Identifying Patients with Bicuspid Aortic Valve Disease in UK Primary Care: A Case–Control Study and Prediction Model

Evans, William; Akyea, Ralph Kwame; Weng, Stephen; Kai, Joe; Qureshi, Nadeem

Abstract

Bicuspid aortic valve disease (BAV) is the most common congenital heart condition, and early detection can improve outcomes for patients. In this case–control study, patients with a diagnosis of BAV were identified from their electronic primary-care records in the UK’s Clinical Practice Research Datalink (CPRD). Each case was propensity-score matched to up to five controls. The clinical features recorded before diagnosis were compared. The proposed clinical features shown to be associated with BAV (p < 0.05) were incorporated into a multivariable regression model. We identified 2898 cases. The prevalence of BAV in the CPRD was 1 in 5181, significantly lower than expected, suggesting that diagnosis and/or recording could be improved. The following biologically plausible clinical features were associated with a subsequent diagnosis of BAV: palpitations (OR: 2.86 (95% CI: 1.60, 3.16)), atrial fibrillation (AF) (OR: 2.25 (95% CI: 1.60, 3.16)) and hypertension (OR: 1.72 (1.48, 2.00)). The best model had an AUC of 0.669 (95% CI: 0.658 to 0.680), a positive predictive value (PPV) of 5.9% (95% CI: 4.0% to 8.7%) and a negative predictive value (NPV) of 99% (95% CI: 99% to 99%) at a population prevalence of 1%. This study indicates that palpitations, hypertension and AF should trigger a clinical suspicion of BAV and assessment via echocardiography. It also demonstrates the potential to develop a prediction model for BAV to stratify patients for echocardiography screening.

Citation

Evans, W., Akyea, R. K., Weng, S., Kai, J., & Qureshi, N. (2022). Identifying Patients with Bicuspid Aortic Valve Disease in UK Primary Care: A Case–Control Study and Prediction Model. Journal of Personalized Medicine, 12(8), Article 1290. https://doi.org/10.3390/jpm12081290

• Journal Article Type: Article
• Acceptance Date: Aug 2, 2022
• Online Publication Date: Aug 5, 2022
• Publication Date: Aug 5, 2022
• Deposit Date: May 3, 2023
• Publicly Available Date: May 11, 2023
• Journal: Journal of Personalized Medicine
• Electronic ISSN: 2075-4426
• Publisher: MDPI
• Peer Reviewed: Peer Reviewed
• Volume: 12
• Issue: 8
• Article Number: 1290
• DOI: https://doi.org/10.3390/jpm12081290
• Public URL: https://nottingham-repository.worktribe.com/output/9906903
• Publisher URL: https://www.mdpi.com/2075-4426/12/8/1290