Thank you for being here today. I'm going to speak to you about sleep disruptors in the intensive care unit. I have no financial disclosures or conflicts of interest. So today, the objectives I want to cover with you are to highlight the interplay between sleep and circadian rhythm in the ICU, address disruptors to sleep and circadian rhythm and review their impact and present a framework for addressing these sleep and circadian disruptors. So I want to just give a little definition of what sleep is here. So sleep is a periodic reversible state of cognitive and sensory disengagement from the external environment. We know it's essential to survival and that animals when they are sleep deprived eventually will die. While the purpose is still not fully understood, there has been lots of exciting research in the past several years highlighting the importance of sleep, especially with cognitive impairment and clearing toxins from the brain. The one thing I want to point out in relation to the ICU though, is that sedation is not sleep. We know this from performing sleep studies and EEGs on patients while they're receiving sedatives. So we can't assume that someone who is sedated is getting restorative sleep. We all know the effects of sleep deprivation are protean, but the things that are important for ICU patients and possibly link sleep and delirium are irritability, cognitive impairment and memory impairment and hallucinations. The other things that are important, especially for our critically ill patients, are impaired immune system and changes in heart rate variability. So we understand sleep and know how it feels to be sleep deprived, but how does circadian rhythm fit into this equation? The circadian rhythm is an endogenous self-sustaining oscillation that follows the approximate 24-hour day-night cycle that's controlled by the daylight of the sun. So sleep is a two-process model that incorporates both the circadian system termed Process C and something called the homeostatic drive for sleep or Process S. And so you can see at the top, the arrows pointing down in blue at the top of the slide, represent the homeostatic process to sleep. And as the day goes on over the course of the day, the drive to sleep increases and this is mediated by adenosine. So eventually you will fall asleep. The yellow arrows at the bottom represent the circadian rhythm. And so the circadian rhythm or your wakefulness increases over the day till you reach a point where melatonin begins to be released around 9 p.m. and that increases our drive to sleep. So these two opposing processes play a role in regulating our sleep. So in mammals, there's a central master clock located within the suprachiasmatic nucleus of the hypothalamus. And this master clock synchronizes peripheral clocks that are found in all organs, as you can see here and picked it in this photo. The largest or the most important entrainer of the circadian rhythm or controller of the circadian rhythm is light or bright light. And we call those entrainers zeitgeibers. So this is why light becomes important when we think about critical illness. Other things that are important in the ICU that we want to think about that are non-photic zeitgeibers are the sleep-wake cycle, physical activity, social time, and meals. And so we'll talk a little bit about these when we talk about sleep disruptors. So we know that sleep disruption is common in the ICU and studies that have gone back almost 30 years have documented this. Patients have increased difficulty falling asleep. They have increased numbers of arousals from sleep and they don't progress normally through the sleep stages. So when studies have been done using polysomnography in critically ill patients, we find there's a predominance of stage N1 and N2 or the light stages of sleep and a lack or absence of stages N3 and REM or rapid eye movement sleep, which are the restorative stages of sleep and felt to be important for things like memory consolidation. We also know that even if patients get a normal number of hours of sleep, that it's fragmented with 50% of it occurring during daytime hours. We also know from more recent studies in the last 10 years or so that circadian rhythm is disrupted in critically ill patients. And this is done by using studies that have used things called after watches, which are accelerometers or similar to Fitbits or activity trackers and measuring markers of circadian rhythm, such as melatonin. And these studies have demonstrated that in patients with sepsis, there is severe circadian disruption as well as a total temporal disorganization of circadian rhythm. Two recent studies that came out in critical care medicine have also demonstrated that there is marked abnormalities in melatonin secretion in patients both with sepsis and intracranial hemorrhage. This was a study that we did using actigraphy to determine circadian rhythm in critically ill patients. The top panel shows a normal circadian rhythm where you see activity, then a period of quiet, and then activity again. Whereas the one in the middle panel is a misaligned circadian pattern. And the one on the bottom demonstrates a total absence of any circadian pattern. The ICU environment is problematic, as those of you who work there know, because there's not much opportunity for sleep. And so what factors contribute to sleep disruption? Well, some of them that we'll discuss today include noise, light, and then things related to care in the ICU itself. Many factors that are similar contribute to circadian disruption in the ICU. So things that impact circadian cues to keep things normal. And again, things including light. Social cues, which has become even more important during COVID when less and less patients are going in rooms and families are not present. Meals are how we feed patients. Exercise or mobility. And then things related to your critical illness itself, including things that we do to you that are modifiable. So how we scent the ventilator, medications we use restraining patients. So this is a qualitative study that looked at patient and staff perceptions of sleep in the ICU and showed that there were three themes that came up. The environmental effects, so the environment is disruptive, which is not surprising, and unpredictable workflows. And then non-environmental effects, so emotional factors. So patients are afraid, they're anxious. And we also have to remember patients come in that have chronic sleep problems, and those should be recognized and addressed