Multiple Sleep Latency Test

Transcription

1 Schwerpunkt Somnologie :28 33 DOI 1.17/s Published online: 9. Januar 213 Springer-Verlag Berlin Heidelberg 213 C. Sauter H. Danker-Hopfe Competence Center of Sleep Medicine, CC1, Charité Universitaetsmedizin, Campus Benjamin Franklin, Berlin Multiple Sleep Latency Test Open questions about methods for the assessment of sleepiness Prevalence of excessive daytime sleepiness Data on the prevalence of excessive daytime sleepiness in adults vary strongly due to the diverse definitions and terms (e.g., daytime sleepiness, excessive daytime sleepiness, hypersomnia, fatigue) and are dependent on the sample characteristics. The prevalence rate of the occurrence of feeling (extremely) sleepy during the day or wishing to go to sleep vary between 4 and 26% of adult populations [21, 26]. Hypersomnia is supposed to affect 2 % of adults [4]. Therefore excessive daytime sleepiness and/or hypersomnia are most relevant medical conditions, which should be diagnosed early, quantified and efficiently treated. Definition of sleepiness Sleepiness is a physical and psychological state in humans and animals, which reflects the need for sleep. The reason for the occurrence of sleepiness or sleep need can be very different: besides the physiological normal sleep need at certain times during the 24-h sleep wake cycle, excessive sleepiness may be a sign of nonrestorative sleep, sleep wake disorders or other medical conditions. The need to test sleepiness, to quantify or predict its occurrence especially in situations where it is life-threatening (e.g. driving a car) is a major challenge in sleep medicine. In the International Classification of Sleep Disorders, Second Edition (ICSD-2), Daytime sleepiness is defined as the inability to stay awake and alert during the major waking episodes of the day, resulting in unintended lapses into drowsiness or sleep (AASM, American Academy of Sleep Medicine; [1]; p. 79). Excessive daytime sleepiness (EDS) is a symptom, which is often used synonymously with the diagnosis of the broader symptom hypersomnia, which includes prolonged night-time sleep, unplanned daytime sleep, and an inability to remain awake when required [2]. All these different but similar terms mirror the first open question on how to define sleepiness, since the operationalization of the measurement relies on the concept or definition. One of the approaches to measure sleepiness is to determine sleep latency, i.e., the time to fall asleep or time of remaining awake. The idea is that the sleepier a person is the more likely he or she is going to fall asleep, which seems plausible. On the other hand, it has been criticized that falling asleep depends on a person s sleepability, since subjects with a high sleepability, who are able to relax, do fall asleep easily without being sleepy [13]. Unpublished results from our own laboratory (. Fig. 1) show that healthy subjects, who are able to fall asleep quickly in an adopted version of the Multiple Sleep Latency Test (MSLT; [7]), are able to remain awake without any problems in the Maintenance of Wakefulness Test [18], and vice versa. Question 1 Number of trials without sleep MSLT MWT Subject Fig. 1 8 Results of a previous study on a modified Multiple Sleep Latency Test (MSLT) and Maintenance of Wakefulness Test (MWT) in 12 healthy nonsleep-deprived subjects, duration of each test 1 min. Number of trials without sleep on 9 MSLT and 9 MWT trials are shown. Tests took place on two separate days and were performed every 2 h from 7: AM until 11: PM 28 Somnologie - Schlafforschung und Schlafmedizin 1 213

2 Tab. 1 Patient characteristics and SOREMP scorings of laboratories: information on sex and gender and number of laboratories, in which more than two SOREMPs were indicated on the evaluation sheet are shown Patient Age Sex s Scoring of laboratories MB 29 Male MB1 MB 2 SOREMPs: 11 MZ 2 Male MZ1 MZ 2 SOREMPs: 27 SL 39 Male SL1 SL 2 SOREMPs: 1 SR 2 Female SR1 SR4 2 SOREMPs: 34 in accredited sleep laboratories has implications on the reliability of results with regard to its comparison with reference data derived from tests performed in a standardized way [1]. The implementation and assessment of multiple sleep latency tests according to published guidelines was therefore strongly recommended. Question 2 