International Journal of Surgery 48 (2017) 266–274

Contents lists available at ScienceDirect

International Journal of Surgery

journal homepage: www.elsevier.com/locate/ijsu

Review

Timing of surgery for aneurysmal subarachnoid hemorrhage: A systematic T

review and meta-analysis
Zhong Yaoa,b, Xin Hua,b, Lu Maa,b, Chao Youa,b, Min Hea,b,∗
a
Department of Neurosurgery, West China Hospital, Sichuan University, China
b
West China Brain Research Centre, West China Hospital, Sichuan University, China

A R T I C L E I N F O A B S T R A C T

Keywords: Introduction: The timing of surgery for aneurysmal subarachnoid hemorrhage influences the outcome, but the
Timing optimal timing remains controversial. We conducted a systematic review to clarify whether early surgery was
Surgery better than late surgery for improving outcome.
Aneurysm Materials and methods: We systematically searched several databases to screen eligible studies. After synthesizing
Subarachnoid hemorrhage
data, an overall effect was shown using a risk ratio (RR) and 95% confidence interval (CI). Subgroup analyses
Meta-analysis
were stratified by multiple variables to control the confounding factors. Sensitivity analyses were applied to
check the robustness of the results. Publication bias was measured with Egger's and Begg's tests.
Results: A total of 14 studies were included in the analysis. Compared with late surgery, early surgery sig-
nificantly decreased the incidence of poor outcome, regardless of whether patients were in good condition (RR,
0.65 [95%CI, 0.50 0.84]; p = 0.001) or in poor condition on admission (RR, 0.71 [95%CI, 0.61 0.83];
p < 0.0001). Moreover, when patients were in good condition on admission, early surgery also effectively
reduced the death rate (RR, 0.61 [95%CI, 0.46 0.82]; p = 0.001). Additionally, early surgery reduced the death
rate compared with late surgery in patients older than 50 years (RR, 0.49 [95%CI, 0.27 0.89]; p < 0.002).
Conclusions: Early surgery was superior to late surgery in reducing a poor outcome and death rate when patients
were in good condition on admission, and decreased the incidence of poor outcome when patients were in poor
condition on admission. Age was a potential confounding factor, influencing the effect of early surgery. Further
study is required on this issue.

1. Introduction intermediate surgical treatment improved outcomes compared with

delayed surgery, although the result was derived from an indirect
Aneurysmal subarachnoid hemorrhage (aSAH) is a devastating comparison [3].
cerebrovascular disease, not only due to the severe effect of initial he- It was reported that the high mortality and morbidity in the early
morrhage but also owing to the complicated treatment schemes. surgery group was attributed to brain swelling and patients' poor tol-
Surgical and endovascular treatments have become the mainstay ap- erance to the surgery. As microsurgery has developed and neurocritical
proaches for occluding the aneurysm. Although detachable coil embo- care has helped patients better tolerate operations, it was unclear
lization is widely applied, many patients still undergo open craniotomy whether advanced microsurgery and basic treatment could offset the
because not all ruptured aneurysms are amenable to endovascular side effects of early surgery and achieve better outcomes than late
treatment. The timing of surgical treatment for a ruptured aneurysm is surgery. We conducted a systematic review to elucidate whether or not
under debate and has experienced a shift from favoring delayed treat- early surgery improves the overall outcomes of patients with aSAH.
ment in the 1970s to preferring early treatment in recent years. A na-
tionwide study found that early surgery resulted in unfavorable out- 2. Materials and methods
comes [1]. In 1990, the International Cooperative Study demonstrated
that early surgery was neither more hazardous nor more beneficial than 2.1. Search strategy
late surgery. Later, a retrospective study showed that early surgery
could benefit patients in good status on admission [2]. Reviewing re- We conducted a comprehensive search through PubMed, Ovid and
levant publications from 1974 to 1998, De Gans et al. found early and Web of Science from their inceptions until August 2017. Relevant

∗
Corresponding author. No. 37 Guo Xue Xiang, Chengdu, 610041, China.
E-mail address: heminhx@aliyun.com (M. He).

