This study highlights some of the controversies surrounding DISE (Drug Induced Sedation Endoscopy, or otherwise known as sleep nasendoscopy in the UK). On one hand it provides some evidence that the BIS (BiSpectral Index), recordings of patients while they are asleep can be recreated during light anaesthetic conditions, which can be construed as evidence for its validity. On the other hand, it reminds the reader that sleep is not a static state with little variation, and to accurately measure it you would need to observe the upper airway over the whole dynamic spectrum that is real physiological sleep.
This article details the complexity of sleep and the difficulties of attempting to recreate it in the operating theatre. In brief the study analysed the obstructive patterns during DISE and noted the changes that occurred with differing sedation levels.
The immediately obvious issues are with the validity of BIS and its as of yet unreleased algorithm for determining sedative levels, and the use of anaesthetic agents to recreate a sleep-type state. In this study, propofol was used, and this is emerging as the most common agent used in DISE globally. Reasonably accurate results seemed to be achieved in this study of 60 patients, and the well-known German authors were very clear in their discussion about the limitations of their study and the time lag issues with BIS monitoring. The argument I think that this article makes is that to accurately assess these patients and get good quality information, you either need to assess the entire range of sleep states recreated as faithfully as you can, or pick a level and base your assessment on that. The former is too long winded for a procedure that is difficult to secure funding for in the NHS, and the latter is placing greater risk on failing to observe important changes to the upper airway that might be useful for the patient’s management decisions. One option may be to base your sedation depth choice on the polysomnography results and deciding which sleep state is causing the most harm to the patient. This however leads to the problem of having to perform a full polysomnography with EEG in all patients. Clearly the arguments will continue, but studies like this are only clarifying the situation, even if it does paradoxically muddy the waters in some respects.