when possible. And then perceptions of sleep that showed that sleep quality is variably reported in studies or to staff. And that both patients and staff believe medications improve sleep. And we'll talk a little bit about how medications may again cause sedation, but actually don't provide normal sleep. So again, why do we care about sleep and circadian disruption? Because it causes altered physiology, which leads to stress, delirium, impaired cognition, and more recent literature suggests sleep impairment in the ICU can lead to post-intensive care unit syndrome. And those characteristics highlighted in red here are the things that are modifiable and we can address. So what does the data actually show? One of the modifiable factors is high ICU sound. And studies starting back from 1973 all the way up to 2019 have demonstrated that ICUs are noisy. The red bar is the World Health Organization max decibels for sound. And you can see in all these studies we're above that. We also know from polysomnography that sound peaks are associated with sleep arousals. This systematic review that came out in 2018 that looked at 20 publications showed that there was heterogeneity between studies, a considerable risk of bias, and that there was a wide variation in how methods were described. So it was difficult to quantify the extent that noise plays in sleep disruption, but they did show that if you expose healthy subjects to ICU noise and do polysomnography, 60% of arousals are associated with sound. If we look at light, we know that ICU light conflicts with circadian rhythm. So light, as I said, is the largest or the most important entrainer of the circadian rhythm. And you can see here where there's this orange circle that there's a very low light in the morning, which is when there should be the brightest light. And then if you look at the overnight hours, the light is too bright. So our light is not mimicking normal circadian light. If we look at what impacts light by time of day, things like the television being on overnight, and then during the day when it's too low, the window shades are closed when they should be open. The other thing is nocturnal room entrances. So we know that we go into patients' rooms a lot, and it will be important for us to think about this. If you look at midnight to 4 a.m., which should be the deepest sleep, there's less than 30 minutes of quiet time for a patient to get any sleep. And we know that you need two hours to get through a whole sleep cycle. So there's not even enough quiet time for a patient to get through one whole normal sleep cycle. Ventilator mode may be another important disruptor to sleep. There are some small studies that have looked at using different ventilator modes. One study looked at pressure support ventilation and showed that using this led to hyperventilation and decreased CO2 levels, which caused central apneas and arousals from sleep. And that pressure support ventilation caused more arousals per hour than assist control ventilation. So the other ways that mechanical ventilation leads to sleep fragmentation is related to increased ventilatory effort, having abnormal gas exchange. Another really important one is patient ventilator to synchronate. Obviously, discomfort from having an endotracheal tube in, and then things related to care of the patient and ventilator. So things like suctioning, ventilator checks, repositioning the ET tube. And so I just want you to remember that adjusting the ventilator to maximize patient comfort may depend more on the settings and not the ventilator mode. So really making sure your patients are as comfortable as possible with the ventilator settings you have chosen. Medications have important impact on sleep. And as I said earlier, sedation is not sleep. And we know this from studies that have looked at polysomnography in patients receiving the medications we use frequently for sedation. So benzodiazepines decreased the restorative stages of sleep, N3 and REM, and increased light stages of N2, and also contribute to abnormal sleep architecture. Propofol also decreases REM, which is the restorative stage. Again, dexmedetomidine has the same effect, decreasing N3 and REM, although a recent study giving dexmedetomidine overnight suggests that it actually may reduce delirium. And again, opioids decreased these restorative stages of sleep and caused dose-dependent reductions in central sleep drive. So again, thinking about medication choices and minimizing medications where possible. We also know that medications affect melatonin. And so these drugs can increase melatonin levels. So given at the wrong times of day or continuously, they may impact our circadian rhythm. Benzos decrease melatonin levels, and vasopressors also increase melatonin levels. The PATIS guidelines, which came out in 2018 from the Society for Critical Care Medicine, have added sleep disruption into their recommendations. And so referencing those, there's a whole host of information about the importance of sleep and what is known about sleep in the ICU. And this study that just came out in 2020 added on to what's known in the PATIS guidelines, and it looked at risk factors for poor sleep during ICU admission, including 44 observational studies. And you can see these are things that we've talked about, noise and light, pain, discomfort, immobility, nursing care, worry, anxiety, fear, and then things like delirium, benzodiazepine, and propofol use, ventilator, asynchrony as well. So when you think about it, this is the interplay between all of these things that can impact sleep and circadian rhythm in the ICU. And so when you think about how to improve it, you really want to think about addressing all of these things and the potential that they have. So things like light and sound where possible, reducing alarms, clustering bedside care. We do know that you can use sleep promotion interventions to impact some of these outcomes. So improving light and providing bright light early in the day. This sleep promotion intervention demonstrated a 32% reduction in in-room entrances and lower peak and mean sound levels. And so you also want to think about providing a chronobundle, and I know other talks will address medications and other sleep promoting interventions. So I'm just going to leave here with thinking about providing bright light, bolus feeding, early mobilization, social interactions, and then again, bundled care, minimizing noise and light, and providing earplugs and eye masks. So I'll stop