P3794MOODS: a novel risk score to identify patients with atrial fibrillation and sleep apnoea

Abstract

Background Sleep-disordered breathing (SDB) is an important risk factor for developing atrial fibrillation (AF), and treatment of concomitant SDB can improve AF rhythm outcomes. Diagnosis of SDB requires sleep studies which can pose a significant time and resource burden. We sought to develop a prediction score based on clinical characteristics that can help identify AF patients who require further assessment for SDB. Methods Prospectively-collected data for 442 consecutive patients treated for AF from 2009 to 2017 were analysed. All patients were considered candidates for rhythm-control and therefore referred for sleep studies. The diagnosis of SDB was confirmed using in-lab polysomnography and classified using the apnoea-hypopnoea-index (AHI), with cut-offs of ≥15/hr and ≥30/hr indicating moderate-to-severe and severe SDB respectively. Patients treated up to 2015 formed the derivation cohort (n=311) and the remainder (n=113) formed the validation cohort. Multivariate logistic regression analysis was used to identify clinical variables predictive of moderate-to-severe SDB. A risk score model was developed based on regression coefficients and tested using receiver-operating-characteristics analyses on the validation cohort. Results Overall, mean age was 60±11 years, mean body mass index (BMI) was 30±5 kg/m2 and 69% were men. The prevalence of moderate-to-severe SDB was 33.7%. There were no significant differences in baseline characteristics between the derivation and validation cohorts. Male gender (score=1), overweight (BMI: 25–29 kg/m2, score=2), obesity (BMI≥30 kg/m2, score=3), diabetes (score=1), and stroke (score=2) were significantly independently predictive of moderate-to-severe SDB and formulated the score. The score performed well to predict moderate-to-severe SDB with a C-statistic of 0.73 (95% CI: 0.67–0.79, P<0.001) in the derivation cohort, and 0.67 (95% CI: 0.57–0.77, P<0.001) in the validation cohort. As a rule-out test, a score of ≤3 had a negative predictive value of 77% for moderate-to-severe SDB (91% for severe SDB). A score of ≥4 had an intermediate positive likelihood ratio (PLR) of 2 for moderate-to-severe SDB (2.2 for severe SDB), while a score of ≥5 had a high PLR of 6.5 and 6.8 for moderate-to-severe SDB and severe SDB respectively. Sensitivity and specificity table Conclusion A novel risk score comprising clinical characteristics can identify patients with AF likely to benefit from further assessment for SDB. Application of this model may aid optimise resource utilisation and facilitate timely patient care.

Cardiorespiratory Polygraphy for Detection of Obstructive Sleep Apnea in Patients With Atrial Fibrillation

“…OSA screening questionnaires and scales are sometimes used in the general population or specific patient populations (e.g., surgical patients), but performance is insufficient to screen for OSA in AF patients. ( 8 – 10 ) Currently, polysomnography (PSG) is the accepted gold standard for diagnosing OSA. Additionally, OSA diagnosis by PSG is mandatory before starting CPAP therapy in several countries ( 11 ).…”

Predictores de apnea del sueño en pacientes con fibrilación auricular

“…En el estudio original de validación de la escala MOODS, realizada sobre 442 pacientes con fibrilación auricular, la prevalencia de AOS moderada-severa fue de 33.7% 26 . En el presente estudio los hallazgos son consistentes con una prevalencia de AOS moderada a severa del 54.8%, mayor que lo comunicado en otras experiencias europeas y norteamericanas 25,26,29 . Es digno de mencionar que un 10.9% de los pacientes no recibían un tratamiento anticoagulante, a pesar de un elevado riesgo de cardioembolia.…”

“…La puntuación MOODS obtuvo su mejor relación S/Sp con tres componentes, a diferencia de la comunicación inicial de Kadhim y cols. 25 en que requirió de cuatro componentes. La identificación clínica de ronquidos demostró ser un buen predictor, aun después de ajustar para múltiples otras variables.…”

“…El BQ se definió de alto riesgo cuando se completaron 2 de 3 categorías. La puntuación MOODS (sexo, sobrepeso, obesidad, diabetes y stroke) se cotejó con los resultados de la poligrafía respiratoria (PR) 25 .…”

“…En general, se aconseja una evaluación clínica detallada y utilizar cuestionarios validados en laboratorios de sueño para identificar aquellos con alta probabilidad. Algunos de los más utilizados son el cuestionario de Berlín 22 (BQ), el cuestionario STOP-BANG 23 (SBQ), la escala de somnolencia de Epworth 24 (ESS) y más recientemente la puntuación MOODS 25 . Sin embargo, estos cuestionarios de recolección de datos clínicos han sido poco estudiados en poblaciones específicas y lo habitual es que los pacientes con FA no informen síntomas relacionados con AOS 26 .…”