Radiological Management of Angiographically Negative, Spontaneous Intracranial Subarachnoid Hemorrhage: A Multicenter Study of Utilization and Diagnostic Yield

Radiological Management of Angiographically Negative, Spontaneous Intracranial Subarachnoid... Abstract BACKGROUND The optimal diagnostic evaluation for patients with angiographically negative subarachnoid hemorrhage (AN-SAH) remains controversial. OBJECTIVE To assess the utilization rate and diagnostic yield of imaging tests routinely obtained in identifying a structural cause for AN-SAH. METHODS In this retrospective multicenter study, consecutive adult patients admitted with nontraumatic, AN-SAH between 01/2010 and 12/2015 were included. Patients with intraparenchymal, subdural, or epidural hematomas in addition to SAH were excluded. Outcomes studied included utilization rate, diagnostic yield, and median time from admission for the following imaging tests: initial computed tomography angiography (CTA) and digital subtraction angiography (DSA), brain and cervical spine magnetic resonance imaging (MRI), and any repeat DSA or CTA performed either during initial admission or at long-term follow-up. RESULTS A total of 752 patients were included (mean age, 53 yr; 54% male). Initial CTA and DSA were performed in 89% and 100% of patients, respectively. Brain MRI was performed in 75% of patients and was positive in 0.7% of cases. Cervical spine MRI was performed in 61% of patients and was positive in 0.2% of cases. Repeat, same-admission follow-up DSA and CTA were performed in 48% and 51% of patients and were positive in 3.3% and 1% of cases, respectively. Delayed follow-up DSA and CTA after discharge were performed in 26% and 7% of patients and were positive in 2% and 3.7% of cases, respectively, all with negative prior imaging studies. CONCLUSION Cervical spine and brain MRI have extremely low diagnostic yield, both are commonly utilized in patients with AN-SAH; while repeat DSA and CTA are utilized less commonly and have slightly higher diagnostic yield. Angiographically negative subarachnoid hemorrhage, Management, Imaging test, Utilization, Diagnostic yield ABBREVIATIONS ABBREVIATIONS AN-SAH angiographically negative subarachnoid hemorrhage CI confidence interval CTA computed tomography angiography DSA digital subtraction angiography EMR electronic medical record IQR median and interquartile ranges IRB institutional review board MRI magnetic resonance imaging NPV negative predictive value SD standard deviations Nontraumatic subarachnoid hemorrhage (SAH) occurs in 30 000 patients per year in the United States1 and accounts for 7% of all strokes2. Nearly half of the SAH survivors may have long-term cognitive impairment due to symptomatic vasospasm, which impacts functional status and quality of life.3,4 In 15% to 20% of SAH patients, angiographic evaluation fails to identify an etiology for the hemorrhage, also known as angiographically negative SAH (AN-SAH).5,6 These patients are at risk for rehemorrhage and its consequences such as vasospasm, if structural lesions are missed on the initial angiogram and treatment is delayed.7 Therefore, early diagnosis of the hemorrhage etiology and aggressive management of SAH is highly recommended by current guidelines.8 The optimal diagnostic evaluation for patients with AN-SAH remains controversial.5,9 Many researchers have suggested that a repeat angiogram may detect etiologies missed on initial angiographic imaging, although the reported diagnostic yield varies considerably.10-12 Currently, the diagnostic protocol for management of AN-SAH varies between institutions and may include a combination of initial and repeat computed tomography angiography (CTA) and/or digital subtraction angiography (DSA) with brain and/or cervical spine magnetic resonance imaging (MRI). However, there are limited data available on the diagnostic yield of each of these imaging tests and these are derived from published studies with small sample sizes.13 Patients with perimesencephalic pattern of SAH14,15 are defined as those in whom center of hemorrhage is anterior to midbrain with or without extension of blood to the anterior part of the ambient cistern or the basal part of the sylvian fissure, incomplete filling of the anterior interhemispheric fissure and no extension of the lateral sylvian fissure or the ventricles.14 These patients less commonly develop hydrocephalus, vasospasm, or other complications and have lower risk of recurrent bleeding.14,16-18 The better overall prognosis in such cases has prompted development of less aggressive management algorithms in benign perimesencephalic pattern of hemorrhage, such as foregoing secondary angiography.19 However, in patients with nonperimesencephalic pattern of AN-SAH, the diagnostic and therapeutic approach is less well defined.17 The purpose of the current study was to assess the utilization rate and diagnostic yield of imaging tests routinely obtained in identifying a structural cause for AN-SAH in a multicenter study. METHODS Institutional Review Board (IRB) approval and a waiver of informed consent were obtained for this retrospective multicenter review. The study was Health Insurance Probability and Accountability Act compliant. Study Population A multicenter radiology informatics system search was conducted for consecutive patients aged 18 yr or older with acute, nontraumatic, intracranial SAH, diagnosed based on noncontrast head CT or xanthochromia on lumbar puncture. A total of 7 university hospitals from 5 academic institutions in the United States and 1 academic institution in the United Kingdom were included in a study period spanning January 2010 to December 2015 in 4 of the participating institutions. In 1 institution, the study period was longer, spanning January 2001 to December 2015. Only patients with a negative initial cerebrovascular DSA were included, regardless of the type of follow-up imaging they received. Patients with misdiagnosis of SAH (eg, due to traumatic spinal tap) were excluded. Further exclusion criteria included patients with acute intraparenchymal, subdural, or epidural hemorrhage; cerebral contusion; recent trauma; surgery to the brain or spine within last 2 wk of presentation; and patients with known etiology for SAH established on CTA or DSA from prior admissions. The identification of eligible patients for the purpose of the study was performed between July 2016 and April 2017. Medical Record Review Demographic information including age at presentation and sex, past medical history of hypertension, diabetes (type 1 or 2), hyperlipidemia, history of any lifetime tobacco use, and Hunt and Hess grade20 at presentation were obtained from the electronic medical record (EMR). We further documented final disposition status: discharge home or to a rehabilitation facility, or death during initial hospitalization. Imaging Analysis The initial noncontrast head CT scan that was obtained within 24 h of admission and prior to a brain MRI was reviewed by investigators from each center, and the perimesencephalic vs nonperimesencephalic pattern of hemorrhage according to Rinkel et al14 was assessed and a Fisher grade21 was calculated. If a head CT was not obtained until after the brain MRI was obtained, or if no head CT was available, pattern of hemorrhage or Fisher grade was not assessed to mitigate potential errors related to redistribution of hemorrhage during the course of admission. The postadmission imaging tests included initial DSA and CTA, brain and cervical spine MRI, repeat DSA and CTA performed during the initial admission and delayed DSA and CTA performed at long-term follow-up. For each included patient, subsequent neurological imaging tests obtained later in admission and after discharge were recorded, and the timing of such imaging logged relative to the time of admission. Formal interpretations were reviewed for any findings potentially explaining initial SAH. For cases in which vague or uncertain potential etiologies were identified, the EMR was further reviewed for follow-up neurology, neurosurgery, and neurointerventional notes, and the findings included only if the hemorrhage was ascribable to the findings in question on the basis of aggregated available clinical information. Statistical Analysis Categorical variables were reported as frequencies and percentages. Numerical variables were presented as mean and standard deviations (SD) or median and interquartile ranges (IQR) if they did not follow a normal distribution. Utilization rate was assessed as a proportion of patients with AN-SAH who received the imaging test. Diagnostic yield was calculated as a proportion of patients with imaging test who had an identifiable etiology for SAH. Negative predictive value (NPV) was calculated for each of the imaging tests separately and compared between nonvascular imaging tests (eg, patient undergoing either brain or cervical spine MRI) and vascular imaging tests (eg, patient undergoing either DSA or CTA) using Fisher's Exact Test, given not all the patients had received all the imaging modalities at follow-up. For nonvascular vs vascular imaging comparison, imaging was considered negative only if both imaging tests in each group (eg, both brain and cervical spine MRI for nonvascular imaging, and both CTA and DSA for vascular imaging) were negative. The analyses were performed using STATA 10 (Stata Corp., College Station, Texas). P values < .05 were considered statistically significant. RESULTS Study Population A total of 752 patients with AN-SAH (406/752 men, age 18-88 yr old [mean 53 yr]) were included in the analyses, with baseline characteristics shown in Table 1. Half of the included patients presented with Hunt and Hess grade 1 (50%, 373/752). Of included patients, 29% (220/752) had perimesencephalic pattern of hemorrhage; while 51% (384/752) had nonperimsencephalic hemorrhage and the remaining 20% (148/752) either did not have CT scan on admission or had no hemorrhage on CT scan (eg, Fisher grade I). TABLE 1. Baseline Characteristics for 752 patients With Angiographically Negative SAH Mean age, yr ± SD  53 ± 14  Gender, % (n)     Male  54 (406)   Female  46 (346)  Past Medical History, % (n)     Hypertension  32 (241)   Diabetes Type I  2 (18)   Diabetes Type II  8 (61)   Hyperlipidemia  13 (99)  Any Life-time Tobacco Use, % (n)  32 (242)  Hunt and Hess Grade,20 % (n)    Not reported  15 (113)   1  50 (373)   2  16 (124)   3  14 (106)   4  4 (31)   5  1 (5)  Fisher Grade based on admission noncontrast Head CT, % (n)    No admission CT scan available or unknown  10 (72)   1  15 (114)   2  21 (159)   3  29 (220)   4  25 (186)  Pattern of Hemorrhage on admission noncontrast Head CT, % (n)    No admission CT scan available or Fisher grade I (no SAH on CT scan)  20 (148)  Perimesencephalic  29 (220)  Nonperimesencephalic  51 (384)  Mean age, yr ± SD  53 ± 14  Gender, % (n)     Male  54 (406)   Female  46 (346)  Past Medical History, % (n)     Hypertension  32 (241)   Diabetes Type I  2 (18)   Diabetes Type II  8 (61)   Hyperlipidemia  13 (99)  Any Life-time Tobacco Use, % (n)  32 (242)  Hunt and Hess Grade,20 % (n)    Not reported  15 (113)   1  50 (373)   2  16 (124)   3  14 (106)   4  4 (31)   5  1 (5)  Fisher Grade based on admission noncontrast Head CT, % (n)    No admission CT scan available or unknown  10 (72)   1  15 (114)   2  21 (159)   3  29 (220)   4  25 (186)  Pattern of Hemorrhage on admission noncontrast Head CT, % (n)    No admission CT scan available or Fisher grade I (no SAH on CT scan)  20 (148)  Perimesencephalic  29 (220)  Nonperimesencephalic  51 (384)  View Large TABLE 1. Baseline Characteristics for 752 patients With Angiographically Negative SAH Mean age, yr ± SD  53 ± 14  Gender, % (n)     Male  54 (406)   Female  46 (346)  Past Medical History, % (n)     Hypertension  32 (241)   Diabetes Type I  2 (18)   Diabetes Type II  8 (61)   Hyperlipidemia  13 (99)  Any Life-time Tobacco Use, % (n)  32 (242)  Hunt and Hess Grade,20 % (n)    Not reported  15 (113)   1  50 (373)   2  16 (124)   3  14 (106)   4  4 (31)   5  1 (5)  Fisher Grade based on admission noncontrast Head CT, % (n)    No admission CT scan available or unknown  10 (72)   1  15 (114)   2  21 (159)   3  29 (220)   4  25 (186)  Pattern