0721 Correlation between Oxygen Desaturation Index and Apnea Hypopnea Index for Diagnosis of Obstructive Sleep Apnea

Abstract

Abstract Introduction Obstructive sleep apnea (OSA) is diagnosed through polysomnography (PSG) which can be done in lab or outpatient. PSG is conventionally not approved by insurance companies’ in-patient which can result in delay in diagnosis and treatment of OSA. High resolution pulse oximetry (HRPO) done inpatient is easy to perform and calculates oxygen desaturation index (ODI) to asses’ nocturnal desaturations which solely, is insufficient for diagnoses of OSA according to current treatment guidelines. We hypothesize that there may be a correlation between ODI and apnea hypopnea index (AHI) which can facilitate in earlier diagnosis of OSA. Methods We conducted a retrospective chart review to compare patients who underwent HRPO resulting in a sleep medicine consult inpatient followed by polysomnography outpatient over a 2-year period at a tertiary care academic center. Demographic data, ODI, AHI and oxygen nadir levels were collected. Results Sixty-five patients (47 males, 18 females; mean age of 59.1 years) with suspected OSA underwent inpatient HRPO during their hospital stay, followed by a PSG in the outpatient setting. The strength of association between ODI and AHI was determined using a Pearson’s analysis after adjusting to a logarithmic scale. There was a statistically significant weakly positive association between ODI and AHI (Pearson correlation=0.33, p=0.008). Linear regression analysis demonstrated a predictive value of 0.419 (p=0.008) between AHI and ODI. However, there was no statistically significant predictive value between ODI and AHI when adjusted for age, sex, body mass index (BMI) and ethnicity (beta =0.169; p=0.330). This may be limited by small sample size. HRPO nadir oxygen saturation (NOS) also correlated with polysomnography NOS with a Pearson correlation coefficient of 0.353 (p=0.006). Linear regression analysis showed a predictive value beta of 0.149 (p=0.006). When adjusted for age, sex, BMI and race, beta was equal to 0.156 (p=0.007). Conclusion ODI calculated through HRPO may be correlated with AHI. NOS determined through HRPO is weakly positively predictive of NOS calculated through PSG. To initiate treatment of OSA sooner, HRPO may be considered for screening or diagnostic purposes. The correlation between ODI and AHI needs to establish further in randomized controlled setting. Support (If Any)