PURPOSE: Conventional angiography is often used in the preoperative work-up of hand surgery patients with systemic sclerosis. The goal of this study was to propose a classification system based on the pattern of arterial involvement in a series of upper extremity angiograms. The authors hypothesized that this classification system would demonstrate high inter- and intrarater reliability.
METHODS: A retrospective review of 110 upper extremity angiograms in patients with systemic sclerosis (obtained between 1996 and 2017) was performed. Images were classified into 4 types based on the patency of the radial and ulnar arteries at the wrist, and into 3 subtypes based on the patency of the superficial and deep palmar arches. Classification reliability was compared with Fleiss' Kappa (for inter-rater) and Cohen's (for intrarater) coefficient between 4 fellowship-trained hand surgeons and a hand fellow.
RESULTS: The inter-rater reliability between all 5 observers using types alone was 0.83 (0.80-0.85), whereas the inter-rater reliability using subtypes was 0.64 (confidence interval [CI] 95%, 0.62-0.65). The intrarater reliability using types alone ranged from 0.80 to 0.95, whereas intrarater reliabilities using subtypes were 0.81 (CI 95%, 0.72-0.90), 0.78 (CI 95%, 0.69-0.87), 0.87 (CI 95%, 0.80-0.95), 0.64 (CI 95%, 0.53-0.75), and 0.92 (CI 95%, 0.86-0.98) for the 4 attendings and a hand fellow, respectively. Fifty-seven percent of angiograms were interpreted as having loss of ulnar artery patency at the wrist (type 2) with 77% having additional loss of superficial palmar arch patency (type 2A).
CONCLUSIONS: This large series of angiograms in patients with systemic sclerosis demonstrates a classification system for conventional angiography that shows high inter-rater and intrarater reliability using type alone. When subtypes were used, the inter-rater and intrarater reliabilities decreased to moderate and moderate-to-high, respectively.
CLINICAL RELEVANCE: This study represents the first step in establishing a classification system that, by grouping patients with similar angiogram findings, may allow for targeted research into risk stratification, monitoring, and treatment in systemic sclerosis.