Prospective evaluation of a Holter-ECG derived severity index for screening of sleep disordered breathing

To prospectively evaluate the applicability of a method to screen overnight Holter-ECGs for sleep disordered breathing (SDB) in an unselected clinical routine Holter sample.

Holter-ECG recordings in 50 cardiologic inpatients were complemented with nocturnal respiratory polygraphy (PG). The respiratory event index (REI) and apnea/hypopnea index (AHI) from the PG served as a reference for an ECG-derived SDB severity estimate using a previously developed method. Agreement with the PG was investigated using Bland-Altman plots color-coded by ectopy level, and screening accuracy for REI≥15/h and AHI≥15/h was assessed.

Prevalence for REI≥15/h was 52%, and 32% for AHI≥15/h. We observed better agreement of the ECG-based estimate with the REI compared to the AHI. Ectopy did not limit the detection of SDB. Binary screening for REI≥15/h provided excellent specificity of 0.96 with a sensitivity of 0.77. Ternary screening for AHI≥15/h yielded 16% borderline classifications and specificity/sensitivity of 0.96/0.86 for the remaining data.

Screening of routine Holter-ECGs for sleep disordered breathing is reasonable and promises earlier identification of a significant part of patients at no additional cost.