Kowalski RG, Claassen J, Kreiter KT, Bates JE, Ostapkovich ND, Connolly ES, Mayer SA. Initial Misdiagnosis and Outcome After Subarachnoid Hemorrhage. JAMA. 2004;291(7):866-869. doi:10.1001/jama.291.7.866
Author Affiliations: Division of Stroke and Critical Care Neurology, Department of Neurology (Mssrs Kowalski and Bates, Ms Ostapkovich, and Drs Claassen, Kreiter, and Mayer), and Department of Neurosurgery (Drs Connolly and Mayer), Columbia University College of Physicians and Surgeons, New York, NY.
Context Mortality and morbidity can be reduced if aneurysmal subarachnoid hemorrhage
(SAH) is treated urgently.
Objective To determine the association of initial misdiagnosis and outcome after
Design, Setting, and Participants Inception cohort of 482 SAH patients admitted to a tertiary care urban
hospital between August 1996 and August 2001.
Main Outcome Measures Misdiagnosis was defined as failure to correctly diagnose SAH at a patient's
initial contact with a medical professional. Functional outcome was assessed
at 3 and 12 months with the modified Rankin Scale; quality of life (QOL),
with the Sickness Impact Profile.
Results Fifty-six patients (12%) were initially misdiagnosed, including 42 of
221 (19%) of those with normal mental status at first contact. Migraine or
tension headache (36%) was the most common incorrect diagnosis, and failure
to obtain a computed tomography (CT) scan was the most common diagnostic error
(73%). Neurologic complications occurred in 22 patients (39%) before they
were correctly diagnosed, including 12 patients (21%) who experienced rebleeding.
Normal mental status, small SAH volume, and right-sided aneurysm location
were independently associated with misdiagnosis. Among patients with normal
mental status at first contact, misdiagnosis was associated with worse QOL
at 3 months and an increased risk of death or severe disability at 12 months.
Conclusions In this study, misdiagnosis of SAH occurred in 12% of patients and was
associated with a smaller hemorrhage and normal mental status. Among individuals
who initially present in good condition, misdiagnosis is associated with increased
mortality and morbidity. A low threshold for CT scanning of patients with
mild symptoms that are suggestive of SAH may reduce the frequency of misdiagnosis.
Subarachnoid hemorrhage (SAH) affects nearly 30 000 individuals
annually in North America and results in serious impairment or death in 40%
to 60% of cases.1 Outcome is highly dependent
on early diagnosis and aggressive intervention.1,2 Immediate
aneurysm repair is particularly crucial because rebleeding occurs in 26% to
73% of patients within days or weeks after the initial rupture if the aneurysm
The reported frequency of misdiagnosis of SAH ranges from 12% to 51%.3- 11 Correct
diagnosis can be confounded because a key symptom of SAH, headache, is among
the most common symptoms reported to emergency physicians.12 Accordingly,
misdiagnosed SAH represents one of the largest sources of emergency department
litigation claims and malpractice settlement payments in the United States.13 We sought to identify the frequency, risk factors,
and impact on outcome of initial misdiagnosis in patients hospitalized with
All patients who were diagnosed with SAH and were admitted to the Neurological
Intensive Care Unit at Columbia-Presbyterian Medical Center in New York between
August 1996 and August 2001 were invited to enroll in the Columbia University
SAH Outcomes Project. The study was approved by the hospital's institutional
review board, and written informed consent was obtained from the patient or
a surrogate. SAH was diagnosed according to computed tomography (CT) or by
xanthochromia of cerebrospinal fluid (CSF). Patients with spontaneous aneurysmal
and nonaneurysmal SAH were included. Individuals with SAH caused by trauma,
arteriovenous malformations, or other secondary causes were excluded, as were
patients younger than 18 years and those admitted more than 14 days after
their most recent hemorrhage.
Demographic data, medical and social history, and clinical features
at admission were obtained through patient and surrogate interviews and medical
record review by a study neurointensivist. Details about symptoms at the onset
of hemorrhage, admission Hunt-Hess grade, and CT and angiographic findings
were recorded, as described previously.14- 17 Sentinel
headaches were defined as discrete episodes of severe headache that preceded
the headache that initially led the patient to seek medical attention. Aneurysms
were designated right-sided if located on the right middle cerebral artery
(MCA) or internal carotid artery (ICA) and left-sided if located on the left
MCA or ICA.