https://doi.org/10.1016/j.ijsu.2017.11.033
Received 14 September 2017; Accepted 15 November 2017
1743-9191/ © 2017 IJS Publishing Group Ltd. Published by Elsevier Ltd. All rights reserved.
Z. Yao et al. International Journal of Surgery 48 (2017) 266–274

Fig. 1. Flow chart of literature search.

publications in English were screened. In the search process, medical 2.3. Definitions of outcomes
subject headings and keywords were used in combination as follows:
“subarachnoid hemorrhage,” “subarachnoid hemorrhage,” “intracranial Good clinical condition on admission or before surgery was defined
aneurysm,” “craniotomy,” “surgical,” “operative,” “microsurgical,” as I-III class in Hunt and Hess classification (HH), 1–3 score in the
“time factors,” “early,” “delay,” “timing,” “24 h,” “48 h” and “72 h”The World Federation of Neurological Surgeons (WFNS) scale, 13–15 score
reference lists of the extracted publications were manually searched. in the Glasgow Coma Scale (GCS) and the corresponding clinical
We also attempted to contact the authors of primary studies to obtain manifestations described in the studies. The poor clinical condition was
further data. This review was conducted in accordance with the opposite the good clinical condition. Early surgery was defined as
Preferred Reporting Items for Systematic Reviews and Meta-Analyses carrying out the operation within 3 days of aSAH onset. Late surgery
[4]. was performed 7–10 days after the ictus. Intermediate surgery between
these two phases has been proven to lead to an unfavorable outcome in
previous reports, as the intermediate time phase corresponds to the
2.2. Including criteria peak period of cerebral vasospasm [5–8]. The definition of poor out-
come was a 1–3 score in the Glasgow Outcome Scale (GOS), 3–6 in the
We included studies meeting the following criteria: (1) research on modified Rankin Scale (mRS) and their corresponding clinical mani-
patients with the diagnosis of ruptured intracranial aneurysms; (2) festations. To take the greatest advantage of the eligible data, char-
microsurgery as the main treatment for ruptured aneurysms; (3) pa- acteristic in more than 95% of patients represented the characteristic of
tients who were assigned to different therapeutic time windows; (4) the whole population (i.e., if 95% of participants were in good clinical
clinical conditions were recorded on admission or before surgery; (5) condition, we considered the whole population to be in good clinical
overall management outcomes described rather than only the surgical condition).
outcomes; (6) outcomes were reported separately by different clinical
conditions and different therapeutic time windows. Studies were ex-
cluded for reasons as follows: (1) therapeutic time windows were de- 2.4. Data extraction
scribed with items of early and late, rather than exact measurement
units of hours or days; (2) reported both surgical and endovascular Data extraction forms were used to collect pertinent information,
treatment, from which we could not extract surgical treatment sepa- including the names of authors, years of publication, study types,
rately. duration of studies, female ratio, ratio of anterior circulation aneurysm,
Two researchers independently screened eligible publications, and clinical condition measurement, outcome measurement, numbers of
differences were resolved by discussion and consensus. When different events of interest and follow-up time.
studies contained overlapping participants partially or totally, we only
chose the study with a larger sample size and more detailed informa-
tion.

267
Z. Yao et al. International Journal of Surgery 48 (2017) 266–274

2.5. Quality assessment

NOS

6
8
7
9

8
6

7
8
7
7
7
6

6
Every included study underwent quality assessment with the
Follow-up time

at discharge

at discharge

at discharge

at discharge
Newcastle-Ottawa Scale (NOS), which evaluates studies in the three
12 months
44 months

11 months
6 months

3 months
3 months
6 months

3 months
3 months

6 months
categories of selection, comparability and outcome. Low quality,
moderate quality and high quality correspond to scores of 0–3, 4–6, and
7–9, respectively.

