Impact of Age On Short-Term Outcomes After Discharge in Patients With Aneurysmal Subarachnoid Hemorrhage

Changes in outcome after discharge of patients with aneurysmal subarachnoid hemorrhage (aSAH) have not been examined in detail in recent aging society. This study aimed to clarify the prevalence and factors of changes in outcome in these patients based on a registry database in Japan. Overall, 1,111 radically treated patients with aSAH in the acute stage were selected between January 2010 and December 2012 for inclusion in the modied World Federation of Neurosurgical Societies (mWFNS) scale study. The modied Rankin Scale (mRS) score at discharge and 3 months after onset, prevalence of improvement and decline between these two periods, and their factors were evaluated by age groups through multivariate logistic regression analysis. The rates of improvement and decline were 28.4% and 2.4%, respectively. Factors that improved the mRS score were WFNS and mWFNS scale grade (cid:0) versus (vs.) (cid:0) (odds ratio: 0.47, 95% condence interval: 0.31–0.71), grade (cid:0) vs. (cid:0) (0.49, 0.31–0.76), and cerebral vasospasm category 4 vs. 1 (0.43, 0.25–0.77). Factors that decreased the mRS score were pre-elderly (65–74 years) vs. non-elderly (<65 years) (5.65, 1.09–29.20), elderly ( ≥ 75 years) vs. non-elderly (19.20, 2.36–156.50), WFNS scale grade III vs. I, and aneurysmal location. Mild neurological grade on admission and no cerebral infarction due to cerebral vasospasm were the factors improving outcomes, and aging was a factor declining outcomes between discharge and at 3 months after onset. Careful assessment timing of outcome in elderly patients is needed in the future aging society.


Introduction
Aneurysmal subarachnoid hemorrhage (aSAH) is a severe type of hemorrhagic stroke with a mortality rate up to 35%.
[28] Approximately one-third of the survivors remain severely disabled and functionally dependent. [19] Although there are many reports examining the outcomes several months after onset of aSAH, [3, 4, 6, 8, 10-12, 15, 20, 21, 29] few reports have discussed the changes in the prevalence of outcome at discharge and after discharge in recent aging situation. [2,5,18,24,25,30] There are only two reports, to date, of changes in the outcome of individual patients after discharge. [5,30] According to these reports, approximately 60% of patients with aSAH had improved modi ed Rankin Scale (mRS) scores at 6 months after onset, and only one report mentioned a decrease in the mRS score after discharge. [30] Since the number of elderly patients with aSAH has been increasing recently, the number of cases in which the mRS score decreases after discharge may also increase. Therefore, to examine not only increases but also decreases in the mRS score after discharge would be desired in future aging population. In addition, short-term changes in post-discharge outcomes of elderly patients with aSAH should be examined.
Although the World Federation of Neurosurgical Societies (WFNS) scale has been used to classify the severity of aSAH, the modi ed WFNS (mWFNS) scale has been proposed to more accurately assess the clinical status of patients, in which patients with a Glasgow Coma Scale (GCS) score of 14 are assigned to grade II and those with a GCS score of 13 are assigned to grade III, regardless of the presence of neurologic de cits. [22] As a result, the mWFNS scale presumably enables more accurate prognostic assessment of patients with aSAH than the original scale. Because the changes in outcomes after discharge in relation to the WFNS and mWFNS scales have never been examined, we tried to investigate these relationships.
This study aimed to elucidate the prevalence and factors of changes in the outcome of patients with aSAH by age groups based on the registry database of the mWFNS scale study.

Data sources
Data for the present study were obtained from the 1,864 participants enrolled in the mWFNS scale study. This was a multicenter, prospective, observational study that included 38 neurosurgical institutions across Japan. Patients were enrolled from January 2010 to December 2012. The mWFNS scale study prospectively collected the following relevant clinical data: age, sex, location of the ruptured aneurysm, each component of the GCS score, presence/type of associated neurologic de cits (weakness and/or aphasia), premorbid mRS score, medical complications, type of treatment used for aneurysm obliteration, type of perioperative complications, type of Fisher computed tomography (CT) group, and mRS scores at discharge and 3 months after onset. The details of data collection and management have been published elsewhere. [17,22] All patients were aged ≥ 15 years, and the interval between onset and admission was ≤ 72 hours in this cohort. Age was classi ed as follows: non-elderly (NE), < 65 years; pre-elderly (PE), 65-74 years; and elderly (E), ≥ 75 years; according to the classi cation of the World Health Organization and the Japan Geriatrics Society.
[16] The locations of the ruptured cerebral aneurysms were categorized as follows: anterior cerebral artery, anterior communicating artery, internal carotid artery-posterior communicating artery (ICPC), middle cerebral artery, other internal carotid artery (ICA), vertebral artery and basilar artery, and others. Other ICA was de ned as any aneurysm that occurred in the ICA except for the ICPC. "Others" was de ned as the aneurysm with an unclassi ed location. The cerebral vasospasm (VS) category due to aSAH was classi ed as follows: 1, none; 2, slight and reversible decline without additional treatment; 3, de nitive but reversible decline with additional treatment; and 4, permanent de cit and/or appearance of a low-density area on CT. [23] In the current study, cerebral VS category 4 was de ned as severe cerebral VS. Patients' outcomes were examined using the mRS score at discharge and 3 months after onset.
Changes in the mRS score between discharge and at 3 months after onset were classi ed as follows: improvement, no change, and decline. Improvement was de ned as a decrease in the mRS score, whereas decline was de ned as an increase in the mRS score at 3 months after onset compared to that at admission.
In the improvement group, patients with a mRS score of 0 at discharge were excluded. Multivariate logistic regression analysis was performed with the dichotomized improvement values of the mRS score from discharge versus (vs.) the follow-up at 3 months (1 = improvement; 0 = no improvement) as the dependent variable. The dichotomized decline values of the mRS score were 1 = decline and 0 = no decline.