of Hemorrhage on admission noncontrast Head CT, % (n)    No admission CT scan available or Fisher grade I (no SAH on CT scan)  20 (148)  Perimesencephalic  29 (220)  Nonperimesencephalic  51 (384)  Mean age, yr ± SD  53 ± 14  Gender, % (n)     Male  54 (406)   Female  46 (346)  Past Medical History, % (n)     Hypertension  32 (241)   Diabetes Type I  2 (18)   Diabetes Type II  8 (61)   Hyperlipidemia  13 (99)  Any Life-time Tobacco Use, % (n)  32 (242)  Hunt and Hess Grade,20 % (n)    Not reported  15 (113)   1  50 (373)   2  16 (124)   3  14 (106)   4  4 (31)   5  1 (5)  Fisher Grade based on admission noncontrast Head CT, % (n)    No admission CT scan available or unknown  10 (72)   1  15 (114)   2  21 (159)   3  29 (220)   4  25 (186)  Pattern of Hemorrhage on admission noncontrast Head CT, % (n)    No admission CT scan available or Fisher grade I (no SAH on CT scan)  20 (148)  Perimesencephalic  29 (220)  Nonperimesencephalic  51 (384)  View Large Utilization Rate of Imaging Tests for Management of AN-SAH Initial postadmission DSA was performed for 100% (n = 752) of patients within 10 d of admission (median = 1 d postadmission, IQR = 1). Initial postadmission CTA was performed for 89% (668/752) of patients within 10 d of admission (median = 0 d, IQR = 1). Brain MRI and cervical spine MRI were performed in 75% (562/752) and 61% (457/752) of patients, respectively. The median time for both tests was 2 (IQR = 29) d after admission. Repeat, same-admission follow-up DSA and CTA were performed in 48% (362/752) and 51% (380/752) of patients, respectively. Only 16% (121/752) of patients received both DSA and CTA follow-up tests during the same admission and 17% (131/752) received neither of these 2 follow-up tests during the same admission. The median time for repeat, same-admission follow-up DSA was 10 (IQR = 7) d after admission and for CTA was 7 (IQR = 4) d after admission (Figure 1). FIGURE 1. View largeDownload slide Utilization rate of imaging tests for management of angiographically negative subarachnoid hemorrhage in 752 patients. C-spine, cervical spine; Re-DSA, repeat same admission follow-up DSA; Re-CTA, repeat same admission follow-up CTA; Del-DSA, delayed post discharge follow-up DSA; Del-CTA, delayed post discharge follow-up CTA. FIGURE 1. View largeDownload slide Utilization rate of imaging tests for management of angiographically negative subarachnoid hemorrhage in 752 patients. C-spine, cervical spine; Re-DSA, repeat same admission follow-up DSA; Re-CTA, repeat same admission follow-up CTA; Del-DSA, delayed post discharge follow-up DSA; Del-CTA, delayed post discharge follow-up CTA. Delayed follow-up DSA and CTA after discharge were performed in 26% (197/752) and 7% (53/752) of patients, respectively. Only 2% (14/752) of patients received both follow-up tests after discharge and 69% (516/752) received neither of these 2 follow-up tests after discharge. The median time for delayed postdischarge follow-up DSA was 90 (IQR = 40) d after admission and for CTA was 93 (IQR = 90) d after admission (Figure 1). Diagnostic Yield of Imaging Tests for Management of AN-SAH Overall, regardless of type of follow-up imaging, 2.7% (21/752) of patients had a positive finding to explain the etiology of SAH on at least one of the follow-up imaging tests. Of these 21 patients with positive findings, 19 had nonperimesencephalic pattern of hemorrhage and 2 did not have an admission head CT available for review due to being transferred from another hospital. However, per EMR the pattern of hemorrhage for both of these patients was reported as nonperimsencephalic at the outside hospital. In 1.6% of patients (12/752), an aneurysm considered to be the etiology of the SAH (Table 2). Remaining etiologies for SAH identified on follow-up imaging were as follows: reversible cerebral vasoconstriction syndrome (0.5% [4/752]), inflammatory vasculopathy (0.3% [2/752]), dural arteriovenous fistula in the posterior fossa (0.1% [1/752]), brain cavernous malformation (0.1% [1/752]), and cervical spine dural arterivenous fistula (0.1% [1/752]; Table 2; Figure 2). FIGURE 2. View largeDownload slide Axial CT scan in a 61-yr-old male showing subarachnoid hemorrhage at the level of foramen magnum (A) with intraventricular extension (B). CTA of the head did not show any etiology for subarachnoid hemorrhage. Sagittal T1 (C), T2 (D), and T1 postcontrast (E) images of cervical spine showed a cervical dural arteriovenous fistula, and this was later confirmed on dedicated cervical angiography. FIGURE 2. View largeDownload slide Axial CT scan in a 61-yr-old male showing subarachnoid hemorrhage at the level of foramen magnum (A) with intraventricular extension (B). CTA of the head did not show any etiology for subarachnoid hemorrhage. Sagittal T1 (C), T2 (D), and T1 postcontrast (E) images of cervical spine showed a cervical dural arteriovenous fistula, and this was later confirmed on dedicated cervical angiography. TABLE 2. Imaging Characteristics for 21 AN-SAH Patients With Positive Findings on Follow-up Imaging to Explain SAH Patient  Pattern of Hemorrhage  Etiology  Brain MRI  C-Spine MRI  Re-DSA  Re-CTA  Del-DSA  Del-CTA  Aneurysmal etiologies                  42 yo F  Nonperimesencephalic  Supraclinoid ICA  +  ND  +  +  ND  ND  69 yo M  Nonperimesencephalic  Supraclinoid ICA  −  ND  +  −  ND  ND  48 yo F  Nonperimesencephalic  ACOM  −  −  +  +  ND  ND  43 yo F  Nonperimesencephalic  ACOM  −  −  ND  −  +  ND  47 yo F  Nonperimesencephalic  ACOM  −  −  ND  −  +  ND  62 yo F  Nonperimesencephalic  ACA  −  −  +  ND  ND  ND  43 yo M  Nonperimesencephalic  ACA  ND  ND  +  ND  ND  ND  74 yo M  Nonperimesencephalic  Paraophthalmic  ND  ND  +  ND  ND  ND  60 yo F  Nonperimesencephalic  PCA  −  −  +  ND  ND  ND  42 yo F  No head CT*  PCA  −  −  −  ND  +  +  52 yo F  No head CT*  PICA  −  −  ND  +  ND  ND  53 yo F  Nonperimesencephalic  Micronaneurysms in posterior circulation  −  −  ND  −  +  +  Nonaneurymal etiologies  56 yo F  Nonperimesencephalic  RCVS  −  −  +  ND  ND  ND  62 yo F  Nonperimesencephalic  RCVS  −  −  +  +  ND  ND  61 yo F  Nonperimesencephalic  RCVS  ND  ND  +  ND  ND  ND  59 yo F  Nonperimesencephalic  RCVS  −  −  +  ND  ND  ND  47 yo F  Nonperimesencephalic  Vasculopathy  +  −  ND  ND  ND  ND  77 yo M  Nonperimesencephalic  Vasculopathy  +  ND  ND  ND  ND  ND  49 yo M  Nonperimesencephalic  Dural AVF in posterior fossa  −  −  +  ND  ND  ND  31 yo F  Nonperimesencephalic  Brain cavernous malformation  +  −  ND  ND  ND  ND  61 yo M  Nonperimesencephalic  C-spine dural AVF  −  +  ND  ND  ND  ND  Patient  Pattern of Hemorrhage  Etiology  Brain MRI  C-Spine MRI  Re-DSA  Re-CTA  Del-DSA  Del-CTA  Aneurysmal etiologies                  42 yo F  Nonperimesencephalic  Supraclinoid ICA  +  ND  +  +  ND  ND  69 yo M  Nonperimesencephalic  Supraclinoid ICA  −  ND  +  −  ND  ND  48 yo F  Nonperimesencephalic  ACOM  −  −  +  +  ND  ND  43 yo F  Nonperimesencephalic  ACOM  −  −  ND  −  +  ND  47 yo F  Nonperimesencephalic  ACOM  −  −  ND  −  +  ND  62 yo F  Nonperimesencephalic  ACA  −  −  +  ND  ND  ND  43 yo M  Nonperimesencephalic  ACA  ND  ND  +  ND  ND  ND  74 yo M  Nonperimesencephalic  Paraophthalmic  ND  ND  +  ND  ND  ND  60 yo F  Nonperimesencephalic  PCA  −  −  +  ND  ND  ND  42 yo F  No head CT*  PCA  −  −  −  ND  +  +  52 yo F  No head CT*  PICA  −  −  ND  +  ND  ND  53 yo F  Nonperimesencephalic  Micronaneurysms in posterior circulation  −  −  ND  −  +  +  Nonaneurymal etiologies  56 yo F  Nonperimesencephalic  RCVS  −  −  +  ND  ND  ND  62 yo F  Nonperimesencephalic  RCVS  −  −  +  +  ND  ND  61 yo F  Nonperimesencephalic  RCVS  ND  ND  +  ND  ND  ND  59 yo F  Nonperimesencephalic  RCVS  −  −  +  ND  ND  ND  47 yo F  Nonperimesencephalic  Vasculopathy  +  −  ND  ND  ND  ND  77 yo M  Nonperimesencephalic  Vasculopathy  +  ND  ND  ND  ND  ND  49 yo M  Nonperimesencephalic  Dural AVF in posterior fossa  −  −  +  ND  ND  ND  31 yo F  Nonperimesencephalic  Brain cavernous malformation  +  −  ND  ND  ND  ND  61 yo M  Nonperimesencephalic  C-spine dural AVF  −  +  ND  ND  ND  ND  C-spine, cervical spine; Re-DSA, repeat same admission follow-up DSA; Re-CTA, repeat same admission follow-up CTA; Del-DSA, delayed postdischarge follow-up DSA; Del-CTA, delayed postdischarge follow-up CTA; F, female; M, Male; ICA, internal carotid artery; ACA, anterior cerebral artery; ACOM, anterior communicating artery; PCA, posterior cerebral artery; PICA, posterior inferior cerebellar artery; RCVS, reversible cerebral vasoconstriction syndrome; ND, not DONE. *These 2 patients did not have an admission head CT available due to being transferred from another hospital. However, per EMR the pattern of hemorrhage for both of these patients was reported as nonperimsencephalic at the outside hospital. View Large TABLE 2. Imaging Characteristics for 21 AN-SAH Patients With Positive Findings on Follow-up Imaging to Explain SAH Patient  Pattern of Hemorrhage  Etiology  Brain MRI  C-Spine MRI  Re-DSA  Re-CTA  Del-DSA  Del-CTA  Aneurysmal etiologies                  42 yo F  Nonperimesencephalic  Supraclinoid ICA  +  ND  +  +  ND  ND  69 yo M  Nonperimesencephalic  Supraclinoid ICA  −  ND  +  −  ND  ND  48 yo F  Nonperimesencephalic  ACOM  −  −  +  +  ND  ND  43 yo F  Nonperimesencephalic  ACOM  −  −  ND  −  +  ND  47 yo F  Nonperimesencephalic  ACOM  −  −  ND  −  +  ND  62 yo F  Nonperimesencephalic  ACA  −  −  +  ND  ND  ND  43 yo M  Nonperimesencephalic  ACA  ND  ND  +  ND  ND  ND  74 yo M  Nonperimesencephalic  Paraophthalmic  ND  ND  +  ND  ND  ND  60 yo F  Nonperimesencephalic  PCA  −  −  +  ND  ND  ND  42 yo F  No head CT*  PCA  −  −  −  ND  +  +  52 yo F  No head CT*  PICA  −  −  ND  +  ND  ND  53 yo F  Nonperimesencephalic  Micronaneurysms in posterior circulation  −  −  ND  −  +  +  Nonaneurymal etiologies  56 yo F  Nonperimesencephalic  RCVS  −  −  +  ND  ND  ND  62 yo F  Nonperimesencephalic  RCVS  −  −  +  +  ND  ND  61 yo F  Nonperimesencephalic  RCVS  ND  ND  +  ND  ND  ND  59 yo F  Nonperimesencephalic  RCVS  −  −  +  ND  ND  ND  47 yo F  Nonperimesencephalic  Vasculopathy  +  −  ND  ND  ND  ND  77 yo M  Nonperimesencephalic  Vasculopathy  +  ND  ND  ND  ND  ND  49 yo M  Nonperimesencephalic  Dural AVF in posterior fossa  −  −  +  ND  ND  ND  31 yo F  Nonperimesencephalic  Brain cavernous malformation  +  −  ND  ND  ND  ND  61 yo M  Nonperimesencephalic  C-spine dural AVF  −  +  ND  ND  ND  ND  Patient  Pattern of Hemorrhage  Etiology  Brain MRI  C-Spine MRI  Re-DSA  Re-CTA  Del-DSA  Del-CTA  Aneurysmal etiologies                  42 yo F  Nonperimesencephalic  Supraclinoid ICA  +  ND  +  +  ND  ND  69 yo M  Nonperimesencephalic  Supraclinoid ICA  −  ND  +  −  ND  ND  48 yo F  Nonperimesencephalic  ACOM  −  −  +  +  ND  ND  43 yo F  Nonperimesencephalic  ACOM  −  −  ND  −  +  ND  47 yo F  Nonperimesencephalic  ACOM  −  −  ND  −  +  ND  62 yo F  Nonperimesencephalic  ACA  −  −  +  ND  ND  ND  43 yo M  Nonperimesencephalic  ACA  ND  ND  +  ND  ND  ND  74 yo M  Nonperimesencephalic  Paraophthalmic  ND  ND  +  ND  ND  ND  60 yo F  Nonperimesencephalic  PCA  −  −  +  ND  ND  ND  42 yo F  No head CT*  PCA  −  −  −  ND  +  +  52 yo F  No head CT*  PICA  −  −  ND  +  ND  ND  53 yo F  Nonperimesencephalic  Micronaneurysms in posterior circulation  −  −  ND  −  +  +  Nonaneurymal etiologies  56 yo F  Nonperimesencephalic  RCVS  −  −  +  ND  ND  ND  62 yo F  Nonperimesencephalic  RCVS  −  −  +  +  ND  ND  61 yo F  Nonperimesencephalic  RCVS  ND  ND  +  ND  ND  ND  59 yo F  Nonperimesencephalic  RCVS  −  −  +  ND  ND  ND  47 yo F  Nonperimesencephalic  Vasculopathy  +  −  ND  ND  ND  ND  77 yo M  Nonperimesencephalic  Vasculopathy  +  ND  ND  ND  ND  ND  49 yo M  Nonperimesencephalic  Dural AVF in posterior fossa  −  −  +  ND  ND  ND  31 yo F  Nonperimesencephalic  Brain cavernous malformation  +  −  ND  ND  ND  ND  61 yo M  Nonperimesencephalic  C-spine dural AVF  −  +  ND  ND  ND  ND  C-spine, cervical spine; Re-DSA, repeat same admission follow-up DSA; Re-CTA, repeat same admission follow-up CTA; Del-DSA, delayed postdischarge follow-up DSA; Del-CTA, delayed postdischarge follow-up CTA; F, female; M, Male; ICA, internal carotid artery; ACA, anterior cerebral artery; ACOM, anterior communicating artery; PCA, posterior cerebral