Which test procedure(s) should be used? Quantification of sleepiness: multiple sleep latency test and maintenance of wakefulness test To date, no single agreed upon marker or direct physiological indicator for the quantification of sleepiness is available. The attempts to measure sleepiness are still only an approach or indirect measure of the underlying actual sleepiness. There are different subjective and objective methods, which are widely used, but none of them is superior to the others (for overview see [17]). The diversity of these methods is also reflected in the different types of problems occurring when applying these techniques. It is supposed that the EEG-based techniques Multiple Sleep Latency Test (MSLT; [6]) and Maintenance of Wakefulness Test (MWT; [18]) are the gold standards, although they have several shortcomings, which will be discussed in the context of the surveys on the implementation of MSLT and MWT in DGSM-accredited sleep laboratories. In the ICSD-2 [1], MSLT and MWT are recommended for the quantification of daytime sleepiness, as well as the Epworth Sleepiness Scale [1]. In this article we will concentrate on the EEG-based measures MSLT and MWT. In 2 the AASM published a comprehensive review on The clinical use of the MSLT and MWT [3], and a related article on Practice Parameters for Clinical Use of the Multiple Sleep Latency Test and the Maintenance of Wakefulness Test [16]. These two articles were based on the need for universally valid guidelines for the conduction and interpretation of the MSLT and MWT. German versions of the recommendations for the MSLT [11] and MWT [2] protocols have been published in this journal. In general, the targets and main outcomes of MSLT and MWT are sleep latencies and especially in the case of the MSLT the number of Sleep Onset REM Periods (SOREMPs), which arise from REM sleep latencies. A mean sleep latency of 8 min and 2 SOREMPs are required for the diagnosis of narcolepsy [1, 17]. Arand et al. [3] reported a sensitivity of.78 and a specificity of.93 for 2 SOREMPs for the diagnosis of narcolepsy. Question 3 Are the MSLT and MWT, which are conducted in German sleep laboratories, conducted in the same standardized way? Results of the German questionnaire surveys on MSLT and MWT in DGSM-accredited laboratories In Germany, two surveys on the current practice of performing the MWT in DGSM-accredited laboratories or sleep clinics have shown an increasing percentage of application from 24.6% in 24 [11] to 4.9% in 29 [2]. The MSLT is performed much more often than the MWT: 82.1% applied the MSLT in their laboratories [11]. Both surveys revealed a high variability in MSLT and MWT protocols. Discrepancies were observed regarding the following: F the suspected diagnoses or indications for MSLT/MWT, F the number of test sessions per patient, F the time schedules, F the EEG montages, F the length of epochs, F the definitions of sleep and REM latencies, F criteria of test termination, and F interpretation of the result as pathological finding. The described heterogeneity in the implementation and assessment of MSLTs Question 4 Is the MSLT, which is conducted in German sleep laboratories, evaluated and interpreted in a standardized way? German study on interrater reliability of MSLT scorings Focus on REM latency and SOREMPs In a next step (sleep) stage scorings according to Rechtschaffen and Kales [22] and the outcome of this staging concerning sleep latency and possible sleep onset REM phases were studied. In May 28, a CD with altogether 19 MSLT recordings from four patients was sent to 31 DGSMaccredited sleep laboratories. The MSLTs were recorded by the sleep laboratory Klingenmuenster (Head: Dr. Hans-Günter Weeß) specifically for the purpose for the present study. All recordings had a total length of 3.±1. min or 7±3 3-s epochs in order to prevent participants from drawing inferences from the duration of the tests about the sleep and/or REM latencies. The laboratories were asked to report sleep latencies (of individual MSLT recordings and patient-specific means) as well as REM latencies and the number of observed sleep onset REM phases (SOREMPs) on an evaluation sheet and to rate the degree of daytime sleepiness for all patients (no, mild, moderate, severe). Furthermore, the laboratories were asked to send the results of their scorings by epochs in EDF (European Data Format) files. Alternatively the results of the scoring by epochs could be sent as paper print out. The only information that was provided about the patients were age and sex. Characteristics of patients are shown in. Tab. 1. For the present paper we focus on problems resulting from REM sleep analysis. The data on sleep latencies and scoring of sleep stages other than REM, as well on the classification of daytime sleepiness will be presented separately. Somnologie - Schlafforschung und Schlafmedizin

