Abstract P101: Clinical Risk Prediction Scores for Primary Aldosteronism

Introduction: Primary aldosteronism (PA) is a common and often underdiagnosed secondary cause of hypertension with cardiovascular, kidney, and metabolic disease rates far in excess of essential hypertension matched for blood pressure levels. Current clinical scoring algorithms could reduce patient and healthcare burdens associated with expanded PA screening but have not been extensively validated in external cohorts. Methods: We evaluated six PA scoring algorithms in two populations of patients seen in the Southern Illinois University (SIU) Hypertension clinic: all patients screened for PA where a negative screen was assumed to rule out PA without further testing (n = 592) and only screen positive patients with subsequent confirmatory testing (n = 90). Area under the receiver operating characteristic curve (AUC) was used to determine their overall performance predicting PA. Negative predictive value (NPV) was used to assess the ability to identify patients with a low probability of PA. Results: Only one score showed an acceptable ability to predict PA (AUC > 0.7). NPV was inflated in the all screen cohort compared to confirmatory testing, where known status yielded low NPV for all scores. Conclusion: Currently available scoring algorithms poorly predict those with PA under current diagnostic guidelines. Equating a negative screen with a negative confirmatory test, as was common in score development, artificially inflates the NPV of these scores. Scores show limited utility in identifying screen positives that do not need to undergo further evaluation. More accurate PA risk prediction scores are needed to guide decision-making in patients suspected of having PA.