independent, dependent, death

good, recovery, disable, dead

excellent, good, fair, poor, 2.6. Statistical analyses
Rankin Disability Score
Outcome measurement

good, fair, poor, death

good, fair, poor, death

Data were synthesized with a random-effects model (Mantel-
Haenszel statistics). Overall effect was shown with a risk ratio (RR) and

survive, death
its 95% confidence interval (CI). A two-tailed p < 0.05 was considered
statistically significant. Heterogeneity was measured with the Cochrane
death

GCS, Glasgow Coma Scale; GOS, Glasgow Outcome Scale; mRS, modified Rankin Scale; NA, not available; NOS, Newcastle-Ottawa Scale; WFNS, World Federation of Neurosurgical societies.
Q test and I2 test with a threshold of p < 0.10 or I2 > 50% indicating
mRS
GOS

GOS

GOS

GOS

GOS

GOS

substantial heterogeneity. Sensitivity analyses, by excluding a single

study each time and calculating the overall effect of the remaining
before surgery

before surgery

before surgery

before surgery

studies, were applied to check the robustness of the results. Subgroup

on admission
on admission
on admission

on admission

on admission

on admission
on admission
on admission

on admission

on admission
analyses were stratified by study type, country, year of publication, age,
condition
Clinical

clinical condition before surgery, clinical condition on admission and

follow-up time. Publication bias was assessed with Begg's and Egger's
tests, and a p < 0.05 showed pronounced bias. All the statistical tests
were implemented with Stata 14.0 software (College Station, Texas).
Clinical condition measurement

3. Results
corresponding clinical

corresponding clinical

corresponding clinical

3.1. Literature search

Hunt and Kosnik

manifestation

manifestation

manifestation
Hunt Hess

Hunt Hess
Hunt Hess

Hunt Hess

Hunt Hess
Hunt Hess

Hunt Hess

A total of 1986 publications were identified after removing all du-

plicates. Of these, 1960 citations were excluded based on irrelevant
WFNS

WFNS

GCS

titles and abstracts, leaving 26 studies for full-text assessment. Several

studies were excluded for the following reasons: outcomes not sepa-
rately reported by clinical conditions [9–11], mixed intermediate sur-
Anterior circulation

gery and late surgery [2,12], or reported endovascular treatment

[13,14]. Finally, 14 eligible studies were included in the quantitative
synthesis. The detailed search process is depicted in Fig. 1.
100%

100%

100%
ratio

95%

95%
94%

88%

80%

92%

3.2. Main characteristics of included studies

NA

NA

NA

NA

NA
Female ratio

One randomized study, one prospective cohort study and 12 retro-

spective studies were extracted. There were 3 multicenter studies and
65%
64%

68%

76%

66%
50%

54%
56%
61%

11 single-center studies. The mean age was 53 years, and one study
NA

NA

NA

NA

NA

[15] exclusively researched the older population. A female pre-

Mean age

dominance was shown in the whole population, and anterior circulation

aneurysms accounted for most cases. Ten studies measured the clinical
NA

NA

NA
55
52
74
49

50
49

52
43

50
54

58

conditions on admission, and the other 4 studies measured the condi-

tions before surgery. As for follow-up time, 4 studies evaluated out-
multinational

multinational

multinational

comes at discharge, and 10 studies recorded a definite follow-up time.

Lithuania
America
Country

Finland

Canada

As for study quality, 9 studies were of high quality and 5 were classified
Japan

Japan

Japan

Japan

Japan
China
Italy

as moderate quality, with a mean score of 7 in the NOS. The detailed

characteristics are summarized in Table 1.
1994–2002
1990–2000
1980–1983

1977–1980
1961–1975

2000–2003

1993–1997
2010–2011

1968–1980

1978–1988
until 1989
Duration

3.3. Synthesis analyses

NA

NA

NA

3.3.1. Overall effect

retrospective
retrospective
retrospective

retrospective
retrospective

retrospective

retrospective
retrospective
retrospective
retrospective

retrospective

retrospective
randomized
prospective
Main characteristics of included studies.

