Good afternoon, everybody, or almost afternoon. My name is Mona. I'm a surgical ICU pharmacist at Columbia, as Melissa has said. I have the task of talking about sleep dysfunction or sleep disruption in the ICU over the next 15 minutes. So I think our goal will be to talk about an overview and a high level of sleep disruption in the ICU. We'll particularly talk about risk factors, etiology, and approaches to management. And I'm sure you've seen in your Congress packets there is a solid one-hour talk on sleep disruption in the ICU this afternoon, of which Dr. Punn is going to be talking at. And so we look forward to that. But consider this as a prelude to that one-hour block. All right, I have no conflicts of interest. All right, so normal sleep cycles varies amongst individuals. But typically speaking, there's about four to five cycles, each lasting about 90 to 100 minutes per night. Most of the time, we spend our time in stage 2 sleep followed by stage 3 sleep and REM. So sleep dysfunction in the ICU is a very common phenomenon. Typically, the total sleep time is very similar between ICU patients and non-ICU patients. But there are things that are very significantly different here. So most of the time, the sleep is occurring during the daytime hours rather than nighttime hours. There's lots and lots of arousals during the sleep time. And most of the time, they're spending most of their time in the lesser sleep cycle, so the stage 1 and stage 2 sleep. And they have very small amount to no amount of time in the deep sleep. So no REM sleep or a little stage 3 sleep. So acute and chronic sleep dysfunction can be associated with many consequences. Now, one thing to highlight is most of the data is in healthy volunteers and non-ICU patients. But what the data does show is that there seems to be a correlation between sleep disruption and biologic outcomes here. And so this relationship seems to be close. And this relationship seems to be bidirectional. So sleep dysfunction has been associated with decreased inspiratory muscle endurance, decreased exercise performance, and dyspnea. There is some literature in the ICU population where, in mechanically ventilated patients who have sleep disruption, there is a higher chance of difficulty weaning off the mechanical ventilator. Sleep disruption is also associated with increased sympathetic nervous system activity. And so there is a higher chance and a higher risk of left atrial dysfunction, QTC prolongation, various arrhythmias as well. It can also cause decreased psychomotor performance, impairment of short-term memory, cognitive processing delays, and possibly even delirium. We've talked a lot about delirium throughout the course here today. But the relationship between sleep dysfunction and delirium is probably not one of cause and effect, but definitely related and definitely seems to be bidirectional as well. Sleep disruption can also cause glucose intolerance. So there's a decreased glucose metabolism, increased insulin intolerance, and circulating cortisol levels. And then lastly, it can also be associated with increased inflammation and decreased immune response, which further increases the risk of infection. So there's several risk factors to think about. These can be modifiable. These can be non-modifiable. They can change between day and day and night to night. But the thing to think about is that the consequences of sleep dysfunction are present. And I think the ideology of sleep dysfunction tends to be multifactorial. So let's think about some of these things. So we have psychological risk factors that include anxiety, fear, feeling of loneliness, disorientation. Physiologic risk factors tend to be around the patient feeling uncomfortable. So are they feeling uncontrolled pain? Are they having difficulty breathing, hungry, thirsty? Next thing to think about is care-related interventions. This is particularly concentrating on the nocturnal care-related interventions. Are we unnecessarily bothering people at night? Are we taking vital signs that may not be necessary? Are we giving medications that are technically not urgent to give? Oftentimes, our patients are complaining about really noisy the environment. Lights might be extra bright for them. And trying to take care of that also can help. And then lastly, think about the number of comorbidities. So the sicker the patient is, the likely the higher the chance of sleep dysfunction in our patients. And those who already have poor home sleep quality, those who are already taking sleep aids at home, and those who are undergoing various withdrawal syndromes can also be at high risk. We think about medications as well. So a lot of this data is also, again, involving healthy volunteers, non-ICU patients. But we should think about some of the things that we commonly use in our ICU patients also. So benzodiazepines, that's been associated with decreased REM sleep, decreased stage 3 sleep, even though it increases total sleep time. Propofol also decreases REM sleep. And opiates has been, the interventions or interactions between opiates and sleep disruptions tends to be variable in terms of how it impacts polysomnography readings. But there seems to be a relationship there that opiates can alter the circadian rhythm as well. I signaled out morphine and methadone because there was a nice paper that was published looking at polysomnography readings in patients who