Suggested Methods Language

Pulmonary arterial pressure (PAP) scores were analyzed using PAP-RIDGE (Markel, 2026), a browser-based decision-support tool for bovine high-altitude disease risk assessment. Observed mPAP values were adjusted for fixed effects of sex, breed, age, and year using a frozen linear model derived from ranch-level PAP testing records. Covariate-adjusted PAP scores were logit-transformed and shrinkage-corrected toward the group mean using an empirical Bayes approach, yielding a latent Z score for each animal. Back-transformed shrunk PAP estimates and Monte Carlo simulation (n = 2,000) were used to derive posterior probabilities of exceeding clinically relevant thresholds (P[PAP ≥ 70 mmHg] and P[PAP ≥ 80 mmHg]). Animals were classified into latent risk categories (Low, Watch, Elevated, High, or Extreme) based on these posterior probabilities. Where available, previous PAP records were incorporated to characterize PAP trajectory.