Study type

The pooled effect demonstrated no significant difference in poor

outcome between the early surgery and late surgery groups (RR, 0.86
[95%CI, 0.63 1.18]; p = 0.36), accompanied by substantial hetero-
geneity (I2 = 90.6%, Fig. 2). Sensitivity analyses by excluding any
Dorhout Mees 2012 [26]

Tamasauskas 2000 [21]

single one study did not change the results statistically. Publication bias
Yamamoto 1992 [22]
Kayama 1978 [27]

Mahaney 2011 [7]

was insignificant in the Egger's test (p = 0.96). With respect to mor-

Taneda 1982 [25]
Kassell 1981 [23]

Ohman 1989 [6]

Ferch 2003 [15]

Sano 1994 [19]

Weir 1981 [20]

Haley 1992 [8]
Abe 1992 [16]

Tan 2014 [24]

tality, early surgery showed no significant difference in reducing the

occurrence of death compared with late surgery (RR, 0.90 [95%CI, 0.49
1.63]; p = 0.72). Heterogeneity was obvious (I2 = 92%, Fig. 3). Sen-
Author
Table 1

sitivity analyses verified the robustness of the results. Only a small

publication bias was shown in the Egger's test (p = 0.98).

268
Z. Yao et al. International Journal of Surgery 48 (2017) 266–274

Fig. 2. Forest plot of studies researching on the poor outcome of all the patients.

3.3.2. Subgroup analyses p = 0.03) and death (RR, 3.38 [95%CI, 2.13 5.38]; p < 0.0001)
There were 10 studies that researched the clinical condition on compared with early surgery. In the subgroup analyses of age, early
admission. Because several studies reported both patients in good surgery reduced the death rate compared with late surgery in patients
condition on admission and in poor condition on admission, the same older than 50 years (RR, 0.49 [95%CI, 0.27 0.89]; p < 0.002; Table 2).
study might appear twice in the forest plot of subgroup analyses. These When subgroup analyses were stratified by study type, country, pub-
participants were independent and did not overlap. Compared with late lication year and follow-up time, there were no significant differences
surgery, early surgery significantly decreased the incidence of outcome, between effects of early surgery and late surgery on a poor outcome and
regardless of whether patients were in good condition (RR, 0.65 death.
[95%CI, 0.50 0.84]; p = 0.001) or in poor condition on admission (RR,
0.71 [95%CI, 0.61 0.83]; p < 0.0001; Fig. 4 and Table 2). The pub- 4. Discussion
lication bias in this subgroup was insignificant, with a p = 0.13 in the
Egger's test, although the funnel plot was slightly asymmetric (Fig. 5a). The timing of surgery for aSAH remains controversial, mainly due to
The sensitivity analyses achieved a consistent result (Fig. 6a). When a lack of well-designed studies. Despite recent studies supporting early
patients were in good condition on admission, early surgery sig- surgery, several confounding factors weakened the comparability be-
nificantly reduced death compared with late surgery (RR, 0.61 [95%CI, tween early and late surgery. On one hand, clinicians tended to assign
0.46 0.82]; p = 0.001). However, when patients were in poor condition patients in poor condition to the late surgery group, worrying about
on admission, there was a marginally significant difference in the death patients' intolerance to early surgery [16]. This selection bias would
between early surgery and late surgery (RR, 0.61 [95%CI, 0.37 1.00]; generate a trend towards more poor outcomes in the late surgery group.
p = 0.05; Fig. 7 and Table 2). Publication bias was insignificant, with a On the other hand, classifying patients according to the presurgical
p = 0.39 in the Egger's test and a symmetric funnel plot (Fig. 5b). The clinical condition is worth discussing. In the early surgery group, some
sensitivity analyses achieved a consistent result (Fig. 6b). patients had to undergo operations despite their poor conditions. In the
There were 4 studies that researched the clinical condition before late surgery group, patients whose condition deteriorated and died
surgery. When patients were in good presurgical condition, late surgery while waiting for surgery would not undergo presurgical condition
reduced the occurrence of poor outcome (RR, 1.59 [95%CI, 1.04 2.44]; evaluation. This classification was equal to assigning fewer patients in

269
Z. Yao et al. International Journal of Surgery 48 (2017) 266–274

Fig. 3. Forest plot of studies researching on death in all the patients.