artery; PICA, posterior inferior cerebellar artery; RCVS, reversible cerebral vasoconstriction syndrome; ND, not DONE. *These 2 patients did not have an admission head CT available due to being transferred from another hospital. However, per EMR the pattern of hemorrhage for both of these patients was reported as nonperimsencephalic at the outside hospital. View Large Diagnostic yields for follow-up imaging are shown in Figure 3 and were as follows: brain MRI, 0.7% (4/562); cervical spine MRI, 0.2% (1/457); repeat, same-admission follow-up DSA, 3.3% (12/362); repeat, same-admission follow-up CTA, 1% (4/380); delayed postdischarge DSA, 2% (4/197); and delayed postdischarge CTA, 3.7% (2/53). FIGURE 3. View largeDownload slide Diagnostic yield of imaging tests for management of AN-SAH. The diagnostic yield is calculated as the proportion of patients undergoing each imaging test who have positive finding in the test to explain subarachnoid hemorrhage. Please note that the denominator is variable among different imaging tests as not all patients received all the tests. C-spine, cervical spine; Re-DSA, repeat same admission follow-up DSA; Re-CTA, repeat same admission follow-up CTA; Del-DSA, delayed postdischarge follow-up DSA; Del-CTA, delayed postdischarge follow-up CTA. FIGURE 3. View largeDownload slide Diagnostic yield of imaging tests for management of AN-SAH. The diagnostic yield is calculated as the proportion of patients undergoing each imaging test who have positive finding in the test to explain subarachnoid hemorrhage. Please note that the denominator is variable among different imaging tests as not all patients received all the tests. C-spine, cervical spine; Re-DSA, repeat same admission follow-up DSA; Re-CTA, repeat same admission follow-up CTA; Del-DSA, delayed postdischarge follow-up DSA; Del-CTA, delayed postdischarge follow-up CTA. Diagnostic yields for follow-up imaging in patients with nonperimesencephalic pattern of hemorrhage were as follows: brain MRI, 1.2% (4/308); cervical spine MRI, 0.3% (1/264); repeat, same-admission follow-up DSA, 7.4% (12/162); repeat, same-admission follow-up CTA, 1.7% (4/235); delayed postdischarge DSA, 3.1% (4/128); and delayed postdischarge CTA, 7.7% (2/26). Given the follow-up imaging was not positive in any of the patients with perimesencephalic pattern of hemorrhage, the diagnostic yield of all the follow-up imaging in these patients was 0%. Only in 4 patients with a positive finding for SAH, both repeat follow-up DSA and CTA were performed during the same admission. The 2 tests were concordantly positive in 3 patients, while in 1 patient with a supraclinoid internal carotid artery aneurysm the finding was reported only on DSA and the CTA was negative. Further, only in 2 patients with positive findings for SAH, both repeat, follow-up DSA and CTA were performed after discharge and both of these tests identified the same etiology explaining the SAH (Table 2). NPV of Imaging Tests for Management of AN-SAH NPVs for follow-up imaging were as follows: brain MRI, 97.5% (95% confidence interval [CI], 95.8%-98.6%); cervical spine MRI, 96.9% (95% CI, 94.9%-98.3%); repeat, same-admission follow-up DSA, 99.7% (95% CI, 98.4%-100%); repeat, same-admission follow-up CTA, 98.7% (95% CI, 96.9%-99.6%); delayed postdischarge DSA, 100% (95% CI, 98.1%-100%); and delayed postdischarge CTA, 100% (95% CI, 93%-100%). When comparing the NPV of follow-up nonvascular MR imaging (eg, patient undergoing either brain or cervical spine MRI; 97.7% [95% CI, 96.5%-98.9%]) with that of vascular same-admission follow-up imaging (eg, DSA or CTA; 99.3% [95% CI, 98.7%-100%]), the vascular same-admission imaging has a higher NPV (P = .03). Similarly, vascular delayed after discharge imaging has a higher NPV (100% [95% CI, 100%-100%]) compared to nonvascular imaging (P = .01); but is not significantly different from vascular same-admission imaging (P = .58). Patients’ Disposition A total of 92% (694/752) of patients were discharged home, while 6% (47/752) were sent to rehabilitation facilities and 2% (11/752) died in the hospital. Of 47 patients sent to rehabilitation facilities, 4 had positive findings on imaging that would explain the etiology for SAH. Of 11 patients who died, all had nonperimesencephalic pattern of hemorrhage, and none had positive findings on imaging to explain SAH. DISCUSSION The results of this retrospective multicenter study of 752 AN-SAH show extremely low diagnostic yield of cervical spine MRI (0.2%) and brain MRI (0.7%) for the management of AN-SAH, both commonly performed in AN-SAH patients. This diagnostic yield remained low (0.3% for cervical spine MRI and 1.2% for brain MRI) in patients with nonperimesencephalic pattern of hemorrhage. Repeat same-admission follow-up CTA and DSA performed in nearly half the patients and delayed after-discharge follow-up DSA and CTA performed in one-quarter and one-tenth of the patients, respectively. All have a slightly higher diagnostic yield than cervical spine or brain MRI, ranging between 1% and 4% regardless of pattern of hemorrhage. In patients with perimesencephalic hemorrhage, none of the follow-up imaging was positive for any etiology to explain SAH, while in those with nonperimsencephalic hemorrhage the diagnostic yield for follow-up DSA and CTA ranged between 1% and 8%. The NPV for all the follow-up tests was more than 96% with the vascular imaging tests having higher NPV compared to nonvascular imaging tests. Our multicenter study has several strengths. First, this is the largest available study assessing the diagnostic management of patients with AN-SAH. Second, with the availability of long-term angiographic follow-up (ranging from 1 to 7 yr) for patients with AN-SAH, we were able to determine with higher confidence whether imaging findings were clinically relevant to explain the etiology of SAH. The results from our large sample size support those from prior, smaller studies. Recent meta-analysis of literature reported a diagnostic yield of 1.3% (95% CI, 0.5%-2.5%) for cervical spine MRI in 538 patients with AN-SAH,22 with individual published studies reporting yields ranging from 0% to 4%.5,9,16,23-26 For brain MRI, a majority of studies reported a diagnostic yield close to 0%,6,16,27-29 with only a few older studies published prior to the year 2000 reporting higher diagnostic yields of 7% to11%.9,30 Nearly half of the etiologies identified on repeat angiography (DSA or CTA) in the current study were aneurysms. In fact, the most common cause of SAH is aneurysmal rupture, which may be missed upon initial angiographic evaluation due, for instance, to copious cisternal blood obscuring the aneurysm, vasospasm proximal and distal to the aneurysm preventing sufficient opacification, thrombosis of the aneurysm, small aneurysms in proximity to the calvaria on CTA, or simply due to a technically inadequate examination.31-33 Reported diagnostic yield of a second, repeat angiography (DSA or CTA) in AN-SAH varies between 0% and 30% in prior studies.17,34-37 In patients with nonperimesencephalic pattern of SAH, prior meta-analysis showed a diagnostic yield of 10% (95% CI, 7.4%-13.6%) for repeat DSA.38 In patients with perimesencephalic patterns of SAH, meta-analyses have reported diagnostic rates of 1.6% (95% CI, 0.7%-3.8%) for repeat, same-admission follow-up DSA39 and 0.78% (95% CI, 23%-1.32%) for repeat same-admission or after-discharge follow-up angiography (DSA or CTA).40 Our study findings are consistent with prior studies and demonstrate a diagnostic yield between 1% and 4% for repeat DSA or CTA regardless of pattern of hemorrhage. However, in patients with nonperimsencephalic pattern of hemorrhage, the diagnostic yields of repeat same admission follow-up DSA and CTA were 7.4% and 1.7%, respectively. In AN-SAH, there is high mortality associated with re-bleeding (up to 4% during the first 24 h and up to 15%-25% during the first 14 d, and even 50% in 6 mo).37,41,42 Further, patients are at risk for delayed cerebral ischemia (up to 4%), and hydrocephalus (up to 14%), which may require shunt placement.43 The risk of re-bleeding and consequent morbidity and mortality associated with an undiscovered source of SAH together with the low morbidity of angiography and MRI, tilts the risk-benefit balance for many clinicians toward ordering these tests. However, the cost-effectiveness of these tests given their low diagnostic yield is questionable and the focus of ongoing research. In the current study, the diagnostic yield of ∼4% for repeat angiography may support the cost-effectiveness of its utilization specifically in nonperimsencephalic pattern of hemorrhage (diagnostic yield of ∼8%) as all aneurysmal causes of patients initially diagnosed with AN-SAH were identified on repeat angiography in the present study. This is especially important if vasospasm is present on initial angiographic evaluation because detection of saccular aneurysms may prove difficult in this setting.17,25 Our results are limited in investigating the effectiveness of repeat DSA vs CTA due to small number of patients undergoing both tests in the current study. However, our results support the very low diagnostic yield of brain and cervical spine MRI. In addition, all except for 1 of the etiologies identified by MRI in the current study were nonaneurysmal (eg, vasculopathy, and vascular malformation) and did not include potentially immediately life-threatening conditions in contrast to the aneurysmal etiologies, diagnosed on follow-up angiography. The 1 aneurysm identified on brain MRI was also diagnosed on repeat, same-admission DSA and CTA. Furthermore, brain and cervical spine MRI likely add diagnostic value in only a very small proportion of patients, while adding to the financial burden of remaining patients with normal exams.22 With one-third of Americans reporting difficulty paying healthcare bills, many regularly face decisions about whether the benefits of recommended services justify associated costs,44,45 and in case of AN-SAH patients, brain and cervical spine MRI may represent examples of these services. Limitations We acknowledge several study limitations, including those inherent to our retrospective study design and the potential for selection bias, which is an issue in any retrospective analysis. The imaging analyses were performed by investigators from each center as opposed to centralized investigators reviewing all imaging interpretations, given the IRB approval at each site did not allow sharing the images or their reports with investigators from other sites. CONCLUSION In summary, there are variations in utilization of imaging tests for management of AN-SAH with brain and cervical spine MRI performed in 75% and 61% of cases, while having a very low diagnostic yield (less than 1% regardless of pattern of hemorrhage). Even in patients with nonperimesencephalic hemorrhage the diagnostic yields for cervical spine and brain MRI were less than 1% and 1.2%, respectively. On the other hand, repeat and delayed follow-up DSA and CTA have more reasonable diagnostic yield, but are performed in less than half of patients. As a result, we suggest that MRI for possible etiologies of AN-SAH is not routinely recommended unless there are focal neurological deficits to suggest a cranial or spinal etiology. Future studies to evaluate cost-effectiveness of MRI examinations as well as comparative effectiveness of follow-up CTA versus DSA are recommended. Disclosure The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article. Notes The abstract of this manuscript was presented as an oral presentation at the ASNR annual meeting, April 24-27, 2017 in Long Beach, California. REFERENCES 1. Bederson JB, Connolly ES Jr, Batjer HH et al.   Guidelines for the management of aneurysmal subarachnoid hemorrhage: a statement for healthcare professionals from a special writing group of the Stroke Council, American Heart Association. Stroke . 2009; 40( 3): 994- 1025. Google Scholar CrossRef Search ADS PubMed  2. Feigin VL, Lawes CM, Bennett DA, Anderson CS. Stroke epidemiology: a review of population-based studies of incidence, prevalence, and case-fatality in the late 20th century. Lancet Neurol . 