Baseline Characteristics for two functional outcome groups of CVST patients
The demographic and clinical characteristics of patients were shown in Table 1. Based on mRS, 102 patients were defined as good functional outcome, and the remaining 35 patients had poor functional outcome. The population was comprised of 72 men and 65 women with a mean age of 39.5 years (range, 7 - 80 years). Patients with acute onset were identified frequently in the poor functional outcome group than patients with chronic onset (P = 0.032). Besides, headache with nausea and vomiting was more common among patients with unfavorable functional outcome (P = 0.042). In addition, there are significant difference between the two groups with respect to type of lesion (P < 0.001) and treatment method (P = 0.031). However, there was no significant difference in other factors including age, gender, and etiologies.
Laboratory indicators for two groups of CVST patients
Laboratory indicators of patients were described in Table 2. D-dimer level on admission was significantly higher in patients evaluated as poor functional outcome compared with patients evaluated as good functional outcome (5.02 ± 6.18 vs 1.47 ± 2.21, P = 0.002). Moreover, neutrophil count was significantly lower in the good functional outcome group than the poor functional outcome group (5.35 ± 2.75 vs 7.15 ± 3.54, P = 0.008). Furthermore, the difference on NLR and PRP between the two groups reached statistical significance (P = 0.003 and P = 0.013), but not LMR (P = 0.355) and PLR (P = 0.138).
Independent risk factors for functional prognosis of CVST patients
Potentially significant demographic, clinical, and laboratory indicators extracted in Table 1 and Table 2 were submitted to binary logistic regression analysis, and the details were shown in Table 3. The results demonstrated that chronic onset (OR, 0.211, 95CI%, 0.048-0.933, P = 0.040), D-dimer on admission (OR, 1.177, 95CI%, 1.021-1.356, P = 0.025) and NLR (OR, 1.243, 95CI%, 1.059-1.459, P = 0.008) were significantly associated with patients’ functional outcome, respectively.
The correlation between short-term outcome and significant clinical indicators
The correlation between functional outcome evaluated by mRS score and significant clinical indicators was described in Table 4. The results showed that the functional outcome was positively associated with treatment method (rho=0.221, P = 0.010), D-dimer on admission (rho=0.394, P < 0.001), neutrophil count (rho=0.258, P = 0.002), NLR (rho=0.357, P < 0.001), and PRP(%) (rho=0.298, P < 0.001), among which D-dimer on admission had the closest correlation. However, acute onset was negatively related with patients’ functional outcome (rho=-0.196, P = 0.021).
ROC curves analysis for NLR, D-dimer on admission, and combined detection
Patients with poor functional outcome had higher D-dimer level on admission and NLR than those with good functional outcome. Therefore, ROC curves of the predictive value of D-dimer on admission, NLR, and combined detection were shown in Fig 2. Simultaneously, the outcome of ROC curves for D-dimer on admission, NLR, and combined detection was described in Table 5. The area under the curve (AUC) of D-dimer on admission was 0.696, the optimal cut-off value was 0.845, the sensitivity was 0.743, the specificity was 0.598, and the Youden’s index was 0.341. AUC of NLR was 0.707, the optimal cut-off value was 2.105, the sensitivity was 0.971, the specificity was 0.392, and the Youden’s index was 0.364. In addition, AUC of combined detection was 0.703, the optimal cut-off value was 12.52, the sensitivity was 0.486, the specificity was 0.833, and the Youden’s index was 0.319.