poor condition to the late surgery group, producing a trend toward predispose the anti-vasospasm therapy to best effect. Even patients in
more good outcomes in the late surgery group. In comparison, the poor grade on admission fared better in the early surgery group
clinical condition on admission may better reflect the patient's state and [22–24]. The reasons impeding the application of early surgery in-
avoid selection bias. The rational method is to evaluate the clinical cluded highly reactive brain tissue, increased surgical difficulty and
condition on admission and conduct an “intention to treat” paradigm. patients' poor tolerance to early surgery. As far as advanced perio-
Even patients who received emergency or delayed surgery due to re- perative treatment and microsurgery were concerned, these problems
bleeding or a condition of deterioration still remained in the primary could be reduced and avoided. Some studies further divided early
group when evaluating overall outcomes. surgery and found ultra-early surgery also achieved a favorable out-
In our meta-analyses of overall effects, no significant differences come [13,25]. However, the financial and logistical burden related to
were found between the early and late surgery group, although sub- ultra-early surgery should be taken into consideration.
stantial heterogeneity existed (Figs. 2 and 3). In the subgroup analyses, In the subgroup analyses stratified by clinical condition before
we differentiated the clinical condition on admission from the clinical surgery, late surgery rather than early surgery reduced poor outcome
condition before surgery. In the subgroup analyses stratified by condi- and death when patients were in good presurgical condition. However,
tion on admission, early surgery showed superiority over late surgery in as previously mentioned, classification based on the presurgical con-
reducing poor outcome and death when patients were in good grade on dition led to selection bias and made the results unreliable. In addition,
admission and led to a decreased incidence of poor outcome when fewer studies contributing to the results and heterogeneity precluded
patients were in poor grade on admission (Figs. 4 and 7). Though the further interpretation of the results. Except for the clinical condition,
previous review found no significant differences between early and late the amount of blood in the subarachnoid space and age of the patient
surgery for poor outcome and death [17], the result was not stratified were potential confounding factors influencing outcomes [17]. Due to
by clinical condition on admission, and the analyses mixed en- the lack of data, the influence of the amount of blood was not analyzed
dovascular and surgical treatment [18]. It was reported that rebleeding in this review. In the subgroup analyses stratified by age, early treat-
and vasospasm were mainly responsible for the mortality in patients ment led to less death than late surgery in patients older than 50 years,
surviving the initial hemorrhage [2,19–21]. The surgical removal of in line with previous research [17]. This might be explained by a
blood in the cisterns alleviated vasospasm, and the secured aneurysms greater incidence of rebleeding in older patients who were more vul-
allowed for an aggressive anti-vasospasm treatment. It seemed rational nerable to secondary insults due to impaired cerebral autoregulation. In
to operate on patients at an early stage to avoid rebleeding and addition, an age of 50 years was arbitrarily adopted as a cut-off, the

270
Z. Yao et al. International Journal of Surgery 48 (2017) 266–274

Fig. 4. Forest plot of studies researching on the poor outcome of patients evaluated for clinical condition on admission.

validity of which is obscure. Considering only 4 studies contributed to and death when patients were in good condition on admission and for
these results, conclusions on the issue should be made carefully, as decreasing the incidence of poor outcomes when patients were in poor
more studies are needed. outcome on admission. Age was a potential confounding factor influ-
Both clinical experience and recent studies support early surgery for encing the effect of early surgery and is worthy of more study.
aSAH despite an absence of consistent results [7,26,27]. This systematic Further studies are required to focus on the amount of blood and age
review explains possible confounding factors and draws the conclusion as confounding factors, as well as the effect of ultra-early surgery for
that early surgery holds an advantage over late surgery for reducing poor aSAH.
outcomes and death, even in patients in poor grade on admission.
Nevertheless, several limitations exist in this review. First, omitting gray
literature and non-English articles might lead to publication bias. Second, Conflicts of interest
the definition of late surgery in the included studies was not totally con-
sistent. In addition, as most of the included patients suffered aSAH from None declared.
anterior circulation aneurysms, generalization to patients with posterior
circulation aneurysms should be approached with caution.
Sources of funding
5. Conclusions
This work was supported by the [Science and Technology
After stratifying patients by multiple factors, we found that the Supportive Project of Sichuan Province]; under Grant [No.
clinical condition on admission was an influential confounding factor. 2015SZ0051] [1.3.5 project for disciplines of excellence, West China
Early surgery was superior to late surgery for reducing poor outcomes Hospital, Sichuan University]; under Grant [No. ZY2016102].