3 Abstract Zusammenfassung Somnologie :28 33 DOI 1.17/s Springer-Verlag Berlin Heidelberg 213 C. Sauter H. Danker-Hopfe Multiple Sleep Latency Test. Open questions about methods for the assessment of sleepiness Abstract In the present paper, the assessment of sleepiness in general and the status quo in Deutsche Gesellschaft für Schlafforschung und Schlafmedizin (DGSM; German Sleep Society)-accredited sleep laboratories are examined. The first question deals with the definition of sleepiness; the second question refers to different methods of assessing sleepiness with a special focus on the Multiple Sleep Latency Test (MSLT), which is considered the gold standard to measure sleepiness. The main outcomes are mean sleep latency and the number of sleep onset REM periods (SOREMPs). Two or more SOREMPs are indicative for the diagnosis of narcolepsy; therefore, correct scoring and interpretation are relevant. The third question is related to the implementation of the MSLT in DGSM-accredited laboratories and whether the MSLT is performed in compliance with the latest guidelines. A questionnaire survey on implementation of the MSLT revealed that 82% apply the MSLT. Questions on the protocol and interpretation of MSLTs indicated that the practice is inhomogeneous. Results of a scoring study, in which DGSM-accredited laboratories were asked to score in total 19 MSLT sessions of 4 different patients, indicated that not only the performance of MSLTs but also evaluation varies across laboratories. For the present article, we focus on the results of the REM sleep parameters scored and reported by the laboratories, and discuss possible reasons for the relatively high heterogeneity between the laboratories, and in some cases differences between the scoring of REM and the reported REM latency and/or SOREMPs. While the scoring of REM per se seems difficult and might not be adequately detected, there is no generally applicable definition of SOREMPs. In several guidelines an observation period of 1 min after the onset of sleep is recommended, while in the German Leitlinie S3 1 min is proposed. Since the length of the recording period has consequences for the definition of SOREMPs, the apparently small difference of min can have detrimental effects on the diagnosis of narcolepsy. Keywords Excessive daytime sleepiness Sleepiness MSLT REM latency SOREMP Multipler Schlaflatenztest. Offene Fragen zu Methoden zur Erfassung der Schläfrigkeit Zusammenfassung In diesem Artikel geht es um die Untersuchung der Schläfrigkeit im Allgemeinen und den Ist-Zustand in von der Deutschen Gesellschaft für Schlafforschung und Schlafmedizin (DGSM) akkreditierten Laboren. Die erste Frage beschäftigt sich mit der Definition der Schläfrigkeit, die zweite mit Methoden zur Erfassung der Schläfrigkeit mit Schwerpunkt auf dem Multiplen Schlaflatenztest (MSLT): Der MSLT gilt als Goldstandard zur Messung von Schläfrigkeit. Die wichtigsten Parameter sind die mittlere Schlaflatenz und die Anzahl der Sleep-Onset-REM-Phasen (SOREMP). Zwei oder mehr SOREMP sprechen für die Diagnose der Narkolepsie. Daher ist die korrekte Auswertung und Interpretation für die Diagnose und Behandlung relevant. Die dritte Frage bezieht sich auf die Implementierung des MSLT in DGSM-akkreditierten Laboren und darauf, ob dort die Durchführung des MSLT nach den aktuellen Richtlinien erfolgt. Eine Fragebogenstudie ergab einen relativ hohen Prozentsatz von 82%, die den MSLT anwenden. Fragen zur Durchführung und Interpretation des MSLT zeigten inhomogene Vorgehensweisen. Eine Auswertestudie, in der Labore gebeten wurden, 19 MSLT-Sitzungen von vier verschiedenen Patienten auszuwerten, ergab, dass nicht nur die Durchführung des MSLT, sondern auch die Auswertung zwischen den Laboren variiert. Im vorliegenden