Study population
We excluded 99 patients with a premorbid mRS score ≥ 3; 374 whose mRS score was 6 at discharge; 81 who received treatment other than clipping or coiling; scale grade or mWFNS scale grade at admission (n = 26), mRS score at discharge and 3 months after onset (n = 31), type of treatment (n = 18), location of the aneurysm (n = 21), pre-onset mRS score (n = 15), Fisher CT grade (n = 14), and cerebral VS category (n = 8). The remaining 1,111 patients were analyzed retrospectively ( Fig. 1).

Statistical analyses
All statistical analyses were performed using JMP® Pro version 15 (SAS Institute, Cary, NC). Values are presented as the mean (standard deviation). Categorical variables were compared using the Fisher exact probability test. Continuous variables with normal distributions were analyzed using the Student ttest, and those with non-normal distributions were analyzed using the Mann-Whitney U-test. P-values < 0.05 were considered statistically signi cant. The odds ratio (OR) and 95% con dence interval (CI) of mRS scores that changed between discharge and at 3 months after onset were evaluated through a multivariate logistic regression analysis. Variable selection in the multivariate analysis was determined according to clinical importance and existing literature, and no variable selection method, such as univariate or stepwise selection, was applied. Sex, age group, treatment type, location of the aneurysms, WFNS scale grade, mWFNS scale grade, and cerebral VS category were selected as variables.

Review of the literature
The literature search on the prevalence of changes in outcomes after discharge in patients with aSAH included articles published between January 2000 and January 2021. The search was performed on PubMed with publications restricted to those written in the English language, and the last search date was March 31, 2021. Keywords and free text searches used combinations of the following keywords: aneurysmal subarachnoid hemorrhage, change in outcome, and outcome at discharge. All reference sections of eligible studies and relevant reviews were reviewed further for potential studies. If a study generated multiple publications, the most current report was considered for the analysis.

Results
Patients' average (standard deviation) and median ages were 52.0 (9.3) and 54 years in the non-elderly group, 68.9 (2.8) and 68 years in the pre-elderly group, and 80.1 (3.9) and 80 years in the elderly group, respectively. Table 1 shows the baseline characteristics and outcomes in each group. While several factorsage, sex, location of the aneurysm, WFNS scale grade, mWFNS scale grade, cerebral VS category, mRS score at discharge, mRS score 3 months after onset, and change in the mRS score-were signi cantly different between the groups, the Fisher CT group was not. Of the 1,111 patients, 316 (28.4%) had improvement, 768 (69.1%) had no change, and 27 (2.4%) had decline in the mRS scores. The results of multivariate logistic regression analysis of the group with improved mRS scores between discharge and at 3 months after onset are presented in Table 2 Table 3 and Fig. 3   Among the 98 papers, we selected satisfactory six papers for review. The prevalence of changes in outcome after discharge in patients with aSAH is presented in Table 4. The changes in the prevalence of good outcome (mRS score: 0-2 or Glasgow Outcome Score: 4-5) between discharge and the last follow-up ranged from 8.4-55.2%. [2,5,18,24,25,30] The prevalence of improvement of mRS scores in total cases between discharge and at 6 months ranged from 57.4-61.0%, [5,30] and the prevalence of declined mRS scores during the same period was 17.0%. [30]