271
Z. Yao et al. International Journal of Surgery 48 (2017) 266–274

Table 2
Results of subgroup analyses stratified by study type, country, year of publication, age, clinical condition before surgery, clinical condition on admission and follow-up time.

Subgroup Poor outcome Death

N RR 95% CI p I2 N RR 95% CI p I2

Study type
randomized 1 0.42 0.17, 1.04 0.06 – 1 0.44 0.14, 1.36 0.15 –
prospective 1 0.59 0.44, 0.80 0.0005 – 1 0.62 0.43, 0.89 0.01 –
retrospective 11 0.94 0.66, 1.34 0.72 92% 8 1.01 0.48, 2.11 0.98 93%
Country
single center 10 0.87 0.56, 1.36 0.55 93% 9 0.93 0.47, 1.86 0.84 93%
multinational 3 0.82 0.59, 1.14 0.23 76% 1 0.62 0.43, 0.89 0.01 –
Publication year
before 2000 8 0.88 0.54, 1.43 0.59 94% 8 0.92 0.47, 1.82 0.82 94%
after 2000 5 0.84 0.62, 1.13 0.25 65% 2 0.89 0.46, 1.73 0.73 0
Age (years)
younger than 50 3 0.85 0.28, 2.52 0.76 96% 3 1.09 0.26, 4.48 0.91 92%
older than 50 8 0.78 0.61, 1.00 0.05 66% 4 0.49 0.27, 0.89 0.02 52%
Condition before surgery
good 4 1.59 1.04, 2.44 0.03 76% 2 3.38 2.13, 5.38 < 0.0001 3%
poor 1 1.22 0.89, 1.68 0.22 – 1 2.33 1.28, 4.27 0.006 –
Condition on admission
good 8 0.65 0.50, 0.84 0.001 54% 7 0.61 0.46, 0.82 0.001 0
poor 5 0.71 0.61, 0.83 < 0.0001 8% 6 0.61 0.37, 1.00 0.05 78%
Follow-up time
less than 6 month 7 0.84 0.52, 1.35 0.48 90% 6 1.02 0.53, 1.97 0.95 84%
more than 6 month 6 0.88 0.54, 1.44 0.61 93% 4 0.75 0.23, 2.48 0.64 96%

CI, confidence interval; N, number of studies; RR, risk ratio.

Fig. 5. Begg's funnel plot. a funnel plot of studies researching on the poor outcome of Fig. 6. Sensitivity analyses. a sensitivity analyses of studies researching on the poor
patients evaluated for clinical condition on admission, b funnel plot of studies researching outcome of patients evaluated for clinical condition on admission, b sensitivity analyses
on death in patients evaluated for clinical condition on admission. of studies researching on death in patients evaluated for clinical condition on admission.

272
Z. Yao et al. International Journal of Surgery 48 (2017) 266–274

Fig. 7. Forest plot of studies researching on death in patients evaluated for clinical condition on admission.