2003; 2( 1): 43- 53. Google Scholar CrossRef Search ADS PubMed  3. Hackett ML, Anderson CS. Health outcomes 1 year after subarachnoid hemorrhage: An international population-based study. Neurology . 2000; 55( 5): 658- 662. Google Scholar CrossRef Search ADS PubMed  4. Mayer SA, Kreiter KT, Copeland D et al.   Global and domain-specific cognitive impairment and outcome after subarachnoid hemorrhage. Neurology . 2002; 59( 11): 1750- 1758. Google Scholar CrossRef Search ADS PubMed  5. Germans MR, Coert BA, Majoie CB et al.   Yield of spinal imaging in nonaneurysmal, nonperimesencephalic subarachnoid hemorrhage. Neurology . 2015; 84( 13): 1337- 1340. Google Scholar CrossRef Search ADS PubMed  6. Andaluz N, Zuccarello M. Yield of further diagnostic work-up of cryptogenic subarachnoid hemorrhage based on bleeding patterns on computed tomographic scans. Neurosurgery . 2008; 62( 5): 1040- 1047. Google Scholar CrossRef Search ADS PubMed  7. Kumar R, Das KK, Sahu RK et al.   Angio negative spontaneous subarachnoid hemorrhage: is repeat angiogram required in all cases? Surg Neurol Int . 2014; 5( 1): 125. Google Scholar CrossRef Search ADS PubMed  8. Connolly ES Jr, Rabinstein AA, Carhuapoma JR et al.   Guidelines for the management of aneurysmal subarachnoid hemorrhage: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke . 2012; 43( 6): 1711- 1737. Google Scholar CrossRef Search ADS PubMed  9. Rogg JM, Smeaton S, Doberstein C, Goldstein JH, Tung GA, Haas RA. Assessment of the value of MR imaging for examining patients with angiographically negative subarachnoid hemorrhage. Am J Roentgenol . 1999; 172( 1): 201- 206. Google Scholar CrossRef Search ADS   10. Gupta SK, Gupta R, Khosla VK et al.   Nonaneurysmal nonperimesencephalic subarachnoid hemorrhage: is it a benign entity? Surg Neurol . 2009; 71( 5): 566- 571. Google Scholar CrossRef Search ADS PubMed  11. Kaim A, Proske M, Kirsch E, von Weymarn A, Radu EW, Steinbrich W. Value of repeat-angiography in cases of unexplained subarachnoid hemorrhage (SAH). Acta Neurol Scand . 1996; 93( 5): 366- 373. Google Scholar CrossRef Search ADS PubMed  12. Yu DW, Jung YJ, Choi BY, Chang CH. Subarachnoid hemorrhage with negative baseline digital subtraction angiography: is repeat digital subtraction angiography necessary? J Cerebrovasc Endovasc Neurosurg . 2012; 14( 3): 210- 215. Google Scholar CrossRef Search ADS PubMed  13. Sadigh G, Allen J. National survey of radiologic management of angiographically-negative subarachnoid hemorrhage. ASNR Annual Meeting  2017. 14. Rinkel GJ, Wijdicks EF, Vermeulen M et al.   Nonaneurysmal perimesencephalic subarachnoid hemorrhage: CT and MR patterns that differ from aneurysmal rupture. AJNR Am J Neuroradiol . 1991; 12( 5): 829- 834. Google Scholar PubMed  15. Fontanella M, Rainero I, Panciani PP et al.   Subarachnoid hemorrhage and negative angiography: clinical course and long-term follow-up. Neurosurg Rev . 2011; 34( 4): 477- 484. Google Scholar CrossRef Search ADS PubMed  16. Akcakaya MO, Aydoseli A, Aras Y et al.   Clinical course of nontraumatic nonaneurysmal subarachnoid hemorrhage: A single institution experience over 10 years and review of the contemporary literature. Turk Neurosurg . 2016; 27( 5): 732- 742. 17. Canneti B, Mosqueira AJ, Nombela F, Gilo F, Vivancos J. Spontaneous subarachnoid hemorrhage with negative angiography managed in a stroke unit: Clinical and prognostic characteristics. J Stroke Cerebrovasc Dis . 2015; 24( 11): 2484- 2490. Google Scholar CrossRef Search ADS PubMed  18. Rinkel GJ, Wijdicks EF, Hasan D et al.   Outcome in patients with subarachnoid haemorrhage and negative angiography according to pattern of haemorrhage on computed tomography. Lancet . 1991; 338( 8773): 964- 968. Google Scholar CrossRef Search ADS PubMed  19. Rinkel GJ, van Gijn J, Wijdicks EF. Subarachnoid hemorrhage without detectable aneurysm. A review of the causes. Stroke . 1993; 24( 9): 1403- 1409. Google Scholar CrossRef Search ADS PubMed  20. Hunt WE, Hess RM. Surgical risk as related to time of intervention in the repair of intracranial aneurysms. J Neurosurg . 1968; 28( 1): 14- 20. Google Scholar CrossRef Search ADS PubMed  21. Fisher CM, Kistler JP, Davis JM. Relation of cerebral vasospasm to subarachnoid hemorrhage visualized by computerized tomographic scanning. Neurosurgery . 1980; 6( 1): 1- 9. Google Scholar CrossRef Search ADS PubMed  22. Sadigh G, Holder CA, Switchenko JM, Dehkharghani S, Allen JW. Is there added value in obtaining cervical spine MRI in the assessment of nontraumatic angiographically negative subarachnoid hemorrhage? A retrospective study and meta-analysis of the literature. J Neurosurg . 2017: 1- 7. 23. Woodfield J, Rane N, Cudlip S, Byrne JV. Value of delayed MRI in angiogram-negative subarachnoid haemorrhage. Clin Radiol . 2014; 69( 4): 350- 356. Google Scholar CrossRef Search ADS PubMed  24. Little AS, Garrett M, Germain R et al.   Evaluation of patients with spontaneous subarachnoid hemorrhage and negative angiography. Neurosurgery . 2007; 61( 6): 1139- 1151. Google Scholar CrossRef Search ADS PubMed  25. Topcuoglu MA, Ogilvy CS, Carter BS, Buonanno FS, Koroshetz WJ, Singhal AB. Subarachnoid hemorrhage without evident cause on initial angiography studies: Diagnostic yield of subsequent angiography and other neuroimaging tests. J Neurosurg . 2003; 98( 6): 1235- 1240. Google Scholar CrossRef Search ADS PubMed  26. Lin N, Zenonos G, Kim AH et al.   Angiogram-negative subarachnoid hemorrhage: relationship between bleeding pattern and clinical outcome. Neurocrit Care . 2012; 16( 3): 389- 398. Google Scholar CrossRef Search ADS PubMed  27. Maslehaty H, Petridis AK, Barth H, Mehdorn HM. Diagnostic value of magnetic resonance imaging in perimesencephalic and nonperimesencephalic subarachnoid hemorrhage of unknown origin. J Neurosurg . 2011; 114( 4): 1003- 1007. Google Scholar CrossRef Search ADS PubMed  28. Wijdicks EF, Schievink WI, Miller GM. MR imaging in pretruncal nonaneurysmal subarachnoid hemorrhage: Is it worthwhile? Stroke . 1998; 29( 12): 2514- 2516. Google Scholar CrossRef Search ADS PubMed  29. Maslehaty H, Barth H, Petridis AK, Doukas A, Maximilian Mehdorn H. Special features of subarachnoid hemorrhage of unknown origin: a review of a series of 179 cases. Neurol Res . 2012; 34( 1): 91- 97. Google Scholar CrossRef Search ADS PubMed  30. Renowden SA, Molyneux AJ, Anslow P, Byrne JV. The value of MRI in angiogram-negative intracranial haemorrhage. Neuroradiology . 1994; 36( 6): 422- 425. Google Scholar CrossRef Search ADS PubMed  31. Broderick JP, Brott TG, Duldner JE, Tomsick T, Leach A. Initial and recurrent bleeding are the major causes of death following subarachnoid hemorrhage. Stroke . 1994; 25( 7): 1342- 1347. Google Scholar CrossRef Search ADS PubMed  32. McMahon J, Dorsch N. Subarachnoid haemorrhage of unknown aetiology: What next? Crit Rev Neurosurg . 1999; 9( 3): 147- 155. Google Scholar CrossRef Search ADS PubMed  33. van Gijn J, Rinkel GJ. Subarachnoid haemorrhage: diagnosis, causes and management. Brain . 2001; 124( 2): 249- 278. Google Scholar CrossRef Search ADS PubMed  34. Nishioka H, Torner JC, Graf CJ, Kassell NF, Sahs AL, Goettler LC. Cooperative study of intracranial aneurysms and subarachnoid hemorrhage: A long-term prognostic study. Arch Neurol . 1984; 41( 11): 1147- 1151. Google Scholar CrossRef Search ADS PubMed  35. Juul R, Fredriksen TA, Ringkjob R. Prognosis in subarachnoid hemorrhage of unknown etiology. J Neurosurg . 1986; 64( 3): 359- 362. Google Scholar CrossRef Search ADS PubMed  36. Pathirana N, Refsum SE, McKinstry CS, Bell KE. The value of repeat cerebral angiography in subarachnoid haemorrhage. Br J Neurosurg . 1994; 8( 2): 141- 146. Google Scholar CrossRef Search ADS PubMed  37. Iwanaga H, Wakai S, Ochiai C, Narita J, Inoh S, Nagai M. Ruptured cerebral aneurysms missed by initial angiographic study. Neurosurgery . 1990; 27( 1): 45- 51. Google Scholar CrossRef Search ADS PubMed  38. Bakker NA, Groen RJ, Foumani M et al.   Repeat digital subtraction angiography after a negative baseline assessment in nonperimesencephalic subarachnoid hemorrhage: a pooled data meta-analysis. J Neurosurg . 2014; 120( 1): 99- 103. Google Scholar CrossRef Search ADS PubMed  39. Potter CA, Fink KR, Ginn AL, Haynor DR. Perimesencephalic hemorrhage: Yield of Single versus multiple DSA examinations-a single-center study and meta-analysis. Radiology . 2016; 281( 3): 858- 864. Google Scholar CrossRef Search ADS PubMed  40. Kalra VB, Wu X, Matouk CC, Malhotra A. Use of follow-up imaging in isolated perimesencephalic subarachnoid hemorrhage: a meta-analysis. Stroke . 2015; 46( 2): 401- 406. Google Scholar CrossRef Search ADS PubMed  41. Inamasu J, Nakamura Y, Saito R et al.   “Occult” ruptured cerebral aneurysms revealed by repeat angiography: result from a large retrospective study. Clin Neurol Neurosurg . 2003; 106( 1): 33- 37. Google Scholar CrossRef Search ADS PubMed  42. Vaitkevicius G, Gvazdaitis AR, Lukosevicius S. [Spontaneous subarachnoid hemorrhage: patients' examination after aneurysm-negative initial angiograms]. Medicina (Kaunas) . 2002; 38( 10): 976- 981. Google Scholar PubMed  43. Elhadi AM, Zabramski JM, Almefty KK et al.   Spontaneous subarachnoid hemorrhage of unknown origin: hospital course and long-term clinical and angiographic follow-up. J Neurosurg . 2015; 122( 3): 663- 670. Google Scholar CrossRef Search ADS PubMed  44. Obermeyer Z, Makar M, Abujaber S, Dominici F, Block S, Cutler DM. Association between the medicare hospice benefit and health care utilization and costs for patients with poor-prognosis cancer. JAMA . 2014; 312( 18): 1888- 1896. Google Scholar CrossRef Search ADS PubMed  45. Sadigh G, Carlos RC, Krupinski EA, Meltzer CC, Duszak R Jr. Health care price transparency and communication: Implications for radiologists and patients in an era of expanding shared decision making. Am J Roentgenol . 2017; 209( 5): 959- 964. Google Scholar CrossRef Search ADS   COMMENT The authors report on the radiologic management of angiographically negative, spontaneous intracranial subarachnoid hemorrhage as part of a multicenter study of utilization and diagnostic yield. This is a retrospective multicenter study, of consecutive adult patients admitted with none Traumatic, AN-SAH. Outcomes studied included utilization rate, diagnostic yield, and median time from admission for the following imaging tests: initial CTA and DSA, brain and cervical spine MRI, and any repeat DSA or CTA performed either during initial admission or at long-term follow-up. In the 752 patients who were included (mean age, 53 years; 54% male), Initial CTA and DSA were performed in 89% and 100% of patients, respectively. Brain MRI was performed in 75% of patients and was positive in 0.7% of cases. Cervical spine MRI was performed in 61% of patients and was positive in 0.2% of cases. Repeat, same-admission follow-up DSA and CTA were performed in 48% and 51% of patients and were positive in 3.3% and 1% of cases, respectively. Delayed follow-up DSA and CTA after discharge were performed in 26% and 7% of patients and were positive in 2% and 3.7% of cases, respectively, all with negative prior imaging studies. The authors conclude that cervical spine and brain MRI have extremely low diagnostic yield whereas repeat DSA and CTA are utilized less commonly but have slightly higher diagnostic yield. A large amount of data is already available in the management of non-traumatic SAH. The biggest problem with this study in addition to its retrospective study design, is that the pattern of hemorrhage is so important in the type of tests ordered and the expected negative and positive results. Much like routine administrative data studies, this study lacks the details required to truly establish and management paradigm. A patient with a SAH in the lower region of the posterior fossa does require and spinal MRI/A in the face of a negative angiogram versus a sylvan fissure SAH does not. Gavin W. Britz Houston, Texas Copyright © 2018 by the Congress of Neurological Surgeons This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Neurosurgery Oxford University Press

Radiological Management of Angiographically Negative, Spontaneous Intracranial Subarachnoid Hemorrhage: A Multicenter Study of Utilization and Diagnostic Yield

Loading next page...
