Delay in the Diagnosis of Cerebral Vein and Dural Sinus Thrombosis
Influence on Outcome
Background and Purpose— Diagnostic delay of cerebral vein and dural sinus thrombosis may have an impact on outcome.
Methods— In the International Study on Cerebral Vein and Dural Sinus Thrombosis (ISCVT) cohort (624 patients with cerebral vein and dural sinus thrombosis), we analyzed the predictors and the impact on outcome of diagnostic delay. Primary outcome was a modified Rankin Scale score >2 at the end of follow-up. Secondary outcomes were modified Rankin Scale score 0 to 1 at the end of follow-up, death, and visual deficits (visual acuity or visual field).
Results— Median delay was 7 days (interquartile range, 3 to 16). Patients with disturbance of consciousness (P<0.001) and of mental status (P=0.042), seizure (<0.001), and with parenchymal lesions on admission CT/MR (P<0.001) were diagnosed earlier, whereas men (P=0.01) and those with isolated intracranial hypertension syndrome (P=0.04) were diagnosed later. Between patients diagnosed earlier and later than the median delay, no statistically significant differences were found in the primary (P=0.33) and in secondary outcomes: modified Rankin Scale score 0 to 1 (P=0.86) or deaths (P=0.53). Persistent visual deficits were more frequent in patients diagnosed later (P=0.05). In patients with isolated intracranial hypertension syndrome, modified Rankin Scale score >2 at the end of follow-up was more frequent in patients diagnosed later (P=0.02).
Conclusions— Diagnostic delay was considerable in this cohort and was associated with an increased risk of visual deficit. In patients with isolated intracranial hypertension syndrome, diagnostic delay was also associated with death or dependency.
The identification of variables that delay the diagnosis of cerebral vein and dural sinus thrombosis (CVT) is of clinical and medicolegal relevance. Two previous studies1,2 identified clinical variables associated with admission delay, but could not demonstrate an influence of admission delay on outcome. The purposes of this study were to describe the delay in the diagnosis of CVT and to analyze its predictors and its influence on the outcome of a large cohort of patients with CVT.
Diagnostic delay was defined as the interval (days) between the onset of symptoms (day of first symptom judged to be related to CVT by the investigator) and the confirmation of CVT by MR and/or angiography. We also measured the interval (days) between onset of symptoms and hospital admission (first hospital). Early admission was defined as an admission delay below the median of the sample.
We used the International Study on Cerebral Vein and Dural Sinus Thrombosis (ISCVT) cohort consisting of 624 consecutive symptomatic patients with CVT with a mean follow-up of 478 days.3 Follow-up visits were performed at 6 and12 months and yearly thereafter.
The primary outcome was death or dependency defined as a modified Rankin Scale score (mRS) >2 at end of follow-up. Secondary outcomes were all deaths, complete recovery (mRS 0 to 1) at the end of follow-up, and any persistent visual deficit related to CVT (decreased visual acuity or visual field defect) at final follow-up. Confrontation testing of visual fields and measurement of visual acuity were recommended at each follow-up, but perimetry was optional.
We used nonparametric statistics (median and quartile distribution, Mann-Whitney U and Kruskal-Wallis tests). We performed the following subgroup analysis: (1) low CVT risk score (<3) and high risk (CVT risk score ≥3) patients4; (2) patients presenting as isolated intracranial hypertension syndrome; (3) patients in coma (Glasgow Coma Score <9); (4) patients with brain lesions on admission CT/MR; and (5) patients who were anticoagulated in therapeutical dosages after the diagnosis of CVT. Additionally, we performed logistic regression for the primary outcome, forcing diagnostic delay in the ISCVT prognostic model, considering the interactions with unbalanced predictors.
There were 623 valid cases for analysis with information on onset to diagnosis delay. The median delay was 7 days and the interquartile range 3 to 16 days. The distribution showed a strong left-sided skew (skewness=11.36). There were 619 patients with valid information on admission delay. The median admission delay was 4 days and the interquartile range 1 to 11 days.
Early admission was associated with higher gross national income (P=0.02), disturbance of consciousness (P=0.001), mental disturbance (P=0.03), seizure (P=0.001), motor deficit (P=0.02), and parenchymal lesion on admission CT/MR (P=0.001).
