Development, validation, and proof-of-concept implementation of a two-year risk prediction model for undiagnosed atrial fibrillation using common electronic health data (UNAFIED)

Abstract

Background Many patients with atrial fibrillation (AF) remain undiagnosed despite availability of interventions to reduce stroke risk. Predictive models to date are limited by data requirements and theoretical usage. We aimed to develop a model for predicting the 2-year probability of AF diagnosis and implement it as proof-of-concept (POC) in a production electronic health record (EHR). Methods We used a nested case–control design using data from the Indiana Network for Patient Care. The development cohort came from 2016 to 2017 (outcome period) and 2014 to 2015 (baseline). A separate validation cohort used outcome and baseline periods shifted 2 years before respective development cohort times. Machine learning approaches were used to build predictive model. Patients ≥ 18 years, later restricted to age ≥ 40 years, with at least two encounters and no AF during baseline, were included. In the 6-week EHR prospective pilot, the model was silently implemented in the production system at a large safety-net urban hospital. Three new and two previous logistic regression models were evaluated using receiver-operating characteristics. Number, characteristics, and CHA2DS2-VASc scores of patients identified by the model in the pilot are presented. Results After restricting age to ≥ 40 years, 31,474 AF cases (mean age, 71.5 years; female 49%) and 22,078 controls (mean age, 59.5 years; female 61%) comprised the development cohort. A 10-variable model using age, acute heart disease, albumin, body mass index, chronic obstructive pulmonary disease, gender, heart failure, insurance, kidney disease, and shock yielded the best performance (C-statistic, 0.80 [95% CI 0.79–0.80]). The model performed well in the validation cohort (C-statistic, 0.81 [95% CI 0.8–0.81]). In the EHR pilot, 7916/22,272 (35.5%; mean age, 66 years; female 50%) were identified as higher risk for AF; 5582 (70%) had CHA2DS2-VASc score ≥ 2. Conclusions Using variables commonly available in the EHR, we created a predictive model to identify 2-year risk of developing AF in those previously without diagnosed AF. Successful POC implementation of the model in an EHR provided a practical strategy to identify patients who may benefit from interventions to reduce their stroke risk.

Development and Validation of 3‐Year Atrial Fibrillation Prediction Models Using Electronic Health Record With or Without Standardized Electrocardiogram Diagnosis and a Performance Comparison Among Models

“… 18 More recently, Grout et al developed an AF prediction model using 10 variables from the EHR (C‐statistic, 0.81). 19 All the prior EHR‐driven AF prediction models used clinical variables that mainly originated from diagnosis codes or simple body measurements. The models of the present study adopted the detailed information of 130 ECG diagnoses that were generated and transformed from the computerized ECG machine.…”

Perspectives on Sex- and Gender-Specific Prediction of New-Onset Atrial Fibrillation by Leveraging Big Data

“…Two studies utilized EHRs to develop AF prediction models ( 41 , 42 ). EHRs contain real-time, patient-centered data that are instantly available to patients and authorized healthcare providers.…”

“…Over the past years, there has been an exponential interest in using various big data sources to further improve the AF risk prediction beyond the traditional AF risk factors (23,24). Specifically, multiple studies have employed AI-enabled algorithms to evaluate new-onset AF prediction by leveraging various big data modalities including the clinical data, ECGs, EHRs, and wearable devices (23)(24)(25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35)(36)(37)(38)(39)(40)(41)(42). Some of these studies showed that AI-enabled AF prediction models performed similar to or better than established traditional AF prediction models (25,(27)(28)(29)(30).…”

“…Another study developed a model to predict new-onset AF over a 2-year time period with good discrimination (C-statistic of 0.81) using a 10-variable model compromised of covariates commonly available in the EHRs of 53,552 subjects. This study, however, did not compare the developed prediction model to previous validated AF risk prediction models (42). Another study evaluated the improvement in 5-year AF risk prediction when adding novel variables identified by ML to the CHARGE-AF enriched score (27).…”

Is machine learning the future for atrial fibrillation screening?

“…The best model was a stacking model that included deep learning (neural networks) with lasso regression. The most recent study 30 to develop an ML model to apply to electronic healthcare records compared only different logistic regression models. The healthcare records used were derived from the Indiana Network for Patient Care.…”