 
/lp/ou_press/radiological-management-of-angiographically-negative-spontaneous-jdft1x4uSk
Publisher
Congress of Neurological Surgeons
Copyright
Copyright © 2018 by the Congress of Neurological Surgeons
ISSN
0148-396X
eISSN
1524-4040
D.O.I.
10.1093/neuros/nyy225
Publisher site
See Article on Publisher Site

Abstract

Abstract BACKGROUND The optimal diagnostic evaluation for patients with angiographically negative subarachnoid hemorrhage (AN-SAH) remains controversial. OBJECTIVE To assess the utilization rate and diagnostic yield of imaging tests routinely obtained in identifying a structural cause for AN-SAH. METHODS In this retrospective multicenter study, consecutive adult patients admitted with nontraumatic, AN-SAH between 01/2010 and 12/2015 were included. Patients with intraparenchymal, subdural, or epidural hematomas in addition to SAH were excluded. Outcomes studied included utilization rate, diagnostic yield, and median time from admission for the following imaging tests: initial computed tomography angiography (CTA) and digital subtraction angiography (DSA), brain and cervical spine magnetic resonance imaging (MRI), and any repeat DSA or CTA performed either during initial admission or at long-term follow-up. RESULTS A total of 752 patients were included (mean age, 53 yr; 54% male). Initial CTA and DSA were performed in 89% and 100% of patients, respectively. Brain MRI was performed in 75% of patients and was positive in 0.7% of cases. Cervical spine MRI was performed in 61% of patients and was positive in 0.2% of cases. Repeat, same-admission follow-up DSA and CTA were performed in 48% and 51% of patients and were positive in 3.3% and 1% of cases, respectively. Delayed follow-up DSA and CTA after discharge were performed in 26% and 7% of patients and were positive in 2% and 3.7% of cases, respectively, all with negative prior imaging studies. CONCLUSION Cervical spine and brain MRI have extremely low diagnostic yield, both are commonly utilized in patients with AN-SAH; while repeat DSA and CTA are utilized less commonly and have slightly higher diagnostic yield. Angiographically negative subarachnoid hemorrhage, Management, Imaging test, Utilization, Diagnostic yield ABBREVIATIONS ABBREVIATIONS AN-SAH angiographically negative subarachnoid hemorrhage CI confidence interval CTA computed tomography angiography DSA digital subtraction angiography EMR electronic medical record IQR median and interquartile ranges IRB institutional review board MRI magnetic resonance imaging NPV negative predictive value SD standard deviations Nontraumatic subarachnoid hemorrhage (SAH) occurs in 30 000 patients per year in the United States1 and accounts for 7% of all strokes2. Nearly half of the SAH survivors may have long-term cognitive impairment due to symptomatic vasospasm, which impacts functional status and quality of life.3,4 In 15% to 20% of SAH patients, angiographic evaluation fails to identify an etiology for the hemorrhage, also known as angiographically negative SAH (AN-SAH).5,6 These patients are at risk for rehemorrhage and its consequences such as vasospasm, if structural lesions are missed on the initial angiogram and treatment is delayed.7 Therefore, early diagnosis of the hemorrhage etiology and aggressive management of SAH is highly recommended by current guidelines.8 The optimal diagnostic evaluation for patients with AN-SAH remains controversial.5,9 Many researchers have suggested that a repeat angiogram may detect etiologies missed on initial angiographic imaging, although the reported diagnostic yield varies considerably.10-12 Currently, the diagnostic protocol for management of AN-SAH varies between institutions and may include a combination of initial and repeat computed tomography angiography (CTA) and/or digital subtraction angiography (DSA) with brain and/or cervical spine magnetic resonance imaging (MRI). However, there are limited data available on the diagnostic yield of each of these imaging tests and these are derived from published studies with small sample sizes.13 Patients with perimesencephalic pattern of SAH14,15 are defined as those in whom center of hemorrhage is anterior to midbrain with or without extension of blood to the anterior part of the ambient cistern or the basal part of the sylvian fissure, incomplete filling of the anterior interhemispheric fissure and no extension of the lateral sylvian fissure or the ventricles.14 These patients less commonly develop hydrocephalus, vasospasm, or other complications and have lower risk of recurrent bleeding.14,16-18 The better overall prognosis in such cases has prompted development of less aggressive management algorithms in benign perimesencephalic pattern of hemorrhage, such as foregoing secondary angiography.19 However, in patients with nonperimesencephalic pattern of AN-SAH, the diagnostic and therapeutic approach is less well defined.17 The purpose of the current study was to assess the utilization rate and diagnostic yield of imaging tests routinely obtained in identifying a structural cause for AN-SAH in a multicenter study. METHODS Institutional Review Board (IRB) approval and a waiver of informed consent were obtained for this retrospective multicenter review. The study was Health Insurance Probability and Accountability Act compliant. Study Population A multicenter radiology informatics system search was conducted for consecutive patients aged 18 yr or older with acute, nontraumatic, intracranial SAH, diagnosed based on noncontrast head CT or xanthochromia on lumbar puncture. A total of 7 university hospitals from 5 academic institutions in the United States and 1 academic institution in the United Kingdom were included in a study period spanning January 2010 to December 2015 in 4 of the participating institutions. In 1 institution, the study period was longer, spanning January 2001 to December 2015. Only patients with a negative initial cerebrovascular DSA were included, regardless of the type of follow-up imaging they received. Patients with misdiagnosis of SAH (eg, due to traumatic spinal tap) were excluded. Further exclusion criteria included patients with acute intraparenchymal, subdural, or epidural hemorrhage; cerebral contusion; recent trauma; surgery to the brain or spine within last 2 wk of presentation; and patients with known etiology for SAH established on CTA or DSA from prior admissions. The identification of eligible patients for the purpose of the study was performed between July 2016 and April 2017. Medical Record Review Demographic information including age at presentation and sex, past medical history of hypertension, diabetes (type 1 or 2), hyperlipidemia, history of any lifetime tobacco use, and Hunt and Hess grade20 at presentation were obtained from the electronic medical record (EMR). We further documented final disposition status: discharge home or to a rehabilitation facility, or death during initial hospitalization. Imaging Analysis The initial noncontrast head CT scan that was obtained within 24 h of admission and prior to a brain MRI was reviewed by investigators from each center, and the perimesencephalic vs nonperimesencephalic pattern of hemorrhage according to Rinkel et al14 was assessed and a Fisher grade21 was calculated. If a head CT was not obtained until after the brain MRI was obtained, or if no head CT was available, pattern of hemorrhage or Fisher grade was not assessed to mitigate potential errors related to redistribution of hemorrhage during the course of admission. The postadmission imaging tests included initial DSA and CTA, brain and cervical spine MRI, repeat DSA and CTA performed during the initial admission and delayed DSA and CTA performed at long-term follow-up. For each included patient, subsequent neurological imaging tests obtained later in admission and after discharge were recorded, and the timing of such imaging logged relative to the time of admission. Formal interpretations were reviewed for any findings potentially explaining initial SAH. For cases in which vague or uncertain potential etiologies were identified, the EMR was further reviewed for follow-up neurology, neurosurgery, and neurointerventional notes, and the findings included only if the hemorrhage was ascribable to the findings in question on the basis of aggregated available clinical information. Statistical Analysis Categorical variables were reported as frequencies and percentages. Numerical variables were presented as mean and standard deviations (SD) or median and interquartile ranges (IQR) if they did not follow a normal distribution. Utilization rate was assessed as a proportion of patients with AN-SAH who received the imaging test. Diagnostic yield was calculated as a proportion of patients with imaging test who had an identifiable etiology for SAH. Negative predictive value (NPV) was calculated for each of the imaging tests separately and compared between nonvascular imaging tests (eg, patient undergoing either brain or cervical spine MRI) and vascular imaging tests (eg, patient undergoing either DSA or CTA) using Fisher's Exact Test, given not all the patients had received all the imaging modalities at follow-up. For nonvascular vs vascular imaging comparison, imaging was considered negative only if both imaging tests in each group (eg, both brain and cervical spine MRI for nonvascular imaging, and both CTA and DSA for vascular imaging) were negative. The analyses were performed using STATA 10 (Stata Corp., College Station, Texas). P values < .05 were considered statistically significant. RESULTS Study Population A total of 752 patients with AN-SAH (406/752 men, age 18-88 yr old [mean 53 yr]) were included in the analyses, with baseline characteristics shown in Table 1. Half of the included patients presented with Hunt and Hess grade 1 (50%, 373/752). Of included patients, 29% (220/752) had perimesencephalic pattern of hemorrhage; while 51% (384/752) had nonperimsencephalic hemorrhage and the remaining 20% (148/752) either did not have CT scan on admission or had no hemorrhage on CT scan (eg, Fisher grade I). TABLE 1. Baseline Characteristics for 752 patients With Angiographically Negative SAH Mean age, yr ± SD  53 ± 14  Gender, % (n)     Male  54 (406)   Female  46 (346)  Past Medical History, % (n)     Hypertension  32 (241)   Diabetes Type I  2 (18)   Diabetes Type II  8 (61)   Hyperlipidemia  13 (99)  Any Life-time Tobacco Use, % (n)  32 (242)  Hunt and Hess Grade,20 % (n)    Not reported  15 (113)   1  50 (373)   2  16 (124)   3  14 (106)   4  4 (31)   5  1 (5)  Fisher Grade based on admission noncontrast Head CT, % (n)    No admission CT scan available or unknown  10 (72)   1  15 (114)   2  21 (159)   3  29 (220)   4  25 (186)  Pattern of Hemorrhage on admission noncontrast Head CT, % (n)    No admission CT scan available or Fisher grade I (no SAH on CT scan)  20 (148)  Perimesencephalic  29 (220)  Nonperimesencephalic  51 (384)  Mean age, yr ± SD  53 ± 14  Gender, % (n)     Male  54 (406)   Female  46 (346)  Past Medical History, % (n)     Hypertension  32 (241)   Diabetes Type I  2 (18)   Diabetes Type II  8 (61)   Hyperlipidemia  13 (99)  Any Life-time Tobacco Use, % (n)  32 (242)  Hunt and Hess Grade,20 % (n)    Not reported  15 (113)   1  50 (373)   2  16 (124)   3  14 (106)   4  4 (31)   5  1 (5)  Fisher Grade based on admission noncontrast Head CT, % (n)    No admission CT scan available or unknown  10 (72)   1  15 (114)   2  21 (159)   3  29 (220)   4  25 (186)  Pattern of Hemorrhage on admission noncontrast Head CT, % (n)    No admission CT scan available or Fisher grade I (no SAH on CT scan)  20 (148)  Perimesencephalic  29 (220)  Nonperimesencephalic  51 (384)  View Large TABLE 1. Baseline Characteristics for 752 patients With Angiographically Negative SAH Mean age, yr ± SD  53 ± 14  Gender, % (n)     Male  54 (406)   Female  46 (346)  Past Medical History, % (n)     Hypertension  32 (241)   Diabetes Type I  2 (18)   Diabetes Type II  8 (61)   Hyperlipidemia  13 (99)  Any Life-time Tobacco Use, % (n)  32 (242)  Hunt and Hess Grade,20 % (n)    Not reported  15 (113)   1  50 (373)   2  16 (124)   3  14 (106)   4  4 (31)   5  1 (5)  Fisher Grade based on admission noncontrast Head CT, % (n)    No admission CT scan available or unknown  