Clinical Risk Score for Prediction of Urgency in Carotid Cavernous Sinus Fistulas

To develop a clinical risk score for the prediction of urgency in patients with carotid cavernous sinus fistulas (CCFs) and test for the discriminative ability of the diagnostic prediction.

The medical charts of 60 patients with CCFs were retrospectively reviewed. The clinical characteristics of direct and dural CCFs were analyzed by logistic regression. The clinical risk score was developed from the coefficient in the multivariable regression model and used to predict direct CCFs which were more urgent than the dural type. The score prediction was reported as an area under the receiver operating characteristic (AuROC) curve and 95% confidence interval (95% CI).

In a univariable analysis, the clinical characteristics which increased the risk of direct CCFs were age, gender, trauma, underlying diseases, visual acuity (VA) at presentation, bruit, chemosis, and dilated retinal vessels. However, in multivariable analysis, the significant predictors were limited to age, trauma, bruit, underlying diseases and logMAR VA. Regression coefficient of each predictor was converted to a risk score and summation of scores from these predictors for each patient was calculated. The total risk score predicted the urgent direct CCFs correctly with AuROC of 97.77% (95% CI; 93.57, 100).

The clinical risk score for the prediction of urgent direct CCFs has been developed and used in the patients with CCFs in our setting. The discriminative ability of the score prediction is high. This simple clinical risk score may help clinicians suspect direct CCFs and urgently refer the patients to have prompt angiography and treatment.