The diagnostic value of sleep and vigilance tests in central disorders of hypersomnolence

Autor: Johannes Mathis, Daniel Andres, Wolfgang J Schmitt, Claudio L Bassetti, Christian W Hess, David R Schreier
Rok vydání: 2022
Předmět:
Zdroj: Sleep
DOI: 10.48350/174346
Popis: Study Objectives This retrospective cross-sectional observational study explored the diagnostic value of selected sleep and vigilance tests (SVT) beyond the multiple sleep latency test to differentiate between various central disorders of hypersomnolence (CDH) and fatigue syndromes. Methods Data from patients who underwent the multiple sleep latency test and at least one additional SVT were extracted from the Bern sleep database (1997–2018). One thousand three hundred fifty-two patients with a CDH (106 narcolepsy type 1, 90 narcolepsy type 2, 119 idiopathic hypersomnia, 192 nonorganic hypersomnia, 205 insufficient sleep syndrome), fatigue syndromes (n = 183), and a subgroup of patients with sleep apnea (n = 457) were analyzed. Classification based on SVT parameters was compared with the final clinical diagnosis serving as a reference. Results An overall model predicted the final diagnosis in 49.5% of patients. However, for the pairwise differentiation of two clinically suspected diagnoses, many SVT parameters showed a sensitivity and specificity above 70%. While the overall discrimination power of the multiple sleep latency test was slightly better than the one of the maintenance of wakefulness test, the latter differentiated best between narcolepsy and idiopathic hypersomnia with prolonged sleep need. Disproportionally poor results in reaction tests (e.g. steer clear test), despite comparable or lower sleepiness levels (SLAT, WLAT), were valuable for differentiating nonorganic hypersomnia from idiopathic hypersomnia/sleep insufficiency syndrome. Conclusion This study demonstrates how the combination of a careful clinical assessment and a selection of SVTs can improve the differentiation of CDH, whereas it was not possible to establish an overall prediction model based on SVTs alone.
Databáze: OpenAIRE