10 (72)   1  15 (114)   2  21 (159)   3  29 (220)   4  25 (186)  Pattern of Hemorrhage on admission noncontrast Head CT, % (n)    No admission CT scan available or Fisher grade I (no SAH on CT scan)  20 (148)  Perimesencephalic  29 (220)  Nonperimesencephalic  51 (384)  Mean age, yr ± SD  53 ± 14  Gender, % (n)     Male  54 (406)   Female  46 (346)  Past Medical History, % (n)     Hypertension  32 (241)   Diabetes Type I  2 (18)   Diabetes Type II  8 (61)   Hyperlipidemia  13 (99)  Any Life-time Tobacco Use, % (n)  32 (242)  Hunt and Hess Grade,20 % (n)    Not reported  15 (113)   1  50 (373)   2  16 (124)   3  14 (106)   4  4 (31)   5  1 (5)  Fisher Grade based on admission noncontrast Head CT, % (n)    No admission CT scan available or unknown  10 (72)   1  15 (114)   2  21 (159)   3  29 (220)   4  25 (186)  Pattern of Hemorrhage on admission noncontrast Head CT, % (n)    No admission CT scan available or Fisher grade I (no SAH on CT scan)  20 (148)  Perimesencephalic  29 (220)  Nonperimesencephalic  51 (384)  View Large Utilization Rate of Imaging Tests for Management of AN-SAH Initial postadmission DSA was performed for 100% (n = 752) of patients within 10 d of admission (median = 1 d postadmission, IQR = 1). Initial postadmission CTA was performed for 89% (668/752) of patients within 10 d of admission (median = 0 d, IQR = 1). Brain MRI and cervical spine MRI were performed in 75% (562/752) and 61% (457/752) of patients, respectively. The median time for both tests was 2 (IQR = 29) d after admission. Repeat, same-admission follow-up DSA and CTA were performed in 48% (362/752) and 51% (380/752) of patients, respectively. Only 16% (121/752) of patients received both DSA and CTA follow-up tests during the same admission and 17% (131/752) received neither of these 2 follow-up tests during the same admission. The median time for repeat, same-admission follow-up DSA was 10 (IQR = 7) d after admission and for CTA was 7 (IQR = 4) d after admission (Figure 1). FIGURE 1. View largeDownload slide Utilization rate of imaging tests for management of angiographically negative subarachnoid hemorrhage in 752 patients. C-spine, cervical spine; Re-DSA, repeat same admission follow-up DSA; Re-CTA, repeat same admission follow-up CTA; Del-DSA, delayed post discharge follow-up DSA; Del-CTA, delayed post discharge follow-up CTA. FIGURE 1. View largeDownload slide Utilization rate of imaging tests for management of angiographically negative subarachnoid hemorrhage in 752 patients. C-spine, cervical spine; Re-DSA, repeat same admission follow-up DSA; Re-CTA, repeat same admission follow-up CTA; Del-DSA, delayed post discharge follow-up DSA; Del-CTA, delayed post discharge follow-up CTA. Delayed follow-up DSA and CTA after discharge were performed in 26% (197/752) and 7% (53/752) of patients, respectively. Only 2% (14/752) of patients received both follow-up tests after discharge and 69% (516/752) received neither of these 2 follow-up tests after discharge. The median time for delayed postdischarge follow-up DSA was 90 (IQR = 40) d after admission and for CTA was 93 (IQR = 90) d after admission (Figure 1). Diagnostic Yield of Imaging Tests for Management of AN-SAH Overall, regardless of type of follow-up imaging, 2.7% (21/752) of patients had a positive finding to explain the etiology of SAH on at least one of the follow-up imaging tests. Of these 21 patients with positive findings, 19 had nonperimesencephalic pattern of hemorrhage and 2 did not have an admission head CT available for review due to being transferred from another hospital. However, per EMR the pattern of hemorrhage for both of these patients was reported as nonperimsencephalic at the outside hospital. In 1.6% of patients (12/752), an aneurysm considered to be the etiology of the SAH (Table 2). Remaining etiologies for SAH identified on follow-up imaging were as follows: reversible cerebral vasoconstriction syndrome (0.5% [4/752]), inflammatory vasculopathy (0.3% [2/752]), dural arteriovenous fistula in the posterior fossa (0.1% [1/752]), brain cavernous malformation (0.1% [1/752]), and cervical spine dural arterivenous fistula (0.1% [1/752]; Table 2; Figure 2). FIGURE 2. View largeDownload slide Axial CT scan in a 61-yr-old male showing subarachnoid hemorrhage at the level of foramen magnum (A) with intraventricular extension (B). CTA of the head did not show any etiology for subarachnoid hemorrhage. Sagittal T1 (C), T2 (D), and T1 postcontrast (E) images of cervical spine showed a cervical dural arteriovenous fistula, and this was later confirmed on dedicated cervical angiography. FIGURE 2. View largeDownload slide Axial CT scan in a 61-yr-old male showing subarachnoid hemorrhage at the level of foramen magnum (A) with intraventricular extension (B). CTA of the head did not show any etiology for subarachnoid hemorrhage. Sagittal T1 (C), T2 (D), and T1 postcontrast (E) images of cervical spine showed a cervical dural arteriovenous fistula, and this was later confirmed on dedicated cervical angiography. TABLE 2. Imaging Characteristics for 21 AN-SAH Patients With Positive Findings on Follow-up Imaging to Explain SAH Patient  Pattern of Hemorrhage  Etiology  Brain MRI  C-Spine MRI  Re-DSA  Re-CTA  Del-DSA  Del-CTA  Aneurysmal etiologies                  42 yo F  Nonperimesencephalic  Supraclinoid ICA  +  ND  +  +  ND  ND  69 yo M  Nonperimesencephalic  Supraclinoid ICA  −  ND  +  −  ND  ND  48 yo F  Nonperimesencephalic  ACOM  −  −  +  +  ND  ND  43 yo F  Nonperimesencephalic  ACOM  −  −  ND  −  +  ND  47 yo F  Nonperimesencephalic  ACOM  −  −  ND  −  +  ND  62 yo F  Nonperimesencephalic  ACA  −  −  +  ND  ND  ND  43 yo M  Nonperimesencephalic  ACA  ND  ND  +  ND  ND  ND  74 yo M  Nonperimesencephalic  Paraophthalmic  ND  ND  +  ND  ND  ND  60 yo F  Nonperimesencephalic  PCA  −  −  +  ND  ND  ND  42 yo F  No head CT*  PCA  −  −  −  ND  +  +  52 yo F  No head CT*  PICA  −  −  ND  +  ND  ND  53 yo F  Nonperimesencephalic  Micronaneurysms in posterior circulation  −  −  ND  −  +  +  Nonaneurymal etiologies  56 yo F  Nonperimesencephalic  RCVS  −  −  +  ND  ND  ND  62 yo F  Nonperimesencephalic  RCVS  −  −  +  +  ND  ND  61 yo F  Nonperimesencephalic  RCVS  ND  ND  +  ND  ND  ND  59 yo F  Nonperimesencephalic  RCVS  −  −  +  ND  ND  ND  47 yo F  Nonperimesencephalic  Vasculopathy  +  −  ND  ND  ND  ND  77 yo M  Nonperimesencephalic  Vasculopathy  +  ND  ND  ND  ND  ND  49 yo M  Nonperimesencephalic  Dural AVF in posterior fossa  −  −  +  ND  ND  ND  31 yo F  Nonperimesencephalic  Brain cavernous malformation  +  −  ND  ND  ND  ND  61 yo M  Nonperimesencephalic  C-spine dural AVF  −  +  ND  ND  ND  ND  Patient  Pattern of Hemorrhage  Etiology  Brain MRI  C-Spine MRI  Re-DSA  Re-CTA  Del-DSA  Del-CTA  Aneurysmal etiologies                  42 yo F  Nonperimesencephalic  Supraclinoid ICA  +  ND  +  +  ND  ND  69 yo M  Nonperimesencephalic  Supraclinoid ICA  −  ND  +  −  ND  ND  48 yo F  Nonperimesencephalic  ACOM  −  −  +  +  ND  ND  43 yo F  Nonperimesencephalic  ACOM  −  −  ND  −  +  ND  47 yo F  Nonperimesencephalic  ACOM  −  −  ND  −  +  ND  62 yo F  Nonperimesencephalic  ACA  −  −  +  ND  ND  ND  43 yo M  Nonperimesencephalic  ACA  ND  ND  +  ND  ND  ND  74 yo M  Nonperimesencephalic  Paraophthalmic  ND  ND  +  ND  ND  ND  60 yo F  Nonperimesencephalic  PCA  −  −  +  ND  ND  ND  42 yo F  No head CT*  PCA  −  −  −  ND  +  +  52 yo F  No head CT*  PICA  −  −  ND  +  ND  ND  53 yo F  Nonperimesencephalic  Micronaneurysms in posterior circulation  −  −  ND  −  +  +  Nonaneurymal etiologies  56 yo F  Nonperimesencephalic  RCVS  −  −  +  ND  ND  ND  62 yo F  Nonperimesencephalic  RCVS  −  −  +  +  ND  ND  61 yo F  Nonperimesencephalic  RCVS  ND  ND  +  ND  ND  ND  59 yo F  Nonperimesencephalic  RCVS  −  −  +  ND  ND  ND  47 yo F  Nonperimesencephalic  Vasculopathy  +  −  ND  ND  ND  ND  77 yo M  Nonperimesencephalic  Vasculopathy  +  ND  ND  ND  ND  ND  49 yo M  Nonperimesencephalic  Dural AVF in posterior fossa  −  −  +  ND  ND  ND  31 yo F  Nonperimesencephalic  Brain cavernous malformation  +  −  ND  ND  ND  ND  61 yo M  Nonperimesencephalic  C-spine dural AVF  −  +  ND  ND  ND  ND  C-spine, cervical spine; Re-DSA, repeat same admission follow-up DSA; Re-CTA, repeat same admission follow-up CTA; Del-DSA, delayed postdischarge follow-up DSA; Del-CTA, delayed postdischarge follow-up CTA; F, female; M, Male; ICA, internal carotid artery; ACA, anterior cerebral artery; ACOM, anterior communicating artery; PCA, posterior cerebral artery; PICA, posterior inferior cerebellar artery; RCVS, reversible cerebral vasoconstriction syndrome; ND, not DONE. *These 2 patients did not have an admission head CT available due to being transferred from another hospital. However, per EMR the pattern of hemorrhage for both of these patients was reported as nonperimsencephalic at the outside hospital. View Large TABLE 2. Imaging Characteristics for 21 AN-SAH Patients With Positive Findings on Follow-up Imaging to Explain SAH Patient  Pattern of Hemorrhage  Etiology  Brain MRI  C-Spine MRI  Re-DSA  Re-CTA  Del-DSA  Del-CTA  Aneurysmal etiologies                  42 yo F  Nonperimesencephalic  Supraclinoid ICA  +  ND  +  +  ND  ND  69 yo M  Nonperimesencephalic  Supraclinoid ICA  −  ND  +  −  ND  ND  48 yo F  Nonperimesencephalic  ACOM  −  −  +  +  ND  ND  43 yo F  Nonperimesencephalic  ACOM  −  −  ND  −  +  ND  47 yo F  Nonperimesencephalic  ACOM  −  −  ND  −  +  ND  62 yo F  Nonperimesencephalic  ACA  −  −  +  ND  ND  ND  43 yo M  Nonperimesencephalic  ACA  ND  ND  +  ND  ND  ND  74 yo M  Nonperimesencephalic  Paraophthalmic  ND  ND  +  ND  ND  ND  60 yo F  Nonperimesencephalic  PCA  −  −  +  ND  ND  ND  42 yo F  No head CT*  PCA  −  −  −  ND  +  +  52 yo F  No head CT*  PICA  −  −  ND  +  ND  ND  53 yo F  Nonperimesencephalic  Micronaneurysms in posterior circulation  −  −  ND  −  +  +  Nonaneurymal etiologies  56 yo F  Nonperimesencephalic  RCVS  −  −  +  ND  ND  ND  62 yo F  Nonperimesencephalic  RCVS  −  −  +  +  ND  ND  61 yo F  Nonperimesencephalic  RCVS  ND  ND  +  ND  ND  ND  59 yo F  Nonperimesencephalic  RCVS  −  −  +  ND  ND  ND  47 yo F  Nonperimesencephalic  Vasculopathy  +  −  ND  ND  ND  ND  77 yo M  Nonperimesencephalic  Vasculopathy  +  ND  ND  ND  ND  ND  49 yo M  Nonperimesencephalic  Dural AVF in posterior fossa  −  −  +  ND  ND  ND  31 yo F  Nonperimesencephalic  Brain cavernous malformation  +  −  ND  ND  ND  ND  61 yo M  Nonperimesencephalic  C-spine dural AVF  −  +  ND  ND  ND  ND  Patient  Pattern of Hemorrhage  Etiology  Brain MRI  C-Spine MRI  Re-DSA  Re-CTA  Del-DSA  Del-CTA  Aneurysmal etiologies                  42 yo F  Nonperimesencephalic  Supraclinoid ICA  +  ND  +  +  ND  ND  69 yo M  Nonperimesencephalic  Supraclinoid ICA  −  ND  +  −  ND  ND  48 yo F  Nonperimesencephalic  ACOM  −  −  +  +  ND  ND  43 yo F  Nonperimesencephalic  ACOM  −  −  ND  −  +  ND  47 yo F  Nonperimesencephalic  ACOM  −  −  ND  −  +  ND  62 yo F  Nonperimesencephalic  ACA  −  −  +  ND  ND  ND  43 yo M  Nonperimesencephalic  ACA  ND  ND  +  ND  ND  ND  74 yo M  Nonperimesencephalic  Paraophthalmic  ND  ND  +  ND  ND  ND  60 yo F  Nonperimesencephalic  PCA  −  −  +  ND  ND  ND  42 yo F  No head CT*  PCA  −  −  −  ND  +  +  52 yo F  No head CT*  PICA  −  −  ND  +  ND  ND  53 yo F  Nonperimesencephalic  Micronaneurysms in posterior circulation  −  −  ND  −  +  +  Nonaneurymal etiologies  56 yo F  Nonperimesencephalic  RCVS  −  −  +  ND  ND  ND  62 yo F  Nonperimesencephalic  RCVS  −  −  +  +  ND  ND  61 yo F  Nonperimesencephalic  RCVS  ND  ND  +  ND  ND  ND  59 yo F  Nonperimesencephalic  RCVS  −  −  +  ND  ND  ND  47 yo F  Nonperimesencephalic  Vasculopathy  +  −  ND  ND  ND  ND  77 yo M  Nonperimesencephalic  Vasculopathy  +  ND  ND  ND  ND  ND  49 yo M  Nonperimesencephalic  Dural AVF in posterior fossa  −  −  +  ND  ND  ND  31 yo F  Nonperimesencephalic  Brain cavernous malformation  +  −  ND  ND  ND  ND  61 yo M  Nonperimesencephalic  C-spine dural AVF  −  +  ND  ND  ND  ND  C-spine, cervical spine; Re-DSA, repeat same admission follow-up DSA; Re-CTA, repeat same admission follow-up CTA; Del-DSA, delayed postdischarge follow-up DSA; Del-CTA, delayed postdischarge follow-up CTA; F, female; M, Male; ICA, internal carotid artery; ACA, anterior