there and thank you for your time. Good afternoon, everybody. My name is Dr. Laura Baumgartner. I'm an associate professor at Toro University, California College of Pharmacy. I'm also a trauma ICU pharmacist at Highland Hospital, which is located in the Bay Area. So today we're going to be talking about melatonin and if it offers deep sleep support in the ICU. I have no financial disclosures or conflicts of interest. So the objectives today are going to be to review the barriers to sleep quality among critically ill patients and the impact it can have on patients. We'll understand the mechanism of action of melatonin and the role it might play in sleep support. We'll review the current literature as it currently stands on the safety and efficacy of nocturnal melatonin administration in our critically ill adults. So let's go ahead and get started. So a lot of us know sleep disturbance in the ICU is a common complaint among our patients and it's been linked to poor outcomes. There's been an increased focus I'd say over the last couple of years on improving sleep quality among our critically ill adults, particularly through both non-pharmacologic and pharmacologic interventions, which melatonin is what we'll sort of spend the majority of our presentation on today, but it's an endogenous hormone that aids in sleep induction and also regulation of circadian rhythms and it's thought to potentially improve sleep quality among our critically ill patients. So let's review sleep in the ICU. So we know sleep disruption in the ICU is very common and it's been categorized by three things, sleep fragmentation, increased time spent in light sleep, and also relative lack of both REM sleep and also stage 3 sleep. We know subjective quality, sleep quality in the ICU is very poor. When patients recall their time spent in the ICU and things that were moderate or severely bothersome they recall that being lack of sleep due to frequent awakenings, the stress that's associated with decreased sleep, and only being able to take short frequent naps. So what are the patient outcomes that could potentially be linked to poor sleep? The first thing I want to mention is we know studying sleep in the ICU is quite difficult. There's a lot of debate in terms of the best way that we can measure sleep and I want to preface that with saying a lot of the literature we'll review today is you'll notice it's a very low end and there's just not a lot of data in terms of how we not only analyze sleep but also the non-pharmacologic interventions that we could potentially improve sleep. There's debate in terms of whether or not we should be using polysomnography versus bispectral index, actigraphy, or just subjective nursing scales. The data is showing that if we can use polysomnography that would be the best in terms of analyzing patient data, however that's very difficult and you'll notice a lot of these studies use quite different measures to assess sleep. The poor outcomes that's been linked to sleep has been increased ICU delirium, increased time in the ventilator, and overall neurocognitive dysfunction. Because sleep is a potential modifiable risk factor there's been an increased focus like I mentioned over the last couple years and you'll notice that we'll talk about the PATAS 2018 guidelines so there's sort of a surge in our interest in sleep given that it's a potential modifiable risk factor. So there's been an increase in data particularly with non-pharm and pharmacologic management. So what is melatonin? So melatonin is an endogenous hormone that aids in sleep induction and regulation of circadian rhythms through its affinity for both M1 and M2 receptors. It's synthesized in our hypothalamus and secreted by the pineal gland. We know secretion correlates with times of darkness and sort of parallels that light-dark cycle. So what exogenous melatonin products do we have available to us? The one I will focus on is melatonin, however remelaniton is a melatonin agonist. It is FDA approved for insomnia and there has been literature linking remelaniton to decreased ICU length of stay, decreased delirium, not a ton of literature quite yet on its effects on sleep in the ICU, but it is a melatonin agonist that is available that is FDA approved. The one we will focus on today though is melatonin. It is a natural dietary supplement and I want to mention that it is not regulated by the Food and Drug Administration. So the optimal dosing and sort of what is contained in each of these different variations or formulations is quite varying. So I'd say the biggest area of research that we still definitely need is optimal dosing as it's not regulated by the FDA. The way it works is it is an exogenous non-selective melatonin agonist. So it works both at your M1 and your M2 receptors. So the first thing I want to talk about today is impaired melatonin secretion. So there's been a few studies out there, I'll focus on this one right here by Mundegler and colleagues, however that's not the only study we have. We think critically ill patients may have impaired circadian release of melatonin. So what that happens is it results in a decrease endogenous serum levels. So this study what they did is they compared endogenous melatonin secretion in critically ill patients as compared to healthy individuals. So they went in and they looked at urinary levels of melatonin and they sampled that every 4 hours over a 24 hour period just to see what critically ill patients versus healthy individuals, how much melatonin they're secreting in the evening. They looked at 17 septic and sedated critically ill patients, 7 non-septic patients and 21 healthy individuals. I think their results are quite interesting. They found that circadian rhythm of urinary melatonin secretion is actually nearly abolished in our critically ill patients. You can see here our first figure in septic patients, you can see that the time in which we should be secreting melatonin about 9pm, 10pm, there's actually no increase in melatonin secretion for these septic patients. Whereas you can see our control patients, otherwise healthy individuals have that surge of nightly melatonin secretion. So you can see it's nearly abolished in our septic patients. Interesting they found though that the non-septic patients, so those 7 patients who are non-septic actually had exhibited similar circadian release of melatonin as our control patients. So an interesting takeaway that there may be this blunted melatonin response in our critically ill patients. There's