Utility of risk prediction models to detect atrial fibrillation in screened participants

Abstract

Aims: Atrial fibrillation (AF) is associated with higher risk of stroke. While the prevalence of AF is low in the general population, risk prediction models might identify individuals for selective screening of AF. We aimed to systematically identify and compare the utility of established models to predict prevalent AF.

Methods and results: Systematic search of PubMed and EMBASE for risk prediction models for AF. We adapted established risk prediction models and assessed their predictive performance using data from 2.5M individuals who attended vascular screening clinics in the USA and the UK and in the subset of 1.2M individuals with CHA2DS2-VASc ≥2. We assessed discrimination using area under the receiver operating characteristic (AUROC) curves and agreement between observed and predicted cases using calibration plots. After screening 6959 studies, 14 risk prediction models were identified. In our cohort, 10 464 (0.41%) participants had AF. For discrimination, six prediction model had AUROC curves of 0.70 or above in all individuals and those with CHA2DS2-VASc ≥2. In these models, calibration plots showed very good concordance between predicted and observed risks of AF. The two models with the highest observed prevalence in the highest decile of predicted risk, CHARGE-AF and MHS, showed an observed prevalence of AF of 1.6% with a number needed to screen of 63. Selective screening of the 10% highest risk identified 39% of cases with AF.

Conclusion: Prediction models can reliably identify individuals at high risk of AF. The best performing models showed an almost fourfold higher prevalence of AF by selective screening of individuals in the highest decile of risk compared with systematic screening of all cases.

Registration: This systematic review was registered (PROSPERO CRD42019123847).

Keywords: Atrial fibrillation; External validation; Risk prediction models; Selective screening; Stroke.

Figure 1

Flowchart.

Figure 2

Included predictors. An overview of predictors used in the eleven risk prediction models that were developed to predict atrial fibrillation.

Figure 3

Discriminative performance. Squares represent the AUROC curves in the analysis of all 2.5M participants and diamonds in 1.2M participants with CHA₂DS₂-VASc of two or more. The vertical bars represent the 95% CIs. The AUROC curves are based on the regression equation in 12 prediction models, and on the point chart for two prediction models.^, Values are provided in Supplementary material online, eTable 9.

Figure 4

Calibration plots. Calibration plots of the two risk prediction models with the highest observed prevalence of AF in the highest decile of predicted risk: CHARGE-AF and MHS.^, To construct the calibration plots, data of all 2.5M participants (top row) and 1.2M participants with CHA₂DS₂-VASc of two or more (bottom row) were used. Mean predicted risk against the observed risk of AF across deciles of predicted risk (after recalibration with adjusting the intercept) is shown. The boxes represent the mean predicted risk for each decile and the vertical lines represent the 95% confidence intervals. The dotted diagonal line indicates perfect calibration. Boxes above the diagonal line indicate underestimation of risk and below the diagonal line overestimation of risk. The prevalences and number of cases of each decile are provided in Supplementary material online, eTable 9.

Figure 5

Test characteristics. Graph showing the sensitivity and specificity and corresponding observed prevalence and number needed to screen to detect 1 participant with AF using the prediction model developed by Alonso et al. 2013 (left) and Aronson et al. 2018 (right). The squares and circles correspond to selective screening of participants in the highest decile and highest two decile of predicted risk, respectively.