cerebral artery; ACOM, anterior communicating artery; PCA, posterior cerebral artery; PICA, posterior inferior cerebellar artery; RCVS, reversible cerebral vasoconstriction syndrome; ND, not DONE. *These 2 patients did not have an admission head CT available due to being transferred from another hospital. However, per EMR the pattern of hemorrhage for both of these patients was reported as nonperimsencephalic at the outside hospital. View Large Diagnostic yields for follow-up imaging are shown in Figure 3 and were as follows: brain MRI, 0.7% (4/562); cervical spine MRI, 0.2% (1/457); repeat, same-admission follow-up DSA, 3.3% (12/362); repeat, same-admission follow-up CTA, 1% (4/380); delayed postdischarge DSA, 2% (4/197); and delayed postdischarge CTA, 3.7% (2/53). FIGURE 3. View largeDownload slide Diagnostic yield of imaging tests for management of AN-SAH. The diagnostic yield is calculated as the proportion of patients undergoing each imaging test who have positive finding in the test to explain subarachnoid hemorrhage. Please note that the denominator is variable among different imaging tests as not all patients received all the tests. C-spine, cervical spine; Re-DSA, repeat same admission follow-up DSA; Re-CTA, repeat same admission follow-up CTA; Del-DSA, delayed postdischarge follow-up DSA; Del-CTA, delayed postdischarge follow-up CTA. FIGURE 3. View largeDownload slide Diagnostic yield of imaging tests for management of AN-SAH. The diagnostic yield is calculated as the proportion of patients undergoing each imaging test who have positive finding in the test to explain subarachnoid hemorrhage. Please note that the denominator is variable among different imaging tests as not all patients received all the tests. C-spine, cervical spine; Re-DSA, repeat same admission follow-up DSA; Re-CTA, repeat same admission follow-up CTA; Del-DSA, delayed postdischarge follow-up DSA; Del-CTA, delayed postdischarge follow-up CTA. Diagnostic yields for follow-up imaging in patients with nonperimesencephalic pattern of hemorrhage were as follows: brain MRI, 1.2% (4/308); cervical spine MRI, 0.3% (1/264); repeat, same-admission follow-up DSA, 7.4% (12/162); repeat, same-admission follow-up CTA, 1.7% (4/235); delayed postdischarge DSA, 3.1% (4/128); and delayed postdischarge CTA, 7.7% (2/26). Given the follow-up imaging was not positive in any of the patients with perimesencephalic pattern of hemorrhage, the diagnostic yield of all the follow-up imaging in these patients was 0%. Only in 4 patients with a positive finding for SAH, both repeat follow-up DSA and CTA were performed during the same admission. The 2 tests were concordantly positive in 3 patients, while in 1 patient with a supraclinoid internal carotid artery aneurysm the finding was reported only on DSA and the CTA was negative. Further, only in 2 patients with positive findings for SAH, both repeat, follow-up DSA and CTA were performed after discharge and both of these tests identified the same etiology explaining the SAH (Table 2). NPV of Imaging Tests for Management of AN-SAH NPVs for follow-up imaging were as follows: brain MRI, 97.5% (95% confidence interval [CI], 95.8%-98.6%); cervical spine MRI, 96.9% (95% CI, 94.9%-98.3%); repeat, same-admission follow-up DSA, 99.7% (95% CI, 98.4%-100%); repeat, same-admission follow-up CTA, 98.7% (95% CI, 96.9%-99.6%); delayed postdischarge DSA, 100% (95% CI, 98.1%-100%); and delayed postdischarge CTA, 100% (95% CI, 93%-100%). When comparing the NPV of follow-up nonvascular MR imaging (eg, patient undergoing either brain or cervical spine MRI; 97.7% [95% CI, 96.5%-98.9%]) with that of vascular same-admission follow-up imaging (eg, DSA or CTA; 99.3% [95% CI, 98.7%-100%]), the vascular same-admission imaging has a higher NPV (P = .03). Similarly, vascular delayed after discharge imaging has a higher NPV (100% [95% CI, 100%-100%]) compared to nonvascular imaging (P = .01); but is not significantly different from vascular same-admission imaging (P = .58). Patients’ Disposition A total of 92% (694/752) of patients were discharged home, while 6% (47/752) were sent to rehabilitation facilities and 2% (11/752) died in the hospital. Of 47 patients sent to rehabilitation facilities, 4 had positive findings on imaging that would explain the etiology for SAH. Of 11 patients who died, all had nonperimesencephalic pattern of hemorrhage, and none had positive findings on imaging to explain SAH. DISCUSSION The results of this retrospective multicenter study of 752 AN-SAH show extremely low diagnostic yield of cervical spine MRI (0.2%) and brain MRI (0.7%) for the management of AN-SAH, both commonly performed in AN-SAH patients. This diagnostic yield remained low (0.3% for cervical spine MRI and 1.2% for brain MRI) in patients with nonperimesencephalic pattern of hemorrhage. Repeat same-admission follow-up CTA and DSA performed in nearly half the patients and delayed after-discharge follow-up DSA and CTA performed in one-quarter and one-tenth of the patients, respectively. All have a slightly higher diagnostic yield than cervical spine or brain MRI, ranging between 1% and 4% regardless of pattern of hemorrhage. In patients with perimesencephalic hemorrhage, none of the follow-up imaging was positive for any etiology to explain SAH, while in those with nonperimsencephalic hemorrhage the diagnostic yield for follow-up DSA and CTA ranged between 1% and 8%. The NPV for all the follow-up tests was more than 96% with the vascular imaging tests having higher NPV compared to nonvascular imaging tests. Our multicenter study has several strengths. First, this is the largest available study assessing the diagnostic management of patients with AN-SAH. Second, with the availability of long-term angiographic follow-up (ranging from 1 to 7 yr) for patients with AN-SAH, we were able to determine with higher confidence whether imaging findings were clinically relevant to explain the etiology of SAH. The results from our large sample size support those from prior, smaller studies. Recent meta-analysis of literature reported a diagnostic yield of 1.3% (95% CI, 0.5%-2.5%) for cervical spine MRI in 538 patients with AN-SAH,22 with individual published studies reporting yields ranging from 0% to 4%.5,9,16,23-26 For brain MRI, a majority of studies reported a diagnostic yield close to 0%,6,16,27-29 with only a few older studies published prior to the year 2000 reporting higher diagnostic yields of 7% to11%.9,30 Nearly half of the etiologies identified on repeat angiography (DSA or CTA) in the current study were aneurysms. In fact, the most common cause of SAH is aneurysmal rupture, which may be missed upon initial angiographic evaluation due, for instance, to copious cisternal blood obscuring the aneurysm, vasospasm proximal and distal to the aneurysm preventing sufficient opacification, thrombosis of the aneurysm, small aneurysms in proximity to the calvaria on CTA, or simply due to a technically inadequate examination.31-33 Reported diagnostic yield of a second, repeat angiography (DSA or CTA) in AN-SAH varies between 0% and 30% in prior studies.17,34-37 In patients with nonperimesencephalic pattern of SAH, prior meta-analysis showed a diagnostic yield of 10% (95% CI, 7.4%-13.6%) for repeat DSA.38 In patients with perimesencephalic patterns of SAH, meta-analyses have reported diagnostic rates of 1.6% (95% CI, 0.7%-3.8%) for repeat, same-admission follow-up DSA39 and 0.78% (95% CI, 23%-1.32%) for repeat same-admission or after-discharge follow-up angiography (DSA or CTA).40 Our study findings are consistent with prior studies and demonstrate a diagnostic yield between 1% and 4% for repeat DSA or CTA regardless of pattern of hemorrhage. However, in patients with nonperimsencephalic pattern of hemorrhage, the diagnostic yields of repeat same admission follow-up DSA and CTA were 7.4% and 1.7%, respectively. In AN-SAH, there is high mortality associated with re-bleeding (up to 4% during the first 24 h and up to 15%-25% during the first 14 d, and even 50% in 6 mo).37,41,42 Further, patients are at risk for delayed cerebral ischemia (up to 4%), and hydrocephalus (up to 14%), which may require shunt placement.43 The risk of re-bleeding and consequent morbidity and mortality associated with an undiscovered source of SAH together with the low morbidity of angiography and MRI, tilts the risk-benefit balance for many clinicians toward ordering these tests. However, the cost-effectiveness of these tests given their low diagnostic yield is questionable and the focus of ongoing research. In the current study, the diagnostic yield of ∼4% for repeat angiography may support the cost-effectiveness of its utilization specifically in nonperimsencephalic pattern of hemorrhage (diagnostic yield of ∼8%) as all aneurysmal causes of patients initially diagnosed with AN-SAH were identified on repeat angiography in the present study. This is especially important if vasospasm is present on initial angiographic evaluation because detection of saccular aneurysms may prove difficult in this setting.17,25 Our results are limited in investigating the effectiveness of repeat DSA vs CTA due to small number of patients undergoing both tests in the current study. However, our results support the very low diagnostic yield of brain and cervical spine MRI. In addition, all except for 1 of the etiologies identified by MRI in the current study were nonaneurysmal (eg, vasculopathy, and vascular malformation) and did not include potentially immediately life-threatening conditions in contrast to the aneurysmal etiologies, diagnosed on follow-up angiography. The 1 aneurysm identified on brain MRI was also diagnosed on repeat, same-admission DSA and CTA. Furthermore, brain and cervical spine MRI likely add diagnostic value in only a very small proportion of patients, while adding to the financial burden of remaining patients with normal exams.22 With one-third of Americans reporting difficulty paying healthcare bills, many regularly face decisions about whether the benefits of recommended services justify associated costs,44,45 and in case of AN-SAH patients, brain and cervical spine MRI may represent examples of these services. Limitations We acknowledge several study limitations, including those inherent to our retrospective study design and the potential for selection bias, which is an issue in any retrospective analysis. The imaging analyses were performed by investigators from each center as opposed to centralized investigators reviewing all imaging interpretations, given the IRB approval at each site did not allow sharing the images or their reports with investigators from other sites. CONCLUSION In summary, there are variations in utilization of imaging tests for management of AN-SAH with brain and cervical spine MRI performed in 75% and 61% of cases, while having a very low diagnostic yield (less than 1% regardless of pattern of hemorrhage). Even in patients with nonperimesencephalic hemorrhage the diagnostic yields for cervical spine and brain MRI were less than 1% and 1.2%, respectively. On the other hand, repeat and delayed follow-up DSA and CTA have more reasonable diagnostic yield, but are performed in less than half of patients. As a result, we suggest that MRI for possible etiologies of AN-SAH is not routinely recommended unless there are focal neurological deficits to suggest a cranial or spinal etiology. Future studies to evaluate cost-effectiveness of MRI examinations as well as comparative effectiveness of follow-up CTA versus DSA are recommended. Disclosure The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article. Notes The abstract of this manuscript was presented as an oral presentation at the ASNR annual meeting, April 24-27, 2017 in Long Beach, California. REFERENCES 1. Bederson JB, Connolly ES Jr, Batjer HH et al.   Guidelines for the management of aneurysmal subarachnoid hemorrhage: a statement for healthcare professionals from a special writing group of the Stroke Council, American Heart Association. Stroke . 2009; 40( 3): 994- 1025. Google Scholar CrossRef Search ADS PubMed  2. Feigin VL, Lawes CM, Bennett DA, Anderson CS. Stroke epidemiology: a review of population-based studies of incidence, prevalence, and case-fatality in the late 20th century. Lancet Neurol . 