been other studies in addition to this one that's shown that patients in the ICU or patients in sort of an ICU environment have impaired melatonin secretion whereas otherwise healthy individuals have that melatonin secretion that occurs about 9, 10pm. So that brings us to the question, should we be giving exogenous melatonin to our critically ill patients? We think they do have this blunted melatonin secretion, so should we be giving them a melatonin product and will that aid in sleep quality? So I'm going to go through a few of the studies that we have to sort of help us make a recommendation of sort of where we stand. So the first one I'm going to look at is in a simulated ICU environment. This was a prospective randomized trial and they randomized patients. It was a two-part trial. The first part they looked at just a crossover design comparing normal sleep conditions compared to when they were subjected to a simulated ICU, so that was ICU noise and light conditions. The second part is when they randomized the patients to four different groups. You had noise and light, noise and light plus placebo, noise and light plus eye mask plus earplugs, and then noise and light plus melatonin one milligram nightly. What they did is they assessed the sleep quality and serum melatonin levels. So you can see here this is results from the first part of the study. I'll just bring everything up to the screen here. So you can see this is baseline where the patients were sleeping in normal conditions versus ICU noise and light. You can see across the board there was a statistical significant reduction in sleep quality. That was REM sleep, sleep efficiency index and overall patient perception of sleep. That was all decreased when patients were subjected to that noise and light. I think this figure here on the right is really important You can see this is baseline with the circles here, this is elevation. So this is serum melatonin concentrations. You can see when they had their baseline normal sleep they had their elevated melatonin secretion overnight, whereas when they were subjected to that noise and light they actually had quite a blunted response similar to what we were finding in the beginning of this presentation in that when patients were subjected to noise and light that their overall ability to secrete melatonin was blunted. So moving on to the second results here, this is when they were introduced to either placebo, eye mask and earplugs or melatonin. You can see here on your right hand side the melatonin group. Again across the board those patients that had received melatonin had an overall statistically significant improvement in a lot of their sleep factors. That included REM sleep, sleep onset latency and number of wakenings. In terms of patient perception of sleep you can see again statistically significant improvement across the board in those patients that received melatonin. Again looking at the melatonin concentrations you can see here that this was their patients who received exogenous melatonin so you can see an obvious increase in serum melatonin concentration when the drug was administered and then tapered off over time. Again we see that blunted overall response regardless of if you had eye masks or earplugs or you got the placebo they all had a relative lack of increase in melatonin concentrations overnight. So now let's look at a study that was actually conducted in our ICU patients. So we have Boren and colleagues who conducted a small randomized placebo controlled trial and they included adult ICU patients with respiratory failure and it's important to note they stopped all pain and sedation prior to 48 hours to enrollment. A small study, like I mentioned, a lot of the studies we'll be looking at today are quite a small end. So they randomized patients to melatonin 10 mg nightly versus placebo and they measured sleep using a bispectral index. So they define sleep as a bispectral index of less than 80, but then they also looked at the sleep efficiency index. So that was the ratio of patients total sleep over time available for nocturnal sleep. So what did Boren and colleagues found? They actually found that there was a nocturnal sleep time increased one hour in the melatonin group. So patients were sleeping about 2.5 hours in the placebo group versus 3.5 hours in the melatonin group, although not statistically significant could be considered clinically relevant. What was statistically significant though was melatonin resulted in a statistically significant decrease in the overall bispectral index AUC by about 7%. So melatonin was associated with an improvement in overall sleep quality as assessed by that bispectral index. It's important to note as well that no adverse events were noted. It was very well tolerated among patients. I put the results here from table two from the study. You can see the other secondary analysis, actigraphy, nurse assessment, and the Richmond-Campbell sleep questions were no difference in terms of melatonin and placebo. So no difference in our secondary outcomes, but I think it is important to know that, you know, I think it is clinically relevant that their sleep increased by an hour and that the bispectral index also was decreased by about 7% in that melatonin. I also wanted to point out the pharmacokinetic findings that they had. So you can see here this is clock time and then plasma melatonin concentrations. So you can see the expected increase when patients received melatonin at 9 p.m. and then it tapered off over time. I do want to note that this asterisk is associated with a 4 a.m. random serum melatonin level. You can see at 4 a.m. these patients still had quite an elevated melatonin concentration. So it begs the question, is 10 milligrams maybe too much to be giving our patients in the evening, because it did result in still elevated concentrations in the morning and throughout the day until it tapered off near night time. The authors did have the same similar question, is 10 milligrams maybe a bit too much, and again each formulation of melatonin is different as it's not regulated by the FDA so that is quite difficult for us to sort of make concrete conclusions about what's the appropriate dose we should be giving our patients. So last study I want to talk about before we look at the PATAS guidelines is another smaller study by Ibrahim and colleagues. They looked at melatonin in just tracheostomized patients and they randomized patients to either three milligrams nightly or placebo. All