Artikel liegt der Schwerpunkt auf den Laborergebnissen der REM-Schlaf-Parameter und den Unterschieden dabei, unter anderem auch zwischen den Scorings von REM und der daraus resultierenden REM-Latenz und/oder SOREMP. Einerseits scheint die Auswertung von REM per se schwierig, und REM wird möglicherweise nicht korrekt erfasst. Auf der anderen Seite gibt es keine allgemein gültige Definition zu SOREMP. In einigen Richtlinien wird ein Beobachtungszeitraum von 1 min nach dem Einschlafen empfohlen, in der deutschen Leitlinie S3 aber 1 min. Da die Dauer einer Aufzeichnung Konsequenzen für die Definition der SOREMP hat, kann die scheinbar geringe Differenz von min nachteilige Auswirkungen auf die Diagnose einer Narkolepsie haben. Schlüsselwörter Exzessive Tagesschläfrigkeit Schläfrigkeit MSLT REM-Latenz SOREMP Results Altogether 4 laboratories (response rate 14.%) participated. The data on REM sleep latencies and on the number of SOREMPs on the evaluation sheets were compared to the occurrence of REM in the scorings by epoch, and the REM sleep latency was calculated from these scorings. Depending on the principle of majority rule a consensus scoring was derived from all scorings. The REM latencies which were calculated from the first epoch of sleep to the occurrence of REM based on the scorings sent by the laboratories are depicted in. Fig. 2. The results underline a large variability in scoring results. Reported and scored REM sleep latencies ranged from 31 min. The correlation between the scored and indicated REM sleep latency on the evaluation sheet was rather high (r=.8987, n=3, p<.1). In. Fig. 3, it can be seen that if the scoring of REM did not match the REM latency reported in the evaluation sheets, laboratories tended to overestimate REM sleep latencies, i.e., reported a later REM onset than the one calculated from their scorings. The kappa (κ) coefficient for agreement between occurrence of s and reporting of REM on the evaluation sheets for all trials and all laboratories was.781, the agreement in percent was 89.6%. The heterogeneity was also apparent in the number of observed SOREMPs (. Fig. 4). 3 Somnologie - Schlafforschung und Schlafmedizin 1 213

4 REM sleep latency (min) REM sleep latency (min) Scoring based REM sleep latency (min) Patient MB Reported REM sleep latency (min) In two (patient MZ and patient SL) out of the four patients the whole range from SOREMPs was observed. There was a clear maximum in the distribution in three patients: 74.4% of laboratories observed 3 SOREMPs in patient MZ, 6.8% of laboratories classified no SOREMPs in patient SL, and 48.7% of laboratories indicated three sessions containing SOREMPs in patient MZ. The largest differences in the evaluation of SOREMPs occurred in patient MB. Results in this patient varied between no SOREMP (42.1% of answers), SOREMP in one session (28.9%), in two (21.1%) or three sessions (7.9%). REM sleep latency (min) REM sleep latency (min) Patient: MZ Patient: SL Patient: SR Fig. 2 8 Distribution of REM sleep latencies derived from scorings by epoch by patient and sessions Fig. 3 9 Scatterplot of reported REM sleep latencies and REM sleep latencies derived from scorings by epoch The reasons for the observed inhomogeneity could be varying definitions in the scoring of SOREMPs in the MSLT. This becomes obvious, when the number of sessions, in which REM sleep was scored, is compared to number of SOREMPs (. Fig. ). In 26.2% of cases the number of SOREMPs was indicated with, and the scorings contained no REM sleep. In 49.4% the number of SOREMPs was in accordance with the number of sessions in which REM sleep was scored, independently from the duration of the REM latency. In about one quarter of cases (24.4%), there are discrepancies between the statements on the evaluation sheet and the corresponding scorings. For instance in 6.7% of cases no SOREMP was stated although REM sleep was scored in single sessions and REM sleep latencies were indicated on the evaluation sheets. In 1.9% of cases the number of reported SOREMPs is smaller than the number of sessions, in which REM sleep was scored. Therefore, it is most likely that laboratories use different definitions of SOREMPs or REM sleep latencies (e.g., <1 min, <1 min, <2 min). In three cases (1.8%) the number of SOREMPs is higher than the number of