**Evaluation**

The reading movements are simultaneous eye movements, which are characterized by either internal or external amplitude deflections in the EOG. It occurs while reading a specific rhythm EOG, as the eyes "jump" at the end of the line to the next line start. The reading movements can be distinguishing well visually by small and big eye movements (Fig. 3). All reading movements were counted that occurred after a minimum interval of 3 seconds.

Fig. 3. On display are the reading movements of the left and right eye (LEOG and REOG) as a rhythmic, blue wave pattern. The reading movements occur during reading, when the eyes "jump" at the line end of the text (right) to line beginning (left). Large eye movements (e.g. view movements) are characterized by large amplitude fluctuations.

In the present study, the following variables were used and calculated: the average, the highest and the lowest reading frequency (read line per epoch), sleep latency (in minutes) and the number of read pages.

#### **Results**

The average reading rate of the patients (n = 75) was 7.0 +/- 3.5 lines per epoch. In healthy volunteers (n = 16) it was 9.4 +/- 4.0 lines per epoch. All healthy subjects were evaluated for daytime sleepiness than normal, since neither sleep onset tendencies nor decreasing reading frequencies were observed. In 32 of the 70 OSAS patients (45.5%), however, sleep latency was found within 60 minutes. Also the reading frequency decreased over time. Rühle and colleagues calculated for the first time, the sensitivity and specificity of the *Reading test*, finding a cut-off value of greater than 11 for a pathological daytime sleepiness (Rühle et al. 2007). The standardized *Reading test* achieved a sensitivity of 76.2% and a specificity of 66.7% (Erle et al. 2009).

#### **2.8 Conclusion**

126 Sleep Disorders

spontaneous, aloud reading of few sentences, if the patient was able to read 3-5 sentences correctly and fluently. About the intention and the period of reading, the patients were not

The reading movements are simultaneous eye movements, which are characterized by either internal or external amplitude deflections in the EOG. It occurs while reading a specific rhythm EOG, as the eyes "jump" at the end of the line to the next line start. The reading movements can be distinguishing well visually by small and big eye movements (Fig. 3). All reading movements were counted that occurred after a minimum interval of 3 seconds.

Fig. 3. On display are the reading movements of the left and right eye (LEOG and REOG) as a rhythmic, blue wave pattern. The reading movements occur during reading, when the eyes "jump" at the line end of the text (right) to line beginning (left). Large eye movements

In the present study, the following variables were used and calculated: the average, the highest and the lowest reading frequency (read line per epoch), sleep latency (in minutes)

The average reading rate of the patients (n = 75) was 7.0 +/- 3.5 lines per epoch. In healthy volunteers (n = 16) it was 9.4 +/- 4.0 lines per epoch. All healthy subjects were evaluated for daytime sleepiness than normal, since neither sleep onset tendencies nor decreasing reading frequencies were observed. In 32 of the 70 OSAS patients (45.5%), however, sleep latency was found within 60 minutes. Also the reading frequency decreased over time. Rühle and colleagues calculated for the first time, the sensitivity and specificity of the *Reading test*, finding a cut-off value of greater than 11 for a pathological daytime sleepiness (Rühle et al. 2007). The standardized *Reading test* achieved a sensitivity of 76.2% and a specificity of

(e.g. view movements) are characterized by large amplitude fluctuations.

and the number of read pages.

66.7% (Erle et al. 2009).

**Results** 

informed in order to allay apprehensions and expectations.

**Evaluation** 

#### **2.8.1 Effect size analysis of the Epworth Sleepiness Scale**

Rühle and colleagues (2005) researched into an effect size analysis of the ESS the question, if daytime sleepiness could be investigated through a situation. Therefore, the authors analyzed the effect sizes of the eight items. From methodological considerations, it was reasonable to imagine, to come across items with good to very good discriminatory power, because the ESS has a good to very good reliability and validity.

In the study, which took place in the sleep laboratory of the Helios Clinic in Hagen-Ambrock, 209 male OSAS patients and 164 healthy subjects participated. To calculate the effect sizes for each item the difference between of the two item means (of patients and healthy subjects) was divided by the standard deviation of the normal population. Rühle et al. received low to very good effect sizes (ES) between 0.19 to 1.50 The best effect sizes were found for the situation "in reading" (ES = 1.50), "watching TV" (ES = .90), "sit and be passive" (ES = .85) and for "traffic-related stopping" (ES = .61). Similarly, there was an increased mean effect size of ES = .88 for the four selected items, compared to a mean effect size of ES = .68 for the total scale. Some situations of ESS was associated with both healthy subjects and OSAS patients with a high propensity for sleep, e.g. to "lie down to rest" (ES = .19), as a "passenger" (ES = .22) and "talk with someone sitting" (ES = .24). For the development of everyday life and job-related tests - as it had been suggested by Johns (2000), the reading activity was an important characterisation of daytime sleepiness, because it discriminates at the best between OSAS patients and healthy individuals in comparison to the other ESS items.

#### **2.8.2 MSLT and MWT criticism**

Although MWT and MSLT are often used in practice, since years there is the assumption that its operationalization does not correspond to the tonic activation. Johns (1998) excludes that the MSLT is suitable as a predictor of daytime sleepiness in everyday situations, regardless how strict are implementation and evaluation standards. Although have the sleep latency on both tests satisfactory correlations as Sangal and colleagues (1992, 1997a) showed in subjects with various sleep disorders (r = .41, p < .001) and in Narcolepsy patients (r = .52, p < .001). However, the tests clarify maximum of 20-25% of common variance, indicating that the test methods measure different constructs of daytime sleepiness. Reasons for the average correlations according to Sangal et al. (1992) are that patients with pathological MSLT values were able to stay awake in the MWT, while others who fell asleep in the MWT were able to stay awake in the MSLT.