their patients were on a trach, also weaning from mechanical ventilation and it's important again to note their sedative infusions were stopped at least 12 hours before inclusion. They assessed sleep by total number of hours observed during the night by nurse record. So what they actually found was there was no difference across all boards in terms of patients who got melatonin. So you can see on average nocturnal sleep was about 240 minutes in the melatonin group versus 243 minutes in the placebo group and again not statistically significant difference. In addition daytime sleep was also similar between both groups and again melatonin was well tolerated as we've seen with the majority of our melatonin studies is that it's generally well tolerated. So again this study actually showed no difference in sleep effect after receiving melatonin. So figure one over here is your box plot of observed sleep again pretty obvious there was no I would say statistical or clinical difference in terms of overall sleep. Our table three here is actually melatonin concentrations before and after drug administration. They just wanted to make sure serum concentrations of melatonin increased or it was absorbed by our patients and you can see here that it was quite absorbed well by the patients. So you can see about an hour, so 930 is their baseline melatonin, they got the medication and then it increased at 1130 quite significantly. So the melatonin patients were absorbing their medication appropriately but again no difference that they found in terms of sleep outcomes. So where does that leave us? So I do want to talk about the PADIS 2018 guidelines. Given that sleep like I talked about is a potential modifiable risk factor, there's been an increase like I mentioned focus on how do we optimize these patients sleep while in the ICU. Their current recommendations for melatonin as it stands is they say we make no recommendation regarding the use of melatonin to improve sleep quality in critically ill adults. No recommendation, very low quality of evidence and as you can see from the previous slides there's quite a lack of data looking specifically at melatonin for sleep. So it's sort of up to the provider or the healthcare team to determine sort of the risk versus benefit of giving melatonin and I sort of always say or my sort of take on practice is it's been linked to very few minimal side effects. So to me the risk versus benefit always typically points in favor of melatonin, however again I think we need a lot more data to help us make a better sort of educated decision and again I think the biggest concern a lot of us still have is just the overall quality and consistency of the medication. It's not regulated by the FDA so a lot of our formulations contain different varying amounts of melatonin so it's hard to make a hard set rule in terms of what doses we should be giving our patients. With all that being said though just as I was putting this presentation together over the last month this article came out to publication which I was very excited about, this just got published in CritCareMed this past month and it's actually the largest melatonin on sleep study we have to date and it was a group out of Brazil and they looked at the effects of melatonin supplementation on sleep quality and they also assessed serum melatonin concentrations. So I am going to review this study for you guys today because I think it just adds to the body of literature that we have. So this is a double-blind randomized placebo-controlled trial like I mentioned it was conducted in Brazil at two tertiary hospitals, they included adult IC patients and they randomized patients to either 10 mg nightly of melatonin versus placebo, definitely one of the larger studies we have with about 100 patients in each group and they treated patients for a total of seven days. The outcomes they assessed were not only sleep quality but also melatonin levels. So let's look at how they assess sleep. They did two different sleep assessments. They did bedside nursing assessment which assessed the total number of hours the patient slept per night and they also did the Richards-Campbell sleep questionnaire so the RCSQ and as a reminder for us that scored 0 to 100 with 0 being the worst possible sleep and 10 being the best possible sleep. They looked at the five different categories which was sleep depth, sleep latency, awakenings, returning sleep and also sleep quality. So the way they measured sleep quality using the RCSQ was that they added all the five categories together and then they divided by five and then they categorized their patients on the quality of sleep they had. So they set a score of 1 to 25 as very poor sleep, 26 to 50 as poor sleep, 51 to 75 as good sleep and greater than 75 as considered very good sleep. So what did they find? So they found that sleep quality was actually significantly improved in patients who received melatonin. They found no difference in the total number of hours slept per night as recorded by the nurse, however they found that the total RCSQ was statistically significantly improved in the melatonin group. You can see here that it was statistically significant with a p-value of 0.027. You can see in terms of breaking it down by category I think the drivers of that statistical significant appear to be the sleep depth and the returning to sleep, however when taken together with the total RCSQ score it was improved with melatonin. So the authors concluded that melatonin did improve sleep despite equal number of hours slept per night. When breaking down overall sleep based off their different categories whether or not they were categorized as very poor sleep, good sleep or very good sleep, you can see here this first column in the very poor sleep group that melatonin significantly decreased patients who had very poor sleep. You can see about 3% of patients in the melatonin group reported very poor sleep versus about 15% in the placebo and that was statistically significant with a risk reduction of 0.21. So you can see melatonin may have more of an effect on that group who would be classified as very poor sleep. Not statistically significant in these other categories however again just looking at very good sleep melatonin was associated with about 46% of those patients versus 34% of the placebo. Not statistically significant but could be clinically relevant. Now let's take a look at melatonin concentration time profiles. So again you can see at 2 a.m. the patients who