Abstract 5791: A New Score to Improve the Detection of Peripheral Artery Disease
Background: Detection of peripheral artery disease (PAD) typically entails noninvasive imaging or population screening, but whether a risk factor-based model has clinical utility is unclear. Our objective was to derive and internally validate a new score for estimating the probability of PAD utilizing the large US REACH Registry dataset.
Methods: Analyses were conducted on 23,395 US REACH Registry outpatients with a complete baseline risk factor profile. PAD presence was determined by previous or current intermittent claudication associated with an ankle-brachial index <0.9 or previous lower extremity intervention. Multivariable stepwise logistic regression identified cross-sectional correlates of PAD from demographic, clinical, and laboratory variables. Model performance was assessed internally using 10-fold cross validation, and effect estimates were used to generate the PAD score.
Results: Overall PAD prevalence was 9.1% (n=2135), with a mean age of 70 years. Age, smoking, history of congestive heart failure, diabetes, coronary artery disease (CAD), cerebrovascular disease (CVD), BMI, sex, and hypertension stage were predictive of PAD (Table 1⇓). The model had reasonable discrimination on derivation and internal validation (c-index = 0.61 and 0.60, respectively). The model-estimated PAD prevalence varied more than 3-fold from lowest to highest decile (range 4.8% to 16.4%) and corresponded closely with the actual PAD prevalence observed in each.
Conclusions: This new PAD detection score uses easily available clinical variables to estimate the probability of disease prevalence. While the predictive power of the model may be limited, its use should nevertheless improve PAD detection in vulnerable, at-risk populations. These data should provide clinicians and patients with a simple risk-based tool to estimate the probability of PAD prevalence across a wide range of patients with atherothrombosis.