What is the interpretation of eAHI (effective Apnea-Hypopnea Index) in a sleep study?

eAHI (Effective Apnea-Hypopnea Index) Interpretation

The effective AHI (eAHI) is a metric reported by auto-titrating CPAP devices that estimates residual respiratory events during treatment, with an eAHI cutoff of 6-8 events/hour optimally distinguishing adequately treated patients from those with residual obstructive sleep apnea. 1

What eAHI Measures

• eAHI represents the device-calculated estimate of apneas and hypopneas per hour during CPAP therapy, derived from smart card data that records pressure, flow, and leak information 1
• This differs from diagnostic AHI, which is measured during polysomnography before treatment and represents the baseline severity of sleep-disordered breathing 2

Clinical Interpretation Thresholds

For Assessing Treatment Adequacy

• An eAHI ≤6 events/hour indicates optimal CPAP treatment with sensitivity of 0.92 (95% CI 0.76-0.98) and specificity of 0.90 (95% CI 0.82-0.95) for identifying patients without residual OSA 1
• An eAHI >6 events/hour suggests residual sleep-disordered breathing requiring clinical evaluation and possible CPAP adjustment 1
• For detecting more severe residual OSA (PSG AHI ≥10), an eAHI cutoff of 8 events/hour provides optimal accuracy with sensitivity 0.94 and specificity 0.90 1

Comparison to Diagnostic AHI Standards

• Normal diagnostic AHI is <5 events/hour, while OSA diagnosis requires AHI ≥5 with symptoms or ≥15 without symptoms 2, 3
• Mild OSA: AHI 5-14 events/hour 2
• Moderate OSA: AHI 15-29 events/hour 2
• Severe OSA: AHI ≥30 events/hour 2

Critical Pitfalls in eAHI Interpretation

Device Limitations

• eAHI cannot detect arousals since auto-CPAP devices lack EEG monitoring, potentially underestimating sleep fragmentation compared to polysomnography 2
• Large mask leaks can artificially elevate or reduce eAHI accuracy, requiring review of leak data alongside eAHI 1
• Home environment eAHI shows no significant bias compared to laboratory-based smart card estimates, validating its use for remote monitoring 1

When eAHI is Insufficient

• Patients with persistent symptoms despite eAHI <6 require full polysomnography to assess for arousal-based events (RERAs), central apneas, or other sleep disorders not captured by CPAP algorithms 2
• eAHI does not distinguish obstructive from central events, which have different treatment implications 2
• Scoring method variations significantly affect interpretation: different hypopnea definitions can alter AHI values by up to 70%, though this primarily affects diagnostic rather than treatment monitoring 3

Clinical Decision Algorithm

For patients on auto-CPAP therapy:

1. eAHI ≤6 events/hour + symptom resolution → Continue current therapy, routine follow-up 1

2. eAHI >6 events/hour OR persistent symptoms → Check for:

   • Excessive mask leak (review leak data) 1
   • Inadequate pressure settings
   • Poor adherence patterns
   • Consider in-laboratory PSG on CPAP to assess for residual events, arousals, or alternative diagnoses 2

3. eAHI >8 events/hour → High likelihood of significant residual OSA requiring intervention 1

Important Caveats

• eAHI accuracy was validated in patients with varying OSA severity (mean baseline AHI ranged from mild to severe across studies), supporting its use across the disease spectrum 1
• Bland-Altman analysis demonstrates good agreement between auto-CPAP eAHI and polysomnography AHI, with positive likelihood ratio of 9.6 for detecting residual disease 1
• The 40% of lean patients who benefit from CPAP treatment despite low desaturation-based AHI highlights that eAHI may miss arousal-predominant disease, as these devices cannot score RERAs 2