received melatonin had a good serum melatonin concentration and then it tapered off throughout the day into the evening in ready to get their next dose. You can see here again in the placebo as we've been seeing in a lot of the other melatonin concentration studies there's just an overall blunted melatonin concentration response in our critically ill individuals. You can see here that at 2 a.m. these individuals only had a concentration of about 32.5 which is similar to their nighttime or evening concentration prior to when they should be secreting their nocturnal melatonin. So again seeing that overall sort of blunted melatonin secretion response. So taking that new article out in CritCareMed this past month combining it with some of the smaller data that we have, where does that leave us in terms of melatonin for deep sleep in the ICU? I think my conclusions are that administration of melatonin does appear to be safe in our critically ill patients. In terms of effectiveness I think there's limited data. I think that recent study adds to the body of evidence but I would say overall there's still limited data in terms of its effectiveness on nocturnal melatonin administration and whether or not it provides a benefit. I would say right now I think the data is trending in towards that it does provide a benefit and like I mentioned previously I think the risk versus benefit ratio of giving melatonin to our critically ill patients would point in a favorable administration. Like I said it hasn't really been linked to a lot of major side effects. The biggest one patient complains about is often headache. So I think that the risk versus benefit like I mentioned would point to a favorable administration particularly in patients who we know are suffering from poor sleep quality. With all that being said though I still think we need more prospective larger studies to definitively say that it is very effective in improving sleep quality and I think another big question like I've been talking about throughout the study is what's our optimal dose and how should we be dosing? If we do want to use melatonin how should we be dosing these patients in the ICU? Whether or not should we be giving the lower dose of 3 or the higher dose of 10 and I think at this time we can't confidently say what the best dose is. So that's all I have for you guys today. Thank you very much for listening in. I'm happy to answer any questions. My email is listed here for you. Hopefully you got a better understanding of where we currently stand in terms of using melatonin for deep sleep support in the ICU. Thank you. Hello. My name is Reema Bhawesam. I'm an ICU pharmacist here at the UCSF Medical Center. Today I'll be presenting to you on other sleep aids for critically ill patients. I have no financial disclosures or conflicts of interest to disclose. My objectives for today are to review the literature evaluating common IV medications for sleep in critically ill patients. Also to summarize literature assessing the use of common enteral sleep medications in critically ill patients as well. So before we delve into the data, it's important to know why we care about sleep. We care about sleep because we know it's a problem in the ICU that's widespread. Sleep impairment has been linked to worse outcomes and we consider sleep to be a modifiable risk factor. While non-pharmacologic sleep promotion measures certainly should be prioritized, medications are often trialed in efforts to facilitate sleep in the ICU. Data surrounding the use of sleep medications in the ICU is very limited at this time as you'll see. In the previous presentation, we reviewed the barriers to sleep and utility of melatonin for sleep. And in this presentation, for the remainder of this presentation, we'll be focusing on literature evaluating the use of non-melatonin sleep medications in the ICU. I know the last presentation made reference to the 2018 CADIS guidelines and recommendations as they pertain to melatonin specifically. These guidelines, in contrast to previous guidelines, incorporate and acknowledge sleep as a major problem in the ICU overall. However, they also highlight the lack of safety and efficacy data supporting the use of sleep medications in the ICU. In fact, when evaluating different sleep classes of medications, they offer guidance on only two additional medications, propofol and dexmedetomidine, making no recommendation for the use of dexmedetomidine and a conditional recommendation against the use of propofol. In the following slides, we'll review the data as to why. Before we delve into the data supporting each of the medications, I thought it'd be worthwhile to go over some general definitions as they pertain to sleep characteristics. This will be a review for most, so I'll be brief. Sleep latency essentially represents the time to fall asleep. Sleep efficiency represents the percent of time in bed asleep. Sleep fragmentation represents the number of awakenings per hour. Total sleep time represents the total time in minutes of REM and non-REM sleep. Sleep quality has a definition that's variable depending on the study, however, it generally represents the percent of time in stage 3 or 4 non-REM plus REM sleep. This is a reminder of the sleep stages in general. Again, this is just a brief review. Stage 1 and stage 2 are generally considered to be the lighter stages of sleep. Stage 3 or 4 plus REM sleep are really the most important stages of sleep. REM sleep, also known as rapid eye movement sleep, which allows for higher cortical sleep immune system enhancement and is considered to be restorative overall. Stage 3 or 4 non-REM sleep, also known as slow-wave sleep, provide rest and feeling of rejuvenation and is generally associated with muscle tissue repair. So now that we've gone over basic definitions, we'll begin by reviewing the data as it pertains to IV sleep medications, starting with propofol. So we know propofol is a common sedative used in the ICU. It's thought to be a global CNS depressant, primarily through GABA-A receptor agonism and possibly reduced glutamatergic activity through NMDA receptor blockade. Adverse effects associated with propofol include a decrease in cardiac output, hypotension, over sedation, and apnea. Benzodiazepines are also commonly used in the ICU for anxiety, status epilepticus, and sedation. They enhance the effect of the neurotransmitter GABA at the GABA-A receptor, resulting in sedative, hypnotic, anxiolytic, anticonvulsant, and muscle relaxant