2003; 2( 1): 43- 53. Google Scholar CrossRef Search ADS PubMed  3. Hackett ML, Anderson CS. Health outcomes 1 year after subarachnoid hemorrhage: An international population-based study. Neurology . 2000; 55( 5): 658- 662. Google Scholar CrossRef Search ADS PubMed  4. Mayer SA, Kreiter KT, Copeland D et al.   Global and domain-specific cognitive impairment and outcome after subarachnoid hemorrhage. Neurology . 2002; 59( 11): 1750- 1758. Google Scholar CrossRef Search ADS PubMed  5. Germans MR, Coert BA, Majoie CB et al.   Yield of spinal imaging in nonaneurysmal, nonperimesencephalic subarachnoid hemorrhage. Neurology . 2015; 84( 13): 1337- 1340. Google Scholar CrossRef Search ADS PubMed  6. Andaluz N, Zuccarello M. Yield of further diagnostic work-up of cryptogenic subarachnoid hemorrhage based on bleeding patterns on computed tomographic scans. Neurosurgery . 2008; 62( 5): 1040- 1047. Google Scholar CrossRef Search ADS PubMed  7. Kumar R, Das KK, Sahu RK et al.   Angio negative spontaneous subarachnoid hemorrhage: is repeat angiogram required in all cases? Surg Neurol Int . 2014; 5( 1): 125. Google Scholar CrossRef Search ADS PubMed  8. Connolly ES Jr, Rabinstein AA, Carhuapoma JR et al.   Guidelines for the management of aneurysmal subarachnoid hemorrhage: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke . 2012; 43( 6): 1711- 1737. Google Scholar CrossRef Search ADS PubMed  9. Rogg JM, Smeaton S, Doberstein C, Goldstein JH, Tung GA, Haas RA. Assessment of the value of MR imaging for examining patients with angiographically negative subarachnoid hemorrhage. Am J Roentgenol . 1999; 172( 1): 201- 206. Google Scholar CrossRef Search ADS   10. Gupta SK, Gupta R, Khosla VK et al.   Nonaneurysmal nonperimesencephalic subarachnoid hemorrhage: is it a benign entity? Surg Neurol . 2009; 71( 5): 566- 571. Google Scholar CrossRef Search ADS PubMed  11. Kaim A, Proske M, Kirsch E, von Weymarn A, Radu EW, Steinbrich W. Value of repeat-angiography in cases of unexplained subarachnoid hemorrhage (SAH). Acta Neurol Scand . 1996; 93( 5): 366- 373. Google Scholar CrossRef Search ADS PubMed  12. Yu DW, Jung YJ, Choi BY, Chang CH. Subarachnoid hemorrhage with negative baseline digital subtraction angiography: is repeat digital subtraction angiography necessary? J Cerebrovasc Endovasc Neurosurg . 2012; 14( 3): 210- 215. Google Scholar CrossRef Search ADS PubMed  13. Sadigh G, Allen J. National survey of radiologic management of angiographically-negative subarachnoid hemorrhage. ASNR Annual Meeting  2017. 14. Rinkel GJ, Wijdicks EF, Vermeulen M et al.   Nonaneurysmal perimesencephalic subarachnoid hemorrhage: CT and MR patterns that differ from aneurysmal rupture. AJNR Am J Neuroradiol . 1991; 12( 5): 829- 834. Google Scholar PubMed  15. Fontanella M, Rainero I, Panciani PP et al.   Subarachnoid hemorrhage and negative angiography: clinical course and long-term follow-up. Neurosurg Rev . 2011; 34( 4): 477- 484. Google Scholar CrossRef Search ADS PubMed  16. Akcakaya MO, Aydoseli A, Aras Y et al.   Clinical course of nontraumatic nonaneurysmal subarachnoid hemorrhage: A single institution experience over 10 years and review of the contemporary literature. Turk Neurosurg . 2016; 27( 5): 732- 742. 17. Canneti B, Mosqueira AJ, Nombela F, Gilo F, Vivancos J. Spontaneous subarachnoid hemorrhage with negative angiography managed in a stroke unit: Clinical and prognostic characteristics. J Stroke Cerebrovasc Dis . 2015; 24( 11): 2484- 2490. Google Scholar CrossRef Search ADS PubMed  18. Rinkel GJ, Wijdicks EF, Hasan D et al.   Outcome in patients with subarachnoid haemorrhage and negative angiography according to pattern of haemorrhage on computed tomography. Lancet . 1991; 338( 8773): 964- 968. Google Scholar CrossRef Search ADS PubMed  19. Rinkel GJ, van Gijn J, Wijdicks EF. Subarachnoid hemorrhage without detectable aneurysm. A review of the causes. Stroke . 1993; 24( 9): 1403- 1409. Google Scholar CrossRef Search ADS PubMed  20. Hunt WE, Hess RM. Surgical risk as related to time of intervention in the repair of intracranial aneurysms. J Neurosurg . 1968; 28( 1): 14- 20. Google Scholar CrossRef Search ADS PubMed  21. Fisher CM, Kistler JP, Davis JM. Relation of cerebral vasospasm to subarachnoid hemorrhage visualized by computerized tomographic scanning. Neurosurgery . 1980; 6( 1): 1- 9. Google Scholar CrossRef Search ADS PubMed  22. Sadigh G, Holder CA, Switchenko JM, Dehkharghani S, Allen JW. Is there added value in obtaining cervical spine MRI in the assessment of nontraumatic angiographically negative subarachnoid hemorrhage? A retrospective study and meta-analysis of the literature. J Neurosurg . 2017: 1- 7. 23. Woodfield J, Rane N, Cudlip S, Byrne JV. Value of delayed MRI in angiogram-negative subarachnoid haemorrhage. Clin Radiol . 2014; 69( 4): 350- 356. Google Scholar CrossRef Search ADS PubMed  24. Little AS, Garrett M, Germain R et al.   Evaluation of patients with spontaneous subarachnoid hemorrhage and negative angiography. Neurosurgery . 2007; 61( 6): 1139- 1151. Google Scholar CrossRef Search ADS PubMed  25. Topcuoglu MA, Ogilvy CS, Carter BS, Buonanno FS, Koroshetz WJ, Singhal AB. Subarachnoid hemorrhage without evident cause on initial angiography studies: Diagnostic yield of subsequent angiography and other neuroimaging tests. J Neurosurg . 2003; 98( 6): 1235- 1240. Google Scholar CrossRef Search ADS PubMed  26. Lin N, Zenonos G, Kim AH et al.   Angiogram-negative subarachnoid hemorrhage: relationship between bleeding pattern and clinical outcome. Neurocrit Care . 2012; 16( 3): 389- 398. Google Scholar CrossRef Search ADS PubMed  27. Maslehaty H, Petridis AK, Barth H, Mehdorn HM. Diagnostic value of magnetic resonance imaging in perimesencephalic and nonperimesencephalic subarachnoid hemorrhage of unknown origin. J Neurosurg . 2011; 114( 4): 1003- 1007. Google Scholar CrossRef Search ADS PubMed  28. Wijdicks EF, Schievink WI, Miller GM. MR imaging in pretruncal nonaneurysmal subarachnoid hemorrhage: Is it worthwhile? Stroke . 1998; 29( 12): 2514- 2516. Google Scholar CrossRef Search ADS PubMed  29. Maslehaty H, Barth H, Petridis AK, Doukas A, Maximilian Mehdorn H. Special features of subarachnoid hemorrhage of unknown origin: a review of a series of 179 cases. Neurol Res . 2012; 34( 1): 91- 97. Google Scholar CrossRef Search ADS PubMed  30. Renowden SA, Molyneux AJ, Anslow P, Byrne JV. The value of MRI in angiogram-negative intracranial haemorrhage. Neuroradiology . 1994; 36( 6): 422- 425. Google Scholar CrossRef Search ADS PubMed  31. Broderick JP, Brott TG, Duldner JE, Tomsick T, Leach A. Initial and recurrent bleeding are the major causes of death following subarachnoid hemorrhage. Stroke . 1994; 25( 7): 1342- 1347. Google Scholar CrossRef Search ADS PubMed  32. McMahon J, Dorsch N. Subarachnoid haemorrhage of unknown aetiology: What next? Crit Rev Neurosurg . 1999; 9( 3): 147- 155. Google Scholar CrossRef Search ADS PubMed  33. van Gijn J, Rinkel GJ. Subarachnoid haemorrhage: diagnosis, causes and management. Brain . 2001; 124( 2): 249- 278. Google Scholar CrossRef Search ADS PubMed  34. Nishioka H, Torner JC, Graf CJ, Kassell NF, Sahs AL, Goettler LC. Cooperative study of intracranial aneurysms and subarachnoid hemorrhage: A long-term prognostic study. Arch Neurol . 1984; 41( 11): 1147- 1151. Google Scholar CrossRef Search ADS PubMed  35. Juul R, Fredriksen TA, Ringkjob R. Prognosis in subarachnoid hemorrhage of unknown etiology. J Neurosurg . 1986; 64( 3): 359- 362. Google Scholar CrossRef Search ADS PubMed  36. Pathirana N, Refsum SE, McKinstry CS, Bell KE. The value of repeat cerebral angiography in subarachnoid haemorrhage. Br J Neurosurg . 1994; 8( 2): 141- 146. Google Scholar CrossRef Search ADS PubMed  37. Iwanaga H, Wakai S, Ochiai C, Narita J, Inoh S, Nagai M. Ruptured cerebral aneurysms missed by initial angiographic study. Neurosurgery . 1990; 27( 1): 45- 51. Google Scholar CrossRef Search ADS PubMed  38. Bakker NA, Groen RJ, Foumani M et al.   Repeat digital subtraction angiography after a negative baseline assessment in nonperimesencephalic subarachnoid hemorrhage: a pooled data meta-analysis. J Neurosurg . 2014; 120( 1): 99- 103. Google Scholar CrossRef Search ADS PubMed  39. Potter CA, Fink KR, Ginn AL, Haynor DR. Perimesencephalic hemorrhage: Yield of Single versus multiple DSA examinations-a single-center study and meta-analysis. Radiology . 2016; 281( 3): 858- 864. Google Scholar CrossRef Search ADS PubMed  40. Kalra VB, Wu X, Matouk CC, Malhotra A. Use of follow-up imaging in isolated perimesencephalic subarachnoid hemorrhage: a meta-analysis. Stroke . 2015; 46( 2): 401- 406. Google Scholar CrossRef Search ADS PubMed  41. Inamasu J, Nakamura Y, Saito R et al.   “Occult” ruptured cerebral aneurysms revealed by repeat angiography: result from a large retrospective study. Clin Neurol Neurosurg . 2003; 106( 1): 33- 37. Google Scholar CrossRef Search ADS PubMed  42. Vaitkevicius G, Gvazdaitis AR, Lukosevicius S. [Spontaneous subarachnoid hemorrhage: patients' examination after aneurysm-negative initial angiograms]. Medicina (Kaunas) . 2002; 38( 10): 976- 981. Google Scholar PubMed  43. Elhadi AM, Zabramski JM, Almefty KK et al.   Spontaneous subarachnoid hemorrhage of unknown origin: hospital course and long-term clinical and angiographic follow-up. J Neurosurg . 2015; 122( 3): 663- 670. Google Scholar CrossRef Search ADS PubMed  44. Obermeyer Z, Makar M, Abujaber S, Dominici F, Block S, Cutler DM. Association between the medicare hospice benefit and health care utilization and costs for patients with poor-prognosis cancer. JAMA . 2014; 312( 18): 1888- 1896. Google Scholar CrossRef Search ADS PubMed  45. Sadigh G, Carlos RC, Krupinski EA, Meltzer CC, Duszak R Jr. Health care price transparency and communication: Implications for radiologists and patients in an era of expanding shared decision making. Am J Roentgenol . 2017; 209( 5): 959- 964. Google Scholar CrossRef Search ADS   COMMENT The authors report on the radiologic management of angiographically negative, spontaneous intracranial subarachnoid hemorrhage as part of a multicenter study of utilization and diagnostic yield. This is a retrospective multicenter study, of consecutive adult patients admitted with none Traumatic, AN-SAH. Outcomes studied included utilization rate, diagnostic yield, and median time from admission for the following imaging tests: initial CTA and DSA, brain and cervical spine MRI, and any repeat DSA or CTA performed either during initial admission or at long-term follow-up. In the 752 patients who were included (mean age, 53 years; 54% male), Initial CTA and DSA were performed in 89% and 100% of patients, respectively. Brain MRI was performed in 75% of patients and was positive in 0.7% of cases. Cervical spine MRI was performed in 61% of patients and was positive in 0.2% of cases. Repeat, same-admission follow-up DSA and CTA were performed in 48% and 51% of patients and were positive in 3.3% and 1% of cases, respectively. Delayed follow-up DSA and CTA after discharge were performed in 26% and 7% of patients and were positive in 2% and 3.7% of cases, respectively, all with negative prior imaging studies. The authors conclude that cervical spine and brain MRI have extremely low diagnostic yield whereas repeat DSA and CTA are utilized less commonly but have slightly higher diagnostic yield. A large amount of data is already available in the management of non-traumatic SAH. The biggest problem with this study in addition to its retrospective study design, is that the pattern of hemorrhage is so important in the type of tests ordered and the expected negative and positive results. Much like routine administrative data studies, this study lacks the details required to truly establish and management paradigm. A patient with a SAH in the lower region of the posterior fossa does require and spinal MRI/A in the face of a negative angiogram versus a sylvan fissure SAH does not. Gavin W. Britz Houston, Texas Copyright © 2018 by the Congress of Neurological Surgeons This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

NeurosurgeryOxford University Press

Published: May 30, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off