Identifying venous clinical severity score thresholds for Clinical-Etiology-Anatomy-Pathophysiology classifications of venous edema and higher.

Objective: Venous Clinical Severity Score (VCSS) is a widely used standard for assessing and grading the severity of chronic venous disease (CVD). Prior research highlighted its high validity in detecting and quantifying venous disease. However, there is little, if any, known about the precise thresholds at which VCSS discriminates important stages of deep venous disease. This study sought to elucidate the diagnostic accuracy, thresholds, and correlation at which VCSS detects salient CEAP (Clinical-Etiology-Anatomy-Pathophysiology) classes in deep venous disease progression.

Methods: A registry of 840 patients who presented with chronic proximal venous outflow obstruction (PVOO) secondary to non-thrombotic iliac vein lesions from August 2011 to June 2021 was retrospectively analyzed. VCSS and CEAP classifications were used to evaluate preoperative symptoms. VCSS was compared to CEAP classes to determine the precise VCSS composite values at which the instrument was able to detect CEAP C3 and higher, C4 and higher, and C5 and higher. Receiver operative characteristic (ROC) curve and area under the curve (AUC) were used to evaluate VCSS for its ability to discriminate disease at these stages of CEAP classification. Spearman's rank coefficient was used to determine the correlation between CEAP VCSS composite as well as individual VCSS components (pain, varicose vein, edema, pigmentation, inflammation, induration, ulcer number, ulcer size, ulcer duration, compression).

Results: VCSS composite was able to detect venous edema (C3) and higher at a sensitivity of 68.9% and a specificity of 54.8% at an optimized threshold of 8.5 (AUC = 0.648; 95% C.I. = 0.575-0.721). To detect changes in skin and subcutaneous tissue from CVD (C4) and higher, an optimal threshold of 11.5 was found with a sensitivity of 51.7% and specificity of 76.5% (AUC = 0.694; 95% C.I. = 0.656-0.731). Healed venous ulcer (C4) and higher was detectable at an optimized threshold of 13.5 at a sensitivity of 67.7% and a specificity of 88.9% (AUC = 0.819; 95% C.I. = 0.766-0.873). The correlation between VCSS composites and CEAP was weak (ρ = 0.372; p < .001). Attributes of VCSS that reflect more severe venous disease correlated more closely with CEAP classes, namely pigmentation (ρ = 0.444; p < .001), inflammation (ρ = 0.348; p < .001), induration (ρ = 0.352; p < .001), number of active ulcers (ρ = 0.497; p < .001), active ulcer size (ρ = 0.485; p < .001), and ulcer duration (ρ = 0.497; p < .001). The correlation between CEAP class and the other four components of VCSS were not statistically significant.

Conclusion: VCSS composite thresholds of 8.5, 11.5, and 13.5 are threshold values for detecting CEAP classification C3 and higher, C4 and higher, and C5 and higher, respectively. Consistent with prior work, VCSS appears to have a better ability to discriminate CVD at more severe CEAP classifications. In this registry, the correlation between VCSS and CEAP was found to be weak while components of VCSS that suggest more advanced disease exhibited the strongest correlation with CEAP.