properties. Adverse effects of benzodiazepines include over sedation, coma, delirium, and respiratory depression. Now delving into the evidence, there have been several studies evaluating use of propofol for sleep in the ICU, the first one being an open prospective randomized study in 1996 of 40 non-intubated SICU patients. Within the study, they looked at bolus plus infusion dosing of midazolam versus propofol. They used a patient-reported scale of 1 to 10 to evaluate sleep quality and found overall no difference in sleep quality between the two drugs. Of note, this particular open label study included relatively young, low acuity of illness, normal body weight patients, and did not contain a control group. A subsequent study in 2012, which was randomized and crossover, evaluated 12 ICU mechanically ventilated patients and did include a control group. In this study, they looked at propofol infusion versus no sedation. Using polysomnography, they found no difference in sleep efficiency, total sleep time, sleep stage distribution. Of note, they actually found that propofol was associated with a decrease in REM sleep, in fact, REM suppression potentially, hemodynamic side effects, and respiratory depression sometimes resulting in mechanical ventilation. Despite the small sample size in this study and the limited patient population in general, this study does highlight that there may be potential harms associated with the use of propofol in the ICU for sleep. Subsequently, in 2014, the first randomized double-blind trial was published to review the effects of propofol versus benzodiazepine called flunitrazepam. This drug in particular is not currently available in the U.S., but is available in select countries and has been used to treat severe insomnia in the past. In this study, this was a randomized double-blinded trial of 66 ICU patients, and they used bispectral index to evaluate sleep quality and efficiency, as well as total sleep time. What they found was, overall, that the use of propofol was associated with improved sleep quality, sleep efficiency, with no difference in total sleep time or sleep latency. Of note, this study does use BIS monitoring, which exhibits a lot of variability in general. It's not generally considered to be the gold standard when it comes to evaluating stages of sleep. Furthermore, the study does lack a control group and included a very specific population of patients that did not require any additional sedatives, did not require mechanical ventilation, without any renal insufficiency. Therefore, it unfortunately represents a very select group of patients. Given the potential risks of propofol and benzodiazepine use in the ICU at this time, especially via continuous infusion in intubated patients and non-intubated patients, there's still insufficient data to support the use of propofol for sleep at this time. Here, we see several additional studies evaluating the use of benzodiazepines. In the first study by Hardin and colleagues in 2006, they looked at 12 patients requiring mechanical ventilation in an observational study. They compared patients that were on intermittent dosing versus continuous lorazepam. Using polysomnography, they found no difference in total sleep time, sleep efficiency, or sleep fragmentation. They did, however, find an increase in slow-wave sleep with intermittent lorazepam use. A subsequent study by Otto and colleagues in 2011 looked at 22 mechanically ventilated patients in a randomized control trial. They compared patients on intermittent dosing midazolam versus continuous infusion midazolam. And using polysomnography, they found that in patients that received continuous infusion, there was an increase in total sleep time and a decrease in sleep fragmentation. In patients on intermittent dosing, there was an increase in slow-wave sleep and an increase in REM sleep overall when compared to continuous infusion. Although this data appears to suggest some benefit, the data surrounding the use of benzodiazepines as seen in previous studies remains conflicted at this time overall. Furthermore, benzodiazepines have been associated with an increased risk for delirium and coma in previous studies. Therefore, use for this indication of sleep should really be done with caution. Last of the IV medications we'll discuss today is dexmedetomidine. This medication is often used for sedation in the ICU. It's an alpha-2 agonist with anesthetic and sedative properties in the brainstem and peripheral vasculature. Adverse effects associated with dexmedetomidine include bradycardia and hypotension. The first case series evaluating the use of dexmedetomidine for sleep is published by Otto and colleagues in 2012. This was a very small study of 10 mechanically ventilated ICU patients. They used a dexmedetomidine bolus plus continuous infusion titrated to a goal RAS of negative 1 to negative 4. Using polysomnography, they found that patients actually had a severely disturbed sleep architecture with the use of dexmedetomidine. There was also no evidence of slow-wave sleep or REM sleep. There wasn't a control group in this study. However, a subsequent crossover study did include a control. This was, again, a very small study of 13 patients published in 2014. These patients were mechanically ventilated ICU patients. They compared dexmedetomidine continuous infusion versus no treatment. In using polysomnography in this subsequent study, they found an increase in sleep efficiency with no effect on slow-wave sleep or REM sleep. Again, I wanted to point out that these both have very small sample sizes. They also evaluated very specific patient population with no additional sedatives. They excluded patients with any delirium or sepsis. So, extrapolation of these results really should be taken with caution. Subsequent study by Wu et al. in 2016 was a randomized double-blind placebo-controlled study. This was the most robust study to date looking at dexmedetomidine used for sleep. They included 76 ICU patients greater than 65 years of age without mechanical ventilation after non-cardiac surgery. They evaluated dexmedetomidine at a set rate versus placebo for 15 hours. Using polysomnography, they found an increase in stage 2 sleep, an increase in total sleep time, sleep efficiency, subjective sleep quality on day 1, although no difference is seen on days 2 or 3 post-op. And they found ultimately no difference in slow-wave sleep, sleep fragmentation, or delirium. Furthermore, they also found no REM sleep in either group. Adverse