I have seen so many cases where the BIS monitor didn't correlate with the patient's state or where it just didn't work. It seems like the cables aren't very sturdy and frequently break giving erroneous readings.
There's an anesthesia awareness website
www.anesthesiaawareness.com
This website shows a dramatized view of the problem that's probably similar to the TV program referred to above. One noteworthy thing about the site is how many times it asks for donations throughout.
The fact is that awareness under anesthesia is a risk. The risk is small for routine cases. However, for trauma, OB, and even cardiac cases the risk increases and for good reason. In trauma many of the drugs given for anesthesia have significant hemodynamic consequences, When your patient is bleeding to death are you more worried about keeping them alive or keeping them from possibly remembering things. Ideally you want to do both, but sometimes the goals are at odds with each other. For OB, the time course of the anesthetic is done differently in order to reduce the amount of sedatives that the baby receives. Again there's conflicting sides. In most cases it works out fine, but which would you rather have happen: the baby needing to be resuscitated or the mother remembering the sensation of the ET tube or hearing the surgeons/nurses talking for part of the delivery?
Part of the problem is that there is no foolproof way to measure "depth of anesthesia". In fact it's so difficult to determine that the term is really a misnomer. Depth of anesthesia is used frequently, but it is more appropriate to instead discuss the specific components of anesthesia: unconsciousness, amnesia, analgesia, paralysis
Another point is that due to personal variation in metabolism of the various drugs, a patient can be dosed accordingly and still be lacking in one of the components of anesthesia.