Misdiagnosis was defined as failure to correctly identify a subsequently
documented SAH on previous presentation to a health care professional. Types
of encounters included ambulance calls, emergency department visits, clinician
office visits, telephone calls to a health care professional (physician, nurse,
or physician assistant), or hospitalization. For patients whose SAH was initially
misdiagnosed, the date and type of medical contact, preliminary diagnosis,
and initial Hunt-Hess grade according to descriptions of the patients' condition
were recorded. The presumed error in diagnosis and complications that occurred
between the initial misdiagnosis and eventual correct diagnosis (neurologic
deterioration, aneurysm rebleeding, and symptomatic vasospasm or hydrocephalus)
were determined by the admitting study neurointensivist and adjudicated by
consensus of other members of the study team.
Patients and surrogate informants were interviewed at 3 and 12 months
either by telephone or in person. Global outcome and functional status were
assessed with the modified Rankin Scale18 (mRS;
0, no symptoms or disability; 6, dead), and quality of life (QOL) was evaluated
with the Sickness Impact Profile (SIP; 0, best; 100, worst).19 Proxy
responses from an informant were used for patients who were too severely impaired
to complete the mRS or SIP.20 All assessment
instruments were administered in the native language (English or Spanish)
of the patient or surrogate informant.
We analyzed all patients and those initially presenting with normal
mental status (Hunt-Hess grade I or II) to control for the confounding effects
of baseline clinical grade on outcome. Differences in proportions were compared
with the χ2 test or Fisher exact test. The independent samples t test (2-tailed) was used to analyze normally distributed
continuous variables, with correction for unequal variances as appropriate.
Independent predictors of misdiagnosis were identified with forward stepwise
logistic regression. Modified Rankin Scale scores were dichotomized to identify
patients who died (mRS 6), those who died or had severe disability (mRS 4-6),
and those who died or had any disability (mRS 2-6); when 12-month data were
missing, the 3-month score was substituted (last observation carried forward).
Level of significance was set at P<.05. Data were
analyzed with SPSS version 9.0 (SPSS Inc, Chicago, Ill).
Of 482 patients with SAH, 56 (12%) were misdiagnosed on first medical
contact. In 75% of cases, the misdiagnosis occurred in a hospital emergency
department or in a physician's office (Table 1). The interval between initial SAH onset and correct diagnosis
was greater for initially misdiagnosed than correctly diagnosed patients (median,
4.0 [range, 0-35] vs 0.0 [range, 0-14] days, P<.001).
The most common diagnostic error was failure to obtain a CT, and migraine
or tension headache was the most common incorrect diagnosis (Table 1). Twenty-two of the 56 (39%) misdiagnosed patients experienced
at least 1 complication before hospital admission: the most common were a
decreased level of consciousness and rebleeding (Table 1). There were no trends in the rate of misdiagnosis throughout
the 5-year study period.
Compared with patients with an initially correct SAH diagnosis, misdiagnosed
patients had smaller hemorrhages, were less likely to have lost consciousness
at ictus, and were more likely to have experienced sentinel headaches or to
have a right-sided aneurysm (Table 2).
Independent predictors of misdiagnosis in the final multivariate model included
Hunt-Hess grade I or II condition (indicating normal mental status) at first
medical contact (adjusted odds ratio [AOR], 10.8; 95% confidence interval
[CI], 3.2-37.1; P<.001), smaller SAH volume (SAH
score <15) (AOR, 5.1; 95% CI, 1.8-13.9; P = .002),
and right-sided aneurysm location (AOR, 3.3; 95% CI, 1.5-7.0; P = .003).
Among patients who were Hunt-Hess grade I or II at first medical contact,
misdiagnosed patients were also less likely to be fluent in English, to be
married, and to have 12 or more years of education (Table 2). Independent predictors of misdiagnosis in this subset
of patients included less than 12 years of education (AOR, 4.5; 95% CI, 1.9-10.5; P<.001) and smaller SAH volume (AOR, 13.5; 95% CI, 3.0-60.8; P<.001).