test sessions, in which REM sleep was scored. We do not have an explanation for that finding, except for the highly speculative hypothesis that REM sleep was interrupted by epochs of other sleep stages, continued again, and was therefore counted more than once per session. The number of laboratories that indicated at least two SOREMPs on the evaluation sheet are displayed in the right column of. Tab. 1. Discussion Measuring sleepiness remains a challenge. The results of the MSLT scoring study documented large variability in the analysis and interpretation of MSLTs. For all four patients substantial differences concerning the reported number of sleep onset REM phases were observed. This large variability was confirmed by scorings based on epochs, and by differences in the agreement as assessed by percentages. The MSLT trials used had been recorded by the sleep laboratory Klingen muenster specifically for this purpose, and although some of the recordings were difficult to score, the unfamiliar software/screen presentation might have contributed to the heterogeneity of results. Nevertheless, the varying results on SOREMPs between some of the laboratories brings up the question whether the definition of SOREMPs is unclear or handled differently between laboratories. This is especially important, since two or more SOREMPs are required for the diagnosis of narcolepsy [1]. In the survey on the implementation of the MSLT in accredited laboratories [11], Somnologie - Schlafforschung und Schlafmedizin

5 Schwerpunkt MB 29 yrs male it became clear that sleep laboratories defined REM sleep latency differently: whereas.1% out of 18 laboratories answering the question on the definition of REM sleep latency stated that the time elapsed from beginning of sleep or NREM1 until the occurrence of REM corresponds with REM latency, 8.2% calculated REM sleep latency from lights off to the first epoch of REM sleep. The latter might explain some of the results from laboratories, which overestimated REM sleep latency in their reports (. Fig. 3). In nearly all 19 MSLT trials, REM sleep latencies for one and the same session were reported either with a length <1 min, between 1 and 1 min, or above 1 min (. Fig. 2). These seemingly small differences result in different numbers of SOREMPs, depending on the SOREMP definition applied by the laboratories. Already in the first publications on the MSLT, the definition of sleep onset REM sleep was defined differently: Richardson et al. ([24]; p. 621) described sleep onset REM sleep periods as follows: A REM sleep period that occurs 1 min after sleep begins and constitutes REM sleep with an abnormally short latency. Mitler et al. ([19]; p. 48) defined abnormality with respect to REM sleep in the MSLT if REM sleep was preceded by less than 1 min of NREM sleep and so easily qualifies as sleep onset-rem-sleep. In another publication on the MSLT, SOREMPS were likewise defined as REM onset SL 39 yrs male 3 SR 2 yrs female Number Sleep Onset REM Number Sleep Onset REM MZ 2 yrs male 19 Fig. 4 8 Histograms of the distribution of Sleep Onset REM periods (SOREMP) in the MSLT for four different patients (MB, MZ, SL, SR) within 1 min of sleep ([23]; p. 443). The guidelines for the MSLT, which were published in 1986 ([8]; p. 21), recommended a termination of the MSLT after 1 min from the first sleep epoch according to Rechtschaffen and Kales [22]: In order to assess the occurrence of REM sleep in subjects or patients the test should continue for 1 min after the first epoch of sleep. The list of different definitions could easily be extended. The different time limits for termination of the MSLT might be one source of confusion and result in different SOREMP definitions. More recently, the guidelines of the AASM ([16]; p. 116) suggest that SOREMPs are defined as the first epoch of REM sleep at any time during the nap trial. The authors furthermore state in the MSLT protocol ([16]; p. 119): In order to assess for the occurrence of REM sleep, in the clinical MSLT the test continues for 1 min from after the first epoch of sleep. On the contrary, in the German S3 Leitlinie, SOREMPs are defined as the occurrence of REM sleep within 1 min or less after sleep onset ([17]; p. 8). This corresponds with the ICSD [2], but a precise definition is missing in the 2nd edition [1]. Therefore, it is hardly surprising that the scoring and calculation of SOREMPs were different between centers. Furthermore, the test duration of 3 min, regardless of the occurrence of sleep, might have led to confusion in the scoring of REM sleep. Another aspect to consider is the interrater reliability for the scoring of the sleep latency (min), since REM latency depends on the beginning of sleep. Results of the questionnaire survey on the implementation on the MSLT showed a high variability of the definition of sleep latencies in the scoring of MSLTs between laboratories [11]. A study on interrater reliability of two independent scoreres and one consensus scoring revealed an intraclass correlation for the parameter sleep latency in night sleep recordings according to Rechtschaffen and Kales [22] of.731 and for six independent scorings according to the AASM standard [14] of.799 [1]. In the study of Danker-Hopfe et al. [1], the intraclass correlations for the Rechtschaffen and Kales scorings were lowest for the time (%) spent in stage 1 (.414) and stage 2 (.419). In the case that the AASM standard was applied, the intraclass correlation was.7423 for N1 and.3373 for N2. These observed differences between scorers and between scoring standards might also be relevant in the scoring of MSLTs in the present study. Above all, there was no instruction given to the laboratories whether they should apply the Rechtschaffen and Kales or AASM standard. Nevertheless, this should have no consequences for the scoring of sleep or/and REM latencies. For the present study, these parameters were not looked at separately. Some extreme cases revealed the short coming of the MSLT in terms of an objective, standardized method: although standard scorings rules and guidelines for the scoring and interpretation are available, the comparability between laboratories is far from satisfactory. Other studies on the interrater reliability showed excellent inter- and intrarater reliability when comparing only three different trained staff member [9]: the interrater reliability for number of SOREMPs in naps of 44 patients was depicted as a correlation coefficient and ranged between.884 and.936. In another study on MSLT, interrater reliability of three raters, who scored MSLTs from 21 patients, showed satisfactory results, with fair to good agreement for SOREMP (κ coefficient for the absence or presence of REM varied between for each nap, []). In a third study on scoring reliability in MSLT, four clinical and seven technical polysomnog- 32 Somnologie - Schlafforschung und Schlafmedizin 1 213

6 Number of SOREMPs = 49.4 % Number of SOREMPs = = Number of SOREMPs 26.2 % Number of SOREMP > Number of SOREMP < 1.9 % 24.4 % Number of SOREMP unspecified but 6.7 % % Fig. 9 Correspondence between the number of SOREMP in the evaluation sheets and actual REM sleep scorings: frequencies are calculated across all MSLT trials and laboratories raphers were involved in the scoring of MSLTs from 192 patients [12]. After previously completing a sleep scoring training, they scored between 36 and 66 individual records. The coefficient for the mean number of REM onsets during the MSLT was.88 for the interrater reliability. The difference to our study is that all the others compared data from a substantial smaller number of scorers, who received the same training or/and came from the same laboratory. Conclusion There are problems in the performance, scoring, and interpretation of MSLTs which in some cases might be relevant with regard to the diagnosis. For the MSLT to be a reliable diagnostic tool it needs (1) strict adherence to a commonly agreed protocol, (2) a clear definition of SOREMPs, which is universally applied, and (3) well-trained staff for the performing, scoring, and evaluation of MSLTs. It is suggested that actions should be taken to ensure that sleep laboratories are not only able to perform MSLTs and MWTs correctly from the technical point of view but they perform these tests according to the standard protocol and that they are able to score them with a given quality. Corresponding address Dr. C. Sauter Competence Center of Sleep Medicine, CC1, Charité Universitaetsmedizin, Campus Benjamin Franklin Eschenallee 3, 14 Berlin Germany cornelia.sauter@charite.de Acknowledgment. We thank Hans-Günter Weeß of the