0898 Title: Test - Retest Reliability of the Multiple Sleep Latency Test in Non-cataplectic Central Hypersomnia Disorders

Abstract Introduction Differentiation of narcolepsy type II (NT2) and idiopathic hypersomnia (IH) is challenging in clinical practice. Diagnostic criteria rely on the multiple sleep latency test (MSLT), previously shown to have poor reproducibility in these disorders. We leveraged a large polysomnographic database to study test-retest reliability in non-cataplectic central hypersomnia disorders. Methods Patients evaluated at Cleveland Clinic Sleep Disorders Center from 2008-2021 with suspected central hypersomnia disorders excluding narcolepsy type I with >1 PSG/MSLT were included. Studies not meeting validity criteria (< 7hr sleep time on PSG, < 5 MSLT trials, use of REM suppressants/alerting medications within 14 days) were excluded. Diagnostic groupings included NT2, IH and Normal/Indeterminate based on ICSD-3. Reliability and correlation between tests was performed using intra-class and Spearman’s correlations, respectively. Data is presented as mean+SD or median [25th,75th percentiles] for continuous variables and N(%) for categorical variables. Results A total of 58 patients (62.1% female, age 31.9 yr [18.5, 48.6]) were included. Interval between MSLTs was 3.6 + 3.18 yr (same diagnosis on both tests: 3.24 + 3.15; changed diagnosis: 3.98 + 3.22). Diagnosis based on MSLT 1 was NT2 8(13.8%), IH 20(34.5%) and Normal/Indeterminate 30(51.7%). Overall, 33(56.9%) patients remained in the same diagnostic category including 3(5.2%) NT2, 12(20.7%) IH and 18(31.0 %) Normal/Indeterminate. Of 25(43.1%) cases changing diagnosis, 3(5.2%) changed from NT2 to IH and 2(3.4%) to Normal/Indeterminate; 4(6.9%) changed from IH to NT2 and 4(6.9%) to Normal/Indeterminate; 5(8.6%) changed from Normal/Indeterminate to NT2 and 7(12.1%) to IH. Median MSL on tests 1 and 2 were 10.8 [4.6,12.9] and 9.0 [4.5,13.1], respectively. Correlation was moderate MSL [0.59 (0.39, 0.73), p< 0.001] and fair for SOREMPs [0.32 (0.07, 0.53), p=0.015]. Intra-class correlation showed moderate reliability in test-retest for MSL [0.56 (0.36, 0.72)] and SOREMPs [0.54 (0.33, 0.70)]. Conclusion MSLT test-retest reproducibility in non-cataplectic central hypersomnias was moderate in a tertiary care clinic population. While more favorable than prior studies, these findings underscore diagnostic challenges that may contribute to misdiagnosis of patients with hypersomnia. Support (if any)