Discussion
Prevalence of changes in outcomes after discharge In the current study of 1,111 patients, the prevalence of improvement, no change, and decline was 28.4%, 69.1%, and 2.4%, respectively, of mRS scores between discharge and at 3 months after onset in Table 1. As described in Table 4, Hammer et al. reported that 116 of 202 patients (57.4%) had an improved mRS score between discharge and at 6 months. [5] According to Wilson et al.'s study, 88 cases of poor-grade aSAH were analyzed, and 61.0 % of patients had an improved mRS score between discharge and at 6 months. [30] Compared to these reports, our study shows lower prevalence (28.4%) of improved mRS scores. According to past reports, the average length of hospital stay for patients with aSAH in Japan was approximately twice longer (32-37 days) than that in the United States (16.9-18.4 days). [1,9]This accounts for the fact that Japan has a universal health insurance system and provides treatment for cerebral vasospasm and ventriculoperitoneal shunting for hydrocephalus during hospitalization. [7] In addition, our study evaluated the mRS score at 3 months after discharge, which was a shorter term than those reported in previous studies.
In the present study, as shown in Table 3, only 2.4% of patients had deteriorated mRS scores after discharge. As described in Table 4, Wilson et al. reported that 17.0 % of patients had deteriorated mRS scores between discharge and at 6 months after onset. [30] Same reasons of longer hospital stay in Japan and short-term outcome can account for this lower rate of decline. Since results of these studies show that the prevalence of improvement in outcome at several months after discharge is higher than that of decline in outcome, the outcomes after aSAH are likely to improve in the several months after discharge.
The factors for improvement of mRS scores after discharge In this study, WFNS and mWFNS scale grades IV and V were inverse risks compared with grade I, for the improvement of mRS scores between discharge and at 3 months after onset. Namely, mild neurological grade on admission was the factors improving outcomes. Hammer  In our study, cerebral VS category 4 was also an inverse risk compared with category 1 for the improvement of mRS scores between discharge and at 3 months after onset. The absence of a stroke played an important role in the prediction of early outcome improvement in the univariate analysis in another study. [5] Therefore, it is considered that the absence of cerebral infarction due to cerebral vasospasm after sSAH is involved in improving the outcome after discharge.
The factors for decline of the mRS score after discharge Although, there are no reports on the factors for decline of the mRS score after discharge in the past, several factors for declining mRS scores were con rmed in this study.
The risk factors regarding the age for the decline of mRS were PE and E in the multivariate analysis. This fact means that aging is a factor for declining outcomes even in the short-term after discharge. In patients with aSAH, older age may worsen the short-term outcome after discharge; therefore, the assessment timing of outcome for aSAH should be carefully determined in the future aging society.
Regarding the aneurysm location, cases of other ICA location had signi cantly decreased mRS scores between discharge and at 3 months after onset compared to cases of ICPC in the current study. Why other ICA locations caused patients' outcomes to signi cantly decline after discharge compared to the ICPC location may be due to the small number of samples in the decline group.
Regarding the mWFNS scale grade WFNS scale grade III vs. I was the signi cant factor for decreased mRS scores after discharge. WFNS scale grade III is a special entity with focal neurological de cit (GCS score of 13 or 14) that included only 4.0% of our cohort. Previous reports have shown that the proportion of WFNS scale grade III is only 2%-9%, [13,14,26,27,31] which is as infrequent as ours. Given the low proportion of WFNS scale grade III, WFNS scale grade III may be underestimated. On the other hand, the proportion of mWFNS scale grade III is 8.5% as shown in Table 1, which is higher than that of the WFNS scale. As previously reported, the mWFNS scale presumably may enable more accurate prognostic assessment of patients with aSAH compared to the original WFNS scale. [22] Prevalence and changes in good outcome after discharge according to the literature review As shown in Table 4, changes in the prevalence of good outcome after discharge were 8.4-55.2% according to the literature review. [2,5,18,24,25,30] In contrast, that of our study was lower at 6.9%. The reasons for this are thought to be due to the shorter evaluation period after discharge and the higher proportion of good outcomes at discharge in our study compared to those in the previous reports.

Limitations
The present study has several limitations. First, this was not a randomized controlled study, and the initial purpose of the mWFNS scale study was to evaluate the ability of the mWFNS scale to predict outcomes. These data were derived from a post-hoc analysis, and the original study was not designed to account for various factors. Second, because of the lack of independent physicians, the GCS and mRS scores were evaluated by neurosurgeons in each institute who were not blinded to the treatment and these ratings were not standardized; therefore, these values may be potentially biased. Third, in this database, the follow-up period was limited to 3 months after onset. We could not assess long-term statuses of the patients. If possible, a longer observation study may be necessary in the future. Fourth, the current results may not be generalized to other countries with different medical resources and systems. Therefore, the results of this study should be interpreted with caution. These data included only a limited set of aSAH cases in Japan, which has the largest elderly population in the world.
Fifth, in the present study, the e cacy of the mWFNS scale in assessing the prognosis of the patients with aSAH was not compared to that of the original WFNS scale. To determine the e cacy of the mWFNS scale for outcome evaluation in patients with aSAH, more precise evaluation between GCS scores 13 and 14 would be necessary based on various factors and should be re ned in the future. Further randomized controlled trials with diverse participants are also warranted.

Conclusions
Mild neurological grade on admission and absence of cerebral infarction due to cerebral VS category were the factors improving outcomes, and aging was a factor declining outcomes between discharge and at 3 months after onset. Careful assessment timing of outcome is necessary for aSAH in the future aging society. Compliance with ethical standards Con ict of interest The authors report no con ict of interest concerning the materials or methods used in this study or the ndings speci ed in this paper.