sleep laboratory of Klingenmuenster who recorded the MSLTs specifically for this study, and all participating laboratories for their large effort in scoring and evaluating the recordings. We like to thank Andrea Rodenbeck and Peter Geisler for their really helpful input and comments. Conflict of interest. On behalf of all authors, the corresponding author states that there are no conflicts of interest. References 1. American Academy of Sleep Medicine (2) The International Classification of Sleep Disorders: diagnostic and coding manual. American Academy of Sleep Medicine, Westchester 2. American Academy of Sleep Medicine (21) The International Classification of Sleep Disorders: diagnostic and coding manual. American Academy of Sleep Medicine, Westchester 3. Arand D, Bonnet M, Hurwitz T et al (2) The clinical use of the MSLT and MWT. Sleep 28: Bassetti C, Gugger M (2) Hypersomnie Ätiologie, Klinik, Diagnostik und Therapie der exzessiven Schläfrigkeit. Ther Umsch 7: Benbadis SR, Qu Y, Perry MC et al (199) Interrater reliability of the multiple sleep latency test. Electroencephalogr Clin Neurophysiol 9: Carskadon MA, Dement WC (1982) The multiple sleep latency test: what does it measure? Sleep (Suppl 2): Carskadon MA, Dement WC (1977) Sleepiness and sleep state on a 9-min schedule. Psychophysiology 14: Carskadon MA, Dement WC, Mitler MM et al (1986) Guidelines for the multiple sleep latency test (MSLT): a standard measure of sleepiness. Sleep 9: Chen L, Ho CK, Lam VK et al (28) Interrater and intrarater reliability in multiple sleep latency test. J Clin Neurophysiol 2: Danker-Hopfe H, Anderer P, Zeitlhofer J et al (29) Interrater reliability for sleep scoring according to the Rechtschaffen & Kales and the new AASM standard. J Sleep Res 18: Danker-Hopfe H, Binder R, Popp R et al (26) Erhebung zur Praxis der Durchführung von Multiplen Schlaflatenztests (MSLT) in akkreditierten Schlaflaboren. Somnologie 1: Drake CL, Rice MF, Roehrs TA et al (2) Scoring reliability of the multiple sleep latency test in a clinical population. Sleep 23: Harrison Y, Horne JA (1996) High sleepability without sleepiness. The ability to fall asleep rapidly without other signs of sleepiness. Neurophysiol Clin 26: Iber C, Ancoli-Israel S, Chesson A et al (27) The AASM Manual for the scoring of sleep and associated events: rules, terminology and technical specifications, 1. edn. American Academy of Sleep Medicine, Westchester 1. Johns MW (1991) A new method for measuring daytime sleepiness: the Epworth sleepiness scale. Sleep 14: Littner MR, Kushida C, Wise M et al (2) Practice parameters for clinical use of the multiple sleep latency test and the maintenance of wakefulness test. Sleep 28: Mayer G, Fietze I, Fischer J et al (29) S3-Leitlinie. Nicht erholsamer Schlaf/Schlafstörungen. Somnologie 13: Mitler MM, Gujavarty KS, Browman CP (1982) Maintenance of wakefulness test: a polysomnographic technique for evaluation treatment efficacy in patients with excessive somnolence. Electroencephalogr Clin Neurophysiol 3: Mitler MM, Van Den Hoed J, Carskadon MA et al (1979) REM sleep episodes during the Multple Sleep Latency Test in narcoleptic patients. Electroencephalogr Clin Neurophysiol 46: Ohayon MM (28) From wakefulness to excessive sleepiness: what we know and still need to know. Sleep Med Rev 12: Ohayon MM, Priest RG, Zulley J et al (22) Prevalence of narcolepsy symptomatology and diagnosis in the European general population. Neurology 8: Rechtschaffen A, Kales A (1968) A manual of standardized terminology, techniques and scoring system for sleep stages of human subjects. University of California, Brain Information Service/Brain Research Institute, Los Angeles 23. Reynolds CF III, Coble PA, Kupfer DJ et al (1982) Application of the multiple sleep latency test in disorders of excessive sleepiness. Electroencephalogr Clin Neurophysiol 3: Richardson GS, Carskadon MA, Flagg W et al (1978) Excessive daytime sleepiness in man: multiple sleep latency measurement in narcoleptic and control subjects. Electroencephalogr Clin Neurophysiol 4: Sauter C, Danker-Hopfe H (21) Der Multiple Wachbleibetest (MWT). Somnologie 14: Zielinski J, Zgierska A, Polakowska M et al (1999) Snoring and excessive daytime somnolence among Polish middle-aged adults. Eur Respir J 14:946 9 Somnologie - Schlafforschung und Schlafmedizin