were exposed to methadone and morphine, and these were non-ICU patients. And they saw that there was a decrease in stage 3 sleep. Now, I'm not advocating to avoid opiates by all means. We need to make sure that we control uncontrolled pain first, as that can be a risk factor for sleep dysfunction and it can also be a risk factor for delirium, of course. But what I am highlighting is the importance of multimodal analgesia to at least decrease the amount of unnecessary opiate use, if at all possible. Tricyclic antidepressants, tertiary amines tend to be more sedating, like amitriptyline, but secondary amines, such as desipramine, nortriptyline, tend to be more activating. So you have a higher chance of wakefulness there. SSRIs, SNRIs, decrease REM sleep and N3 sleep in non-ICU populations, and then in healthy populations, corticosteroids tend to increase wakefulness. So, now that we have an idea of what the risk factors are, maybe idea of what the etiology is, what can we do to actually improve sleep? Do non-pharmacologic bundles, pharmacologic bundles help improve sleep? Well, let's talk about the non-pharmacologic bundles first, and that should be the first line therapy for all patients in the ICU. There could be several arms that we should think about. Decrease interruptions, so avoiding non-urgent interventions, and this includes clustering care as best as possible. We should promote daytime wakefulness, so promote cognitive engagement, initiate physical therapy, avoid excessive napping, and avoid giving sedating medications in the middle of the daytime, so try to time those towards nighttime. Address environmental factors, so generally speaking, most people tend to sleep a little easier and better when the room is a little cooler, but if the patient has a preference of having a really hot room for some reason, then maybe try to accommodate that. If they prefer particular blankets, positioning, et cetera, try to accommodate that as best as possible. And of course, decreasing nocturnal light, decreasing nocturnal noise is also very, very important. And then lastly, incorporate decreasing psychological distress, so treat uncontrolled pain first, and using multimodal therapy. And then enhance the communication between patients and families, so address their concerns, address their worries, address their questions. And then think about implementing music therapy. So music therapy has a little bit of a caveat, of course, and we've heard about this in previous talks as well, in that some people find music very soothing, some people find it very, very annoying. So I think we just need to talk to our patients, talk to our family members, and say, do you prefer music? And if the answer is yes, then great, figure out what kind of music they like, and then implement that. But if they don't like music, then of course, let's not do that. So the other parts of the bundle include earplugs, eye masks, along with environmental interventions. There's lots of RCTs published, and I've highlighted some of the bigger results here. So there is one randomized control trial that looked at the use of earplugs, eye masks, along with environmental modifications, and they looked at polysomnography readings, and there was an increase in stage three sleep and due case awakenings in those patients who were given this bundle. There's lots of other RCTs that looked at this bundle, and they saw the improved sleep quality when evaluated with self-assessment tools. There are some literature considerations that we should think about. First is that there's varying compliance to intervention. So a lot of these papers didn't talk about compliance to environmental modifications. And some of these patients ended up having, like some of them, up to 30% ended up having noncompliance to earplugs and eye masks, just because a lot of times it's uncomfortable for some patients. And so before we intervene and give earplugs and eye masks, just make sure your patients know and the family members know what the purpose of this is, what we're trying to do, and what the benefit of regulating sleep is going to be. And you might want to avoid it in patients who are confused or delirious. There's differences in methodologies amongst the papers. They're generally smaller sample size, mostly single center, and they generally involve lower acuity. So what should we do from a pharmacologic standpoint? Now a lot of this data, again, is outside of the ICU, and it generally involves healthy volunteers, but there's some data, a hypothesis generating studies, that look at dexmanetomidine's use as a sleep agent in ICU patients, and has seen increased total sleep time, increased stage two sleep. But the data around antidepressants, antipsychotics, remains outside of the ICU, at least from a prospective standpoint. So trazodone in non-ICU patients has been associated with increased sleep, total sleep time increase, and three sleep, similar to mirtazapine. And then when you look at antipsychotics, generally, at least when you look at Haldol, cotiapine, onalanzapine, generally increased total sleep time, increased sleep efficiency, and then variable effects on stage two sleep. There is remeltion and melatonin, both melatonin agonists. Remeltion works on melatonin one, melatonin two receptors, three to 16 times more affinity for the receptors with remeltion compared to melatonin. There's one randomized, prospective randomized