Testing Zero-Inflation in Binomial Regression Models with an Application to Electrocardiogram Monitoring on Atrial Fibrillation

Key Points

• Monitoring atrial fibrillation through binomial regression models reveals significant zero-inflation issues.
• Key evidence is that traditional models fail to account for zero-inflation, disrupting accurate predictions.
• Analysis utilizes binomial regression methods to assess the effectiveness of ECG monitoring in detecting irregular heartbeats.
• Findings underscore the necessity for incorporating zero-inflation in statistical models for better cardiac monitoring accuracy.