Reliability and validity of the Japanese version of the Ullanlinna Narcolepsy Scale and Swiss Narcolepsy Scale for screening Japanese individuals with narcolepsy type 1.

OBJECTIVE

To validate the Japanese versions of the Ullanlinna Narcolepsy Scale (J-UNS) and Swiss Narcolepsy Scale (J-SNS) for screening narcolepsy in the Japanese population and to discuss strategies for their use in hypersomniac individuals.

METHODS

We selected 451 outpatients with excessive daytime sleepiness (EDS) already diagnosed according to the International Classification of Sleep Disorders third edition. They responded to both scales twice at 1-month intervals. After eliminating individuals who met the exclusion criteria, validity and reliability analyses were performed on 408 and 381 participants, respectively.

RESULTS

Patients with narcolepsy type 1 (NT1) displayed higher J-UNS and lower J-SNS scores than those with NT2 and other sleep disorders. The intraclass correlation coefficients and weighted κ coefficient for scale scores in the total participants and patients with NT1 were ≥0.70 and ≥ 0.40, respectively, indicating high reliability. Furthermore, both the sensitivity and specificity of these scales upon using the original cut-off scores (14 for UNS and 0 for SNS) for detecting NT1 were 0.70 or ≥0.70, suggesting high validity. Additionally, the receiver operating characteristic curve analysis revealed that the best cut-off score did not change for the J-SNS but that for the J-UNS, it increased to 18. In our study, the scale's sensitivity and specificity changed from 96% to 82% and 58%-78%, respectively.

CONCLUSIONS

Both scales revealed satisfactory screening abilities for NT1 in the Japanese population. However, it may be better to use J-UNS cut-off scores of 18 for a population with EDS.