broad outcome score GOS and might therefore have affected this finding. Nevertheless, pupillary abnormalities should be interpreted as signs of severe injury in all aSDH patients [12
]. Similarly, a midline shift > 1 cm was only associated with an unfavorable outcome at discharge in the group of younger aSDH patients, suggesting that effects of brain atrophy and wider subdural spaces in patients aged ≥ 80 years might have allowed for more midline shift without having the same impact on outcome.
Value of volumetric aSDH analysis in outcome prognostication
The association between aSDH volume and outcome was already analyzed in the 1980s and 1990s, [20
] with volumes being manually calculated. Favorable outcomes were seen in patients with mean hematoma volumes of 31 ml [23
] (< 100 ml [24
]) and poor recovery was associated with mean hematoma volumes of 104 ml [23
] (> 100 ml [24
]). Additionally, Howard et al. [20
] described significant differences in mean volumes between patients aged over 65 years (mean 96.2 ± 11.72 ml) and patients aged 18–40 years (21.6 ± 27.7 ml). Volumetric measurements of aSDH were already conducted in 2009, [31
] using computer assisted analysis and volumes < 50 ml were associated with higher rates of functional recovery (50%) in comparison to larger bleedings (> 50 ml, 34% functional recovery).
A correlation between larger hematoma volume and worse outcome was also seen in our patient cohort. Patients aged ≥ 80 years who had an unfavorable outcome showed almost twice the size of aSDH volumes than patients with a favorable outcome. This finding was also applicable for the younger aSDH cohort (p = 0.02).
Interestingly, the mean aSDH volume was greater in the elderly patient population compared to the younger patients without yielding statistical significance. Nevertheless, this finding might reflect the higher use of antithrombotic drugs in patients aged 80 years and above, making them prone to more extensive traumatic bleedings.
Our current results underline that while a chance for a good functional outcome exists and while the risk of mortality is reduced, surgical evacuation of traumatic aSDHs in octo- and nonagenarians is still leading to high rates of poor outcome and reduced quality of life – circumstances, under which many older patients might not want to further live their lives. The availability of a patient decree outlining the individual treatment choices and wishes should, therefore, always be inquired and in applicable cases, the decision to perform surgery should only be taken in consideration of the patients written will. In light of frequent preexisting conditions such as care dependency or dementia in the elderly population, the suspected will of the patient has to be respected even when no patient decree is existing or present and thus, her or his relatives should be consulted if possible.
Several limitations of our study should be acknowledged. The analysis was performed at a single institution in a retrospective manner over a long period of time (11 years), limiting external validity and posing the risk of changes in the management of aSDH patients. However, comparison of clinical parameters between the first and the second half of the study period revealed no relevant significant differences over time. Nevertheless, the available patient cohort of aSDH patients aged ≥ 80 years was small and the univariate statistical analysis therefore potentially underpowered. In addition, due to the limited number of eligible aSDH patients overall, mathematical matching of the oldest with the younger patients was not feasible and findings in octo- and nonagenerians were compared with randomly selected younger patients instead. Findings of the comparison between both groups have, therefore, to be interpreted with caution. Due to the small sample size, multivariate analysis was also not feasible, making it impossible to rule out that some variables are dependable in our study. Furthermore, we excluded patients in whom the decision to withhold surgical treatment was taken, hereby potentially biasing our results towards better outcomes. Additionally, an even smaller sample size of patients was available for follow-up analysis which limits the general applicability of our results on health-related quality of life. We also did not differentiate between patients undergoing rehabilitation, biasing outcome at follow-up. Broader prospective studies on long-term outcome of elderly aSDH patients with larger cohorts or national as well as international trauma registries are therefore needed to better understand risks and chances of surgical treatment in this patient subgroup.