Ethical approval aneurysm surgery in Japan, Stroke 16 (1) (1985 Jan-Feb) 48–52 PubMed PMID:
3966265.
[2] M. Miyaoka, K. Sato, S. Ishii, A clinical study of the relationship of timing to out-
This review needs no ethical approval. come of surgery for ruptured cerebral aneurysms. A retrospective analysis of 1622
cases, J. Neurosurg. 79 (3) (1993 Sep) 373–378, http://dx.doi.org/10.3171/jns.
Research registration unique identifying number (UIN) 1993.79.3.0373 PubMed PMID: 8360733.
[3] K. de Gans, D.J. Nieuwkamp, G.J. Rinkel, et al., Timing of aneurysm surgery in
subarachnoid hemorrhage: a systematic review of the literature, Neurosurgery 50
researchregistry3284. (2) (2002 Feb) 336–340 discussion 340-2. PubMed PMID: 11844269.
[4] D. Moher, A. Liberati, J. Tetzlaff, et al., Preferred reporting items for systematic
reviews and meta-analyses: the PRISMA statement, BMJ 339 (2009 Jul 21) b2535,
Author contribution
http://dx.doi.org/10.1136/bmj.b2535 PubMed PMID: 19622551; PubMed Central
PMCID: PMCPMC2714657.
Zhong Yao was responsible for data collections, statistic analyses [5] N.F. Kassell, J.C. Torner, J.A. Jane, et al., The international cooperative study on
the timing of aneurysm surgery. Part 2: surgical results, J. Neurosurg. 73 (1) (1990
and drafting manuscript.
Jul) 37–47, http://dx.doi.org/10.3171/jns.1990.73.1.0037 PubMed PMID:
Xin Hu and Lu Ma were responsible for study design and data col- 2191091.
lections. [6] J. Ohman, O. Heiskanen, Timing of operation for ruptured supratentorial aneur-
Chao You designed the study and revised the manuscript. ysms: a prospective randomized study, J. Neurosurg. 70 (1) (1989 Jan) 55–60,
http://dx.doi.org/10.3171/jns.1989.70.1.0055 PubMed PMID: 2909689.
Min He was responsible for data analysis and revising the manu- [7] K.B. Mahaney, M.M. Todd, J.C. Torner, et al., Variation of patient characteristics,
script. management, and outcome with timing of surgery for aneurysmal subarachnoid
hemorrhage, J. Neurosurg. 114 (4) (2011 Apr) 1045–1053, http://dx.doi.org/10.
3171/2010.11.JNS10795 PubMed PMID: 21250801.
Guarantor [8] E.C. Haley Jr., N.F. Kassell, J.C. Torner, The international cooperative study on the
timing of aneurysm surgery. The north american experience, Stroke 23 (2) (1992
Min He. Feb) 205–214 PubMed PMID: 1561649.
[9] N. Ross, P.J. Hutchinson, H. Seeley, et al., Timing of surgery for supratentorial
aneurysmal subarachnoid haemorrhage: report of a prospective study, J. Neurol.
References Neurosurg. Psychiatry 72 (4) (2002 Apr) 480–484 PubMed PMID: 11909907;
PubMed Central PMCID: PMCPMC1737846.
[10] N. Golchin, S.M. Ramak Hashem, E. Abbas Nejad, et al., Timing of surgery for
[1] A. Nishimoto, K. Ueta, H. Onbe, et al., Nationwide co-operative study of intracranial

