DISCUSSION
To our best knowledge, this study is one of the few researches that attempt to define BIS values during normal sleep in the sleep laboratory while performing PSG for obstructive sleep apnea syndrome (OSAS) diagnosis. Although the patients included in this prospective study were patients suspected to have OSAS; moderate and severe OSAS patients were not included to study to provide a population closer to normal. By this way, a study population with mildly distorted sleep architecture was created.
Only the BIS and sleep stage values in the sleep period that were longer than 2 minutes were taken into account (stable sleep stage) after the end of second minute.
Although the BIS devices are not calibrated to identify the natural sleep stages, there are few studies in the literature investigating the BIS values during normal sleep with conflicting results. In 1999, Slergh et al(6) claimed that; changes in the depth of natural sleep are reflected sensitively by changes in the BIS and BIS may be a simple indicator of the depth of sleep. But their study group was consisted of only 5 subjects. Benini et al(5) also concluded that BIS is sensitive in reflecting the changes on the EEG trace that accompany various stages of natural sleep. To be more specific, they found a progressive decrease in BIS values as the stages of sleep becomes deeper, in their study performed with 15 children. Contrarily, Neuwenhuijs D et al (11) reported that the processed EEG measurements (both BIS and spectral edge frequency) have a limited ability to estimate classical sleep stages as a result of their work with 10 patients.
It has been shown that, BIS can identify stage N3 with satisfactory sensitivity and specificity, but it is unable to distinguish light sleep stage (N1) from REM, sleep or wakefulness (W). (3)Signal quality index is used in different scientific areas to explain some characteristics of determined signals.Therefore the defined SQI is related with signal strength.The SQI is ranged between 0 and 100% (best signal quality). Mohammad et al suggested that BIS signals should be read in relation to the current signal quality index (SQI) and impedance. (3) However there is no validated threshold for the quality of the BIS signal and it is not completely clear how the SQI is calculated and what should be an ideal value.
In our study, we first investigated the BIS scores for SQI values of 50-70, 70-90 and over 90 and we found that there was no statistically significant difference between average BIS values of 70-90 and over 90 SQI values, but there was significant difference between 50-70 and 50-70/over 90 SQI values. So we included only the BIS values with an index of SQI over 70 to study.
According to our study, BIS values decrease progressively as the stages of sleep becomes deeper. Mean BIS values were 87.70, 83.94, 74.41, 56.60 and 74.90 For W, N1, N2, N3 and R3 stages respectively (table 4). Furthermore, BIS values were sensitive in reflecting the stages of natural sleep except REM and N2; REM and stage N2 BIS values were similar. One possible explanation of proximity of the index scores in N2 and REM stages may be the inability of the available BIS monitor to distinguish the brain’s electrical activiy in these two stages. Our results are incompatible with some of the previous studies, which reported BIS is not 100% specific in predicting intraoperative awakening and recall( 2). In their study, Duarte et al also claimed that the model of BIS monitor could influence the interpretation of the results. We found a statistically significant relationship between BIS scores and sleep stages, which may due to use of a higher model of BIS monitor.
All of the currently available monitors need different period of times to calculate and update the index in response to changes in the depth of anesthesia (2). The time to update BIS records can range from 14 to 155 seconds. For this reason, to acquire more reliable results, we evaluated only the data obtained after the second minute of stable stages.
BIS monitoring may be a practical method to identify sleep stages.Because in oder to recognize sleep stages,BIS is a numerical method that everyone can evaluate in stead of EEG which can only be assesed with a specific training. Which BIS values should be used for sleep endoscopy is not exactly known yet.  As a result of their research Babar-Craig *12* et all reported that : average BIS values ranging from  50 to 61 is appropriate for sleep nasendoscopy. Our results suggest that, the 50-70 BIS values correspond to the deep stages of sleep (N3), and it is well known fact that respiratory events are less common in N3 phase of sleep (13). Thus, we recommend that sleep endoscopy should be done at 70-80 BIS values, which corresponds to N1 and N2 phases of sleep where respiratory events are more frequently seen.
In conclusion, according to us BIS is a usefool tool for distinguishing of normal sleep stages except the R2 REM stages and BIS values decreases as the sleep stages become deeper.