Overall, 18% of patients died during hospitalization, and 24% were dead
at 12 months. Mortality and functional disability (Table 3) and SIP total scores at 3 and 12 months were not significantly
different between misdiagnosed and correctly diagnosed patients. Among the
221 patients who were Hunt-Hess grade I or II at first medical contact, misdiagnosed
patients were more likely to have died or to have died or be severely disabled
at 3 and 12 months (Table 3).
Among survivors, misdiagnosed patients also had significantly poorer QOL,
as assessed by the SIP total score at 3 months (21  vs 12 ; P = .03) but not at 12 months (19  vs 13 ; P = .10).
Of 482 patients diagnosed with SAH and admitted to our institution over
5 years, 12% were initially misdiagnosed. Previous studies have generally
reported higher frequencies of initial misdiagnosis, ranging from 12% to 51%.3- 9 These
differences may be explained by variations in patient behavior, clinical practice,
access to emergency medical care, or methods of case ascertainment. Increased
awareness among clinicians about the importance of ruling out SAH in patients
with sudden, severe headache may also explain the lower rate in our study.
Our results confirm that patients with good neurologic grade (ie, those
most likely to benefit from urgent aneurysm repair) have the highest risk
of being misdiagnosed.3,4 Unlike
previous studies that found that the most common diagnostic errors were failure
to interpret subtle CT or CSF findings properly,3,11 we
found that failure to obtain a CT at initial contact was the most frequent
diagnostic error. The spectrum of incorrect working diagnoses (Table 1) and the median delay of 4 days to establish the correct
diagnosis in our study are similar to previous findings.3- 9
To our knowledge, our study is the largest to analyze the association
between initial misdiagnosis and outcome after SAH and the first to identify
risk factors in a multivariate analysis. In addition to good clinical grade,
smaller hemorrhages (perhaps relating to less severe headache) and right-sided
aneurysm location were independently associated with initial misdiagnosis.
Reduced SAH volume might also be explained by the gradual disappearance of
blood on CT before the correct diagnosis was made.21 The
importance of the association with right-sided aneurysm location is unclear;
perhaps right hemisphere dysfunction may lead to neglect or lack of concern
on the part of the patient when he or she describes symptoms to a medical
practitioner. Misdiagnosed patients also were more likely to report antecedent
sentinel headaches than those who were correctly diagnosed, which might be
explained by a tendency for clinicians to incorrectly assume that recent recurrent
headaches make a benign condition such as migraine or tension headache more
Among patients who were Hunt-Hess grade I or II at first medical contact,
lower education (≤12 years), smaller hemorrhages, nonfluency in English,
and being unmarried were also associated with misdiagnosis. These findings
implicate sociocultural barriers and lack of social support in explaining
some SAH misdiagnoses.
Although initial misdiagnosis was not related to outcome in our study
population as a whole, this comparison is confounded by the fact that misdiagnosed
patients were generally in better neurologic condition. When we restricted
our analysis to the 45% of patients who initially presented with a normal
neurologic examination (Hunt-Hess grade I or II), nearly 20% were misdiagnosed,
and the impact was devastating: misdiagnosis was associated with a nearly
4-fold increase in the likelihood of death at 12 months and with worse functional
recovery and QOL among survivors.
Our study has several limitations. First, our hospital-based series
may not accurately reflect the true frequency and impact of SAH misdiagnosis,
because 12% to 50% of patients with SAH who have rebleeding may die without
being hospitalized.22,23 In addition,
we have no data on misdiagnosed patients with a good outcome who were not
subsequently hospitalized. Second, our assessments of initial level of consciousness,
diagnostic error, and complications of delayed diagnosis were based on subjective
assessments of historical data. To address these limitations, these variables
were adjudicated by multiple study team members who obtained their own history.
Finally, our study is a single-center study and may have referral bias.
Important clues to the diagnosis of SAH, in addition to the complaint
of a sudden "thunderclap" headache, include loss of consciousness or vomiting
at onset, meningismus, prominent neck or back pain, retinal subhyaloid hemorrhages,
and recent headaches that have lasted for days at a time.2,11,24 In
addition to careful attention to these symptoms and signs, CT imaging (and
CSF examination when necessary) should be used to exclude SAH for any patient
with mild symptoms and for whom the diagnosis is a consideration.25