273
Z. Yao et al. International Journal of Surgery 48 (2017) 266–274

aneurysmal subarachnoid hemorrhage, Acta Med. Iran. 50 (5) (2012) 300–304 Neurol. Res. 16 (1) (1994 Feb) 23–26 PubMed PMID: 7913524.
PubMed PMID: 22837081. [20] B. Weir, K. Aronyk, Management mortality and the timing of surgery for supra-
[11] S. Dalbayrak, M. Altas, R. Arslan, The effects of timing of aneurysm surgery on tentorial aneurysm, J. Neurosurg. 54 (2) (1981 Feb) 146–150, http://dx.doi.org/10.
vasospasm and mortality in patients with subarachnoid hemorrhage, Acta Neurol. 3171/jns.1981.54.2.0146 PubMed PMID: 7452328.
Belg 111 (4) (2011 Dec) 317–320 PubMed PMID: 22368972. [21] A. Tamasauskas, J. Tamasauskas, G. Bernotas, et al., Management of patients with
[12] K.C. Lee, Surgery of intracranial aneurysms at Yonsei University: 780 cases, Keio J. ruptured cerebral aneurysms in hospital population of Lithuania, Acta Neurochir.
Med. 40 (1) (1991 Mar) 1–5 7. (Wien) 142 (1) (2000) 51–59 PubMed PMID: 10664376.
[13] T.J. Phillips, R.J. Dowling, B. Yan, et al., Does treatment of ruptured intracranial [22] K. Yamamoto, I. Ezuka, N. Takai, et al., Comparison of late and early stage surgery
aneurysms within 24 hours improve clinical outcome? Stroke 42 (7) (2011 Jul) for ruptured intracranial aneurysms, Neurol. Med. Chir. (Tokyo) 32 (1) (1992 Jan)
1936–1945, http://dx.doi.org/10.1161/Strokeaha.110.602888 PubMed PMID: 1–4 PubMed PMID: 1375977.
WOS:000292090900036; English. [23] N.F. Kassell, D.J. Boarini, H.P. Adams, et al., Overall management of ruptured
[14] S.C. Oudshoorn, G.J. Rinkel, A.J. Molyneux, et al., Aneurysm treatment < 24 versus aneurysm - comparison of early and late operation, Neurosurgery 9 (2) (1981)
24-72 h after subarachnoid hemorrhage, Neurocrit Care 21 (1) (2014 Aug) 4–13, 120–128 PubMed PMID: WOS: A1981MC17200002; English.
http://dx.doi.org/10.1007/s12028-014-9969-8 PubMed PMID: 24639201. [24] H. Tan, G. Huang, Z. Li, et al., The impact of surgical timing on the management of
[15] R. Ferch, A. Pasqualin, G. Barone, et al., Surgical management of ruptured aneur- aneurysms with acute hydrocephalus after aneurysmal subarachnoid hemorrhage,
ysms in the eighth and ninth decades, discussion 445, Acta Neurochir. (Wien) 145 Turk Neurosurg. 24 (3) (2014) 385–390, http://dx.doi.org/10.5137/1019-5149.
(6) (2003 Jun) 439–445, http://dx.doi.org/10.1007/s00701-003-0037-y PubMed JTN.9484-13.0 PubMed PMID: 24848179.
PMID: 12836067. [25] M. Taneda, The significance of early operation in the management of ruptured
[16] K. Abe, A. Demizu, K. Kamada, et al., Prostaglandin E1 and carbon dioxide re- intracranial aneurysms–an analysis of 251 cases hospitalized within 24 hours after
activity during cerebral aneurysm surgery, Can. J. Anaesth. 39 (3) (1992 Mar) subarachnoid haemorrhage, Acta Neurochir. (Wien) 63 (1–4) (1982) 201–208
247–252, http://dx.doi.org/10.1007/BF03008785 PubMed PMID: 1551156. PubMed PMID: 7102411.
[17] C. Zhao, Y. Wei, Surgical timing for aneurysmal subarachnoid hemorrhage: a meta- [26] S.M. Dorhout Mees, A.J. Molyneux, R.S. Kerr, et al., Timing of aneurysm treatment
analysis and systematic review, Turk Neurosurg. 27 (4) (2017) 489–499, http://dx. after subarachnoid hemorrhage: relationship with delayed cerebral ischemia and
doi.org/10.5137/1019-5149.JTN.16422-15.0 PubMed PMID: 27759872. poor outcome, Stroke 43 (8) (2012 Aug) 2126–2129, http://dx.doi.org/10.1161/
[18] Z. Qian, T. Peng, A. Liu, et al., Early timing of endovascular treatment for aneur- STROKEAHA.111.639690 PubMed PMID: 22700527.
ysmal subarachnoid hemorrhage achieves improved outcomes, Curr. Neurovasc [27] T. Kayama, T. Yoshimoto, K. Uchida, et al., Intracranial saccular aneur-
Res. 11 (1) (2014 Feb) 16–22 PubMed PMID: 24320010. ysms–surgical results of 1,000 consecutive cases, Tohoku J. Exp. Med. 126 (2)
[19] K. Sano, Grading and timing of surgery for aneurysmal subarachnoid haemorrhage, (1978 Oct) 117–124 PubMed PMID: 715762.

274