In addition, Johns described measurement error as reasons for the variability of individual test results. It argues that the measurements are depended on the situation character, internal attitude and physical condition of the patient. Kotterba and colleagues (2007) reported that the sleep latency of the MSLT corresponds to the individual property to switch off quickly. In the opinion of John (2000) was the MLST least suitable and is no longer regarded as the gold standard.

In handling the tests are very time-consuming and labour intensive (because of multiple tests during the day) as well as it is uneconomical. This would be calling in question the use of the method (Danker-Hopfe et al. 2006, Johns 2000). Because the claim of a standardized implementation and evaluation it could also be performed only by professionally-equipped sleep laboratories (Randerath 1997). Daytime sleepiness can be measured more easily and possibly more effectively with the ESS (Johns 2000).

The Effects of Sleep-Related Breathing Disorders on Waking Performance 129

Pupillography (company Amtech, Weinheim, Wilhelm et al. 1998, 2001). A newer method to

The *phasic activation* is manifested in stimulus situations, in which short-term increases of the activation in the resting state are required. Limitations of the phasic activation can result in delayed reaction rapidities up to omitted reactions. The phasic activation may be tested for example in the kind of reaction time measurements, e.g. with the test battery of Zimmermann and Fimm (TAP / 1994), event-related EEG deductions or on the basis of the

Fig. 4. Proposed relationship between sleep quality and sleepiness-related restrictions

(2000a/b, 2001) have been used to test the vigilance and sustained attention.

*Sustained attention* is the ability to direct attention over a long period to one or more randomly occurring stimuli and to respond to minimal stimuli changes (Davies, Jones and Taylor 1984). *Vigilance*, which is a variant of sustained attention, requires long-term attention performance in minimally and irregularly occurring stimuli. As reliable indicators false (i.e. incorrectly or delayed) and omitted responses as well as reaction times can be measured as an expression of sustained attention and vigilance. Furthermore, particularly in the field of sleep medicine the *Clock Test* by Mackworth (1948), modified by Quatember and Maly (Sturm und Büssing, 1993), and the *vigilance test* "*Carda*" by Randerath et al. (1997, 2000) and Gerdesmeyer et al. (1997) and the *sustained attention test* "*Carsim*" by Büttner et al.

*Selective attention* is also the ability to focus on specific relevant stimuli and to suppress simultaneously occurring irrelevant stimuli. The kind of attention function can be investigated by choice-reaction tasks or orienting responses, e.g. based on the subtest

*Divided attention* describes the capacity for serial and parallel information processing and the flexibility of selecting to switch back and forth at least two different sources of information (Sturm and Zimmermann 2000). Relevant stimuli can each occur in one or two sources of information to which the person have to respond as quickly as possible. Divided attention can be measured with dual-task activities (e.g. using the subtest "*Divided attention*" of the *TAP*). As with many sleep-related disorders, such as hypersomnias and dyssomnias, the victims suffer from, in addition to their nocturnal symptoms, increased daytime sleepiness and the tendency to fall asleep (Büttner et al. 2004b). These difficulties are in turn associated with attention-related deficits and limitations (including Gerdesmeyer et al. 1997, Müller et al. 1997, Randerath et al. 1997, 1998, Weeß 1997, Weeß et al. 1998a/b, Büttner et al. 2003b, 2004b).

1 TAP = German: **T**estbatterie zur **A**ufmerksamkeits**p**rüfung; English translation: Test battery for

quantify the tonic activation is the *Reading test* (Erle et al. 2009).

heartbeat rate or skin conductivity.

during the day

"*Selective attention*" to the *TAP*1.

Attentional Performance

#### **2.8.3 Summary and outlook**

Although the MSLT and MWT have been used frequently, in many studies was found evidence that the reliability and validity of the procedures are unsatisfactory. In addition, the two test methods don't correspond to any real life situation (Johns, 2000). Even if it is objective and standardized measuring instruments, have been repeatedly confirmed weaknesses in the implementing objectivity of the individual tests as well as their generalization ability (Danker-Hopfe et al. 2006). John's criticism is that the reliability and validity verification of MWT and MSLT were not gone in any way according to objective and standardized criteria. The ESS compared to the MSLT and MWT has good reliability and validation criteria, sensitivity and specificity measures. Its only drawback lies in the fact that the subjective assessments are based on individual perception and trust and the honesty of the patient.

Johns (2000) emphasizes the need to find an objective test, such as the ESS is valid and able to quantify the alertness in various everyday situations. Such a test would represent a true gold standard. Result of this strong criticism and of the clinical relevance of developing a new measuring method, Rühle et al. (2005) analyzed the effect sizes of the ESS. They pursued the goal, to detect the daytime sleepiness of life situation as objective, reliable and valid as possible. The analysis of the ESS and its implications led to the experimental derivation, design and construction of the *Reading test* (pilot study: Rühle et al. 2007, main study: Erle et al. 2009).

#### **2.8.4 Conclusion of the reading test**

Both the pilot study and the main study, the alertness impairments in OSAS patients with the reading activity, a simple spiritual activity were operationalized. In contrast to the MSLT and MWT daytime sleepiness was not measured in an experimental laboratory situation, but in an everyday clinical situation. The *Reading Test* is suitable for the determination of daytime sleepiness, because it probably produces a low level of attention. The reading activity will be documented and monitored continuously by EOG. Therefore, the nonreading phases can be observed, e.g. at the beginning of sleep, movement and looking around of the patient. In addition, the behaviour spectrum of patients are also detected in the EMG, as unwanted movements (facial grimacing and head movements), which can prevent sleep. This aspect would be particularly relevant in experiments.