effects were observed and they did find an increase in hypotension in patients that received dexmedetomidine as well as great acardia and total hospitalization. So although this information is certainly compelling in terms of increasing total sleep time and efficiency, ultimately the 2018 PADIS guidelines do not recommend the use of or have no recommendation, I should say, for the use of dexmedetomidine at this time really based off of limited data to support their use as well as increase in adverse effects that may be seen with dexmedetomidine. So we've gone over most of the information already contained in this slide. However, this is a visual summary taken out of a publication from 2016 in the British Medical Journal, which was a systematic review of pharmacologic interventions to improve sleep in hospitalized patients. The results I have highlighted here, again, focus specifically on slow-wave sleep and REM sleep, which are considered to be the most restorative and most impactful forms of sleep. This is to reiterate, again, that none of the continuous infusion medications really showed an improvement in stage 3 or REM sleep other than lorazepam. However, we know that there is evidence that benzodiazepines, including lorazepam, are known to suppress stage 3 sleep, so much of this data is very much conflicted. And all in all, we do not recommend the use of IV sleep medications for this reason at this time. So next, we'll talk about the data supporting enteral sleep medication options within the ICU. I'm going to kind of point out the limitations off the bat, that the majority of these studies evaluated enteral medications in outpatients, and therefore were excluded from this presentation. There are inpatient studies that I have highlighted here. However, ICU patients were largely excluded from these studies. There's also a large variation in study designs that we'll see, and very limited comparative data. So starting with trazodone, trazodone is used for the treatment of depression at lower doses that antagonizes 5-HT2A receptor and H1 receptor, and has sedative qualities. It has a black box warning for increased suicidality, and adverse effects include cardiac arrhythmias, orthostatic hypotension, and serotonin syndrome. Quetiapine is a drug that's used for the treatment of bipolar, schizophrenia, and depression. However, it also has sedative properties and is often used in the ICU. This is an atypical antipsychotic, and largely at the doses that we're using for sleep, it targets the histamine receptors within the body. Adverse effects include increased blood pressure, metabolic changes, drowsiness, and serostemia. Of note, this also has a black box warning for increased suicidality. There are two studies that looked at the use of quetiapine or trazodone in the inpatient setting. The first one was published in 2013 as an observational study of 61 psychiatric inpatients. They compared the use of trazodone versus quetiapine, and using a patient-reported Likert scale, they evaluated total sleep time and fragmentation. They found an increase in total sleep time with trazodone versus quetiapine, and a decrease in sleep fragmentation with the use of trazodone versus quetiapine. However, they do note that patients that received trazodone experienced an increase in GI side effects, including constipation, nausea, and diarrhea. Another study by Moret and colleagues in 1988 was a controlled open-label study of 10 hospitalized inpatients with depression, looking at trazodone alone, and using electroencephalography, they found an increase in total sleep time, a decrease in sleep latency, a decrease in REM latency, and decrease in sleep fragmentation, as well as an increase in sleep efficiency overall. As you can see from both of these studies, they both have a relatively small sample size. Some of the effects were seen at day 7, some as late as 5 weeks, based off of the study from 2013. Therefore, it's really important to take these with a grain of salt, and understand that given that this patient population is very different from the ICU patients, we still have insufficient data supporting the use of trazodone in the ICU at this time. Next, we'll review sedative hypnotics. These are often used for the treatment of insomnia outpatient. Mechanistically, they have a structure that's similar to benzodiazepines, with more targeted action at GABA type A receptor. Adverse effects include cognitive dysfunction, behavioral changes, risk of dependency withdrawal. They've also been associated with delirium in falls in elderly patients in the inpatient setting. The single study that looked at this in the intensive care unit, or post-intensive care unit, I should say, was from 2018. This was a retrospective observational study of patients after they left the ICU. They looked at esopoclone monotherapy versus combination of CBT. Using polysomnography, they found that there was a decrease in sleep latency with the esopoclone plus CBT versus esopoclonalone. Total sleep time and slow waves of sleep were increased in the combination group as well. There was no control group in this study. A subsequent study in 2019 was a single center prospective observational cohort study of 100 inpatients. Again, these are non-ICU patients looking at melatonin versus zolpidem. What they found was there was no difference in perceived sleep efficiency or sleep maintenance using a patient-reported scale. The last of the enteral medications we'll review today is mirtazapine. This is a tetracyclic antidepressant that's used for the treatment of depression. However, it does have sedative properties. Adverse effects include increased appetite, drowsiness, and xerostimia. Although we do use this in the ICU routinely, there unfortunately is no inpatient data published at this time from mirtazapine. To summarize, it's really important to assess for and address for underlying causes of insomnia in general, such as pain, thirst, nausea, or anxiety, and treat accordingly. Pharmacologic sleep strategies should also be optimized first, being that they are relatively low risk in general. As we now know, there's limited data that exists supporting the use of pharmacologic sleep aids in the ICU in general. If used, risks versus benefits should be considered prior to sleep medication administration. With that, I'll open this up to questions. Thank you.

sleep disruption

intensive care unit

ICU

circadian rhythm

melatonin

medications

negative outcomes

sleep medications

improving sleep