study looking at remeltion in ICU patients, eight milligrams versus placebo. The primary outcome was ICU length of stay, but the secondary outcome was sleep quality. And so what this paper saw was there were fewer awakenings per night with remeltion, and there was no difference in sleep hours, though. There's lots of limitations to this one paper. It was a single center study. The sleep quality was determined retrospectively in non-intubated patients, and then there was use of sedation analgesia practices that were not reported that would impact sleep quality. There's lots of melatonin randomized control trials. We're gonna highlight two of those, the most recent ones in 2020 and 2022. So the published paper in 2020 looked at melatonin 10 milligrams versus placebo for seven days, and the primary outcome was sleep quality assessed by the Richard Campbell Sleep Questionnaire. And they ended up seeing improved sleep quality with the melatonin group. The second paper, the ProMedic paper published in 2022 looked at melatonin four milligrams versus placebo for 14 days. The primary outcome in the study was powered for delirium-free assessments, but the secondary outcome was sleep quality. And there was no difference in sleep quality or quantity in either of those groups. Again, there's lots of limitations to the published literature. There's lots of smaller papers, and most of them have very small sample sizes, lower acuity patients again. The compliance to non-pharmacologic therapy generally is not reported, and there's varying doses of melatonin used. So what should we do in a patient with altered sleep now that we've kind of seen some of this data, how these medications may or may not help really? The first and foremost thing is I think we just need to implement non-pharmacologic bundles for all patients. Second is identify and address modifiable risk factors in etiology of sleep dysfunction. If they're withdrawing, then treat the withdrawal accordingly. If they have uncontrolled pain, treat that with multimodal pain therapy. Now if they don't have withdrawal, you've ruled that off your differential, but they're agitated, including signs and symptoms of hyperactive delirium or anxiety where they are increasing harm to themselves and increasing the risk of harm to others, then think about using drug therapy here, of course. If they're mechanically ventilated, again, target light sedation with either propofol or dexmedetomidine, avoiding benzodiazepines. If at nighttime they do need sedation, because again, they're a harm to themselves or others, then minimize propofol, because that does seem to disrupt sleep in the ICU. So at that standpoint, you could maybe think about using dexmedetomidine, maybe some antipsychotics, at least for symptom control only. Melatonin could be used. There's probably little to no harm of melatonin, but in this scenario where we have a very combative patient, I'm not entirely sure if it's gonna help a lot, but you can try, certainly, adding that to other things. And then possibly restarting home sleep aids, but be careful if the home sleep aid is a benzodiazepine, which can certainly increase the risk and further worsen things a little bit. If there's no agitation and they simply have insomnia, then starting home sleep aids or melatonin seems like a very reasonable thing to consider. And then I wanna lastly think about how to implement a bundle. Bundles, of course, are very stressful sometimes and can be challenging to do, but hopefully if we think about implementing a sleep bundle, think about and considering the following. Recognize the common goals. The common goal is let's improve sleep in our ICU. Identify the multidisciplinary champions who share the same goal. Identify the folks who are interested in making a difference in quality improvement, those who are interested in sleep quality, those who are interested in the A through F bundle. Increase awareness, so make our goal known. Our goal known is we need to improve sleep and we need to adapt the A through F bundle. And also highlight the fact that where are we now? How many medications are we giving at night that are unnecessary? How many room entries are we doing that are unnecessary? And then publish that to not only the key stakeholders of that unit, but everybody, every single team member. Design your bundle. So with the first and foremost, highlighting the non-pharmacologic bundle first. And then implement it. So educate, yes, but constantly talk about it on rounds every single day. Sleep dysfunction and treating that or in preventing that should be one of the many goals for the day that should be outlined in the morning and then should be passed off at night to the night team as well. Determine measurable outcomes. So figure out what's successful, figure out what are the barriers. And then based on that, you need to probably reevaluate pieces of your bundle and re-explore what actually works for your unit in order to meet that goal. So in conclusion, sleep dysfunction in ICU is very common. We don't know the long-term outcomes, but sleep disruption may lead to unintended negative outcomes. Despite the need for more studies evaluating sleep protocols, a multi-component bundle with particular focus on optimizing non-pharmacologic therapy should be used. Thank you so much for your time.

sleep dysfunction

intensive care unit

ICU

risk factors

non-pharmacologic interventions