Diagnostic yield and outcomes of hand surgery referrals to rheumatology

ABSTRACT Background: Hand and upper extremity complaints frequently prompt evaluation for systemic inflammatory conditions, leading surgeons to refer patients to rheumatology. However, referral patterns are highly variable and not standardized, and access to rheumatologic care is often limited by prolonged wait times and workforce constraints. Existing literature on rheumatology referrals has largely focused on access and disease-specific pathways, with little attention to referral patterns within surgical subspecialties. Concordance between referral indication and confirmed rheumatologic diagnosis is reported to be low, underscoring the need to better define which referral triggers demonstrate true diagnostic yield. Purpose: To quantify and characterize referrals from hand surgery to rheumatology and identify predictors of diagnostic yield. Study Design/Level of Evidence: Retrospective cohort/3. Methods: We analyzed 183 hand surgery referrals to rheumatology (2018-2024) in this retrospective cohort study. Referral reasons were classified as laboratory-based, history-based, examination-based, or imaging-based. Outcomes were categorized as “low-yield” (Group 1: no rheumatologic diagnosis), “intermediate-yield” (Group 2: diagnosed without need for therapy or follow-up), and “high-yield” (Group 3: diagnosed, requiring systemic therapy or specialty follow-up). Descriptive statistics and multivariable logistic regression (high-yield vs low-yield) were performed. Results: Sixty-six percent of referrals were evaluated. Laboratory abnormalities were the most common referral reason (40%), followed by history (27%), examination findings (24%), and imaging (9%). Thirty-five percent (95% confidence interval [CI], 26.4%–44.0%) of referrals were high-yield, while 49% (95% CI, 39.6%–58.0%) were low-yield and 17% (95% CI, 10.6%–24.6%) were intermediate-yield. Across major referral indications (antinuclear antibodies, synovitis, and polyarthralgia), low-yield outcomes exceeded high-yield outcomes. Imaging-based referrals were significantly more likely to be high yield than laboratory-based referrals (adjusted OR 6.1, 95% CI, 1.5–31.5, P = .016). Examination-based and history-based referrals were not associated with increased diagnostic yield, and age and sex were not independent predictors. Conclusions: One-third of referrals produced a high-yield rheumatologic diagnosis, whereas half resulted in no rheumatologic diagnosis. Laboratory-based and history-based referrals were common but low-yield, whereas those supported by radiographic findings were significantly more likely to be high-yield. Prioritizing referrals with objective clinical or imaging findings may improve diagnostic efficiency in the hand surgery population.