Split-night polysomnography overestimates apnea-hypopnea index in high-risk professions

Introduction: According to the American Academy of Sleep Medicine (AASM) guidelines, split-night polysomnography (SN-PSG) is an acceptable alternative to full-night PSG (FN-PSG) and may be considered in patients with an apnea-hypopnea index (AHI) =20/hr within the first 2 hours of the study. While SN-PSGs are an accurate approximation of moderate-to-severe obstructive sleep apnea (OSA), there remains the potential to misclassify the severity of sleep disordered breathing. Risks associated with the misclassification of OSA severity may be significant in high-risk professions such as active duty service members (ADSMs). The purpose of our study was to determine the accuracy of split-night polysomnography (SN-PSG) in a cohort of ADSMs. Materials and Methods: We conducted a retrospective review of ADSMs undergoing FN-PSG with approval by our institution's Department of Clinical Investigation. FN-PSG data were processed using t-test, ANOVA, Chi-Squared, and logistical regression using JMP v12.0 to obtain partial-night data for the first 2 and 3 hours of recording. Significance was established with p-value less than 0.05. OSA severity was determined by calculating the AHI of each subject's FN-PSG and SN-PSG. Results: Three-hundred patients were included in the study. Overall 79% were male with a mean age of 37.6 ± 8.4 years and mean BMI of 28.5 ± 3.3 kg/m2. Of our cohort, 112 patients (37%) would have qualified for a SN-PSG, of which 94 (84%) were appropriately classified and 18 patients (16%) were misclassified. Conclusions: In the relatively young, non-obese ADSM population, the majority did not qualify for a SN-PSG. The 3-hour SN-PSG accurately determined OSA severity in those with moderate-severe OSA; however, some patients with mild OSA would have been misclassified which can result in unnecessary duty limitations. A SN-PSG may not be ideal for this population.