Median delay (interquartile range) was 9 days (5 to 18) for countries with low-middle gross national income per capita and 7 days (3 to 15) for countries with high gross national income per capita (P=0.013). Diagnostic delay was decreased females and in patients with decreased vigilance, mental disturbance, seizure, and venous infarct on admission CT/MR. There were no significant differences in diagnostic delay in relation to age, centers with <10 patients, patients referred from other centers and patients requiring intra-arterial angiography for diagnosis, focal deficits (motor, sensory or language), the number or site of thrombosed sinus and veins, and any of the associated conditions or risk factors. Diagnostic delay was longer in patients with papilledema and isolated intracranial hypertension syndrome. More patients with diagnostic delay above the median were not prescribed anticoagulants in therapeutic dosages: 11% (22.4% versus 11.4%, χ2=13.5, P<0.001).
No statistically significant differences were found in the primary outcome and in secondary outcomes: mRS 0 to 1 or death. Persistent visual loss (visual field defects in 6 patients and decreased visual acuity in 21 patients; median visual acuity in right and left eyes 8 of 10; range, 9.5 to 5.5 of 10) was more frequent in patients diagnosed later (Table 2). We performed logistic regression with mRS >2 at final follow-up as the outcome and the variables of the ISCVT prognostic model and diagnostic delay as predictors. Diagnostic delay was not retained in the model either when dichotomized in below/above the median (P=0.056; adjusted OR, 0.60; 95% CI, 0.35 to 1.01) or when quartiles were considered (P=0.24; adjusted OR, 0.87; 95% CI, 0.69 to 1.10).
Diagnostic delay was significantly associated with visual loss in patients with isolated intracranial hypertension syndrome and in the low CVT risk score subgroups (Table 3⇓). In patients with isolated intracranial hypertension syndrome, poor outcome was significantly more frequent in patients with longer diagnostic delay.
In this multicenter cohort, the median interval between onset of symptoms and confirmation of CVT diagnosis was 7 days. Patients with more severe clinical presentations, women, and patients in high-income countries tend to be diagnosed earlier. In the overall sample, diagnostic delay had no influence on death or dependency or complete recovery. However, diagnostic delay was associated with an increased risk of poor outcome and of visual deficit in patients presenting with a syndrome of isolated increased intracranial hypertension.
Although visual acuity and confrontation testing of the visual fields were performed in the majority of the patients with visual complaints, information on visual acuity was not always recorded by the investigators and its frequency is probably underestimated. A standardized evaluation of visual consequences of CVT would require formal neuro-ophthalmological examination. We used a pragmatic approach considering any objective visual deficit not explained by an ocular cause as CVT-related. Another limitation of the research on diagnostic delay in CVT is the difficulty in establishing the onset of symptoms in patients with subacute or chronic presentations. This is certainly subject to considerable interexaminer variation, which was not studied in our investigation. Results of this study cannot be generalized to world regions that did not participate or were underrepresented in ISCVT such as Africa or Asia. Median admission delay and predictors of admission delay were consistent with the findings of the 2 previous studies with smaller sample sizes.1,2
Early diagnosis is more likely to have a favorable influence in outcome if an intervention has increased efficacy if applied earlier. There are no studies analyzing the influence on outcome of the delay on starting anticoagulation. Starting treatment earlier in patients with good initial condition may prevent pulmonary embolism and thrombus progression. Local thrombolysis, hematoma evacuation, or hemicraniectomy were used in a minority of patients in serious condition who, as we have shown, tend to present and be diagnosed earlier.
We thank the investigators who participated in the ISCVT. Their names and centers are listed in reference 3.
Sources of Funding
This study was supported by grants from the “Fundação para a Ciência e Tecnologia” PRAXIS C/SAU/10248/1998 and the “Associação para o Desenvolvimento da Investigação da Doença Vascular Cerebral.”
- Received April 2, 2009.
- Accepted May 8, 2009.
de Bruijn SFTM, de Haan RJ, Stam J; for the Cerebral Venous Thrombosis Study Group. Clinical features and prognostic factors of cerebral venous sinus thrombosis in a prospective series of 59 patients. J Neurol Neurosurg Psychiatry. 2001; 70: 105–108.
Ferro JM, Canhão P, Stam J, Bousser M-G, Barinagarrementeria F; for the ISCVT Investigators. Prognosis of cerebral vein and dural sinus thrombosis: results of the International Study on Cerebral Vein and Dural Sinus Thrombosis (ISCVT). Stroke. 2004; 35: 664–670.
Ferro JM, Rodrigues T, Bacelar-Nicolau L, Bacelar-Nicolau H, Canhão P, Crassard I, Dutra A, Massaro A, Mackowiak-Cordiolani MA, Leys D, Fontes J. Development of a risk score to predict the prognosis of cerebral vein and dural sinus thrombosis (CVT). Cerebrovasc Dis. 2007; 23 (suppl 2): 141.