Thank you to the organizers for the invitation to speak. The title of my talk is Interventions for Good Sleep. My name is Elizabeth Wilcox. I'm an associate professor at the University of Toronto, and my research interests include understanding the interplay between cognition, sleep, circadian rhythm, during and after acute illness. I have no conflicts of interest to disclose. The learning objectives for my talk are the following. To understand what are the key factors in sleep and circadian disruption in acute illness, specifically in the ICU. Evaluate non-pharmacologic interventions to improve sleep and circadian disruption. Examine pharmacologic interventions to improve sleep and circadian disruption. And then lastly, to identify future studies to improve sleep and circadian disruption in the ICU. So what are the disrupting factors? The magnitude of ICU sleep and circadian disruption has been very difficult to quantify due to study heterogeneity, difficulties in measuring sleep itself in the ICU, and related limitations in crafting clinically meaningful definitions for sleep and circadian disruption. Notably, ICU sleep and circadian disruption can differ substantially between individual patients, between subgroups, and may even change over time as patients recover from their acute illness. Baseline risk factors for ICU sleep and circadian disruption have therefore been very difficult to establish. Consistently, however, patients report pain and anxiety as important risk factors for sleep and circadian disruption. And in addition, reporting poor sleep or using sleep medications at home prior to admission are factors that have been consistently implicated in ICU sleep disruption. And the impact of pre-admission sleep abnormalities on ICU outcomes and management has not been well studied. Thinking about the environmental risk factors for sleep and circadian disruption, though evidence is limited, inadequate daytime and excessive nighttime light are also common in the ICU and may significantly disrupt sleep and circadian rhythms. In addition, numerous studies demonstrate that the ICU environment exceeds recommended sound levels almost around the clock. Polysomnography studies have shown that excessive noise contributes to approximately 20% of awakenings in ICU patients. And also, frequent care interruptions are also big contributors to excess noise and light exposure can induce pain or anxiety, and they've been reported repeatedly by patients as being sleep disruptive. Associations between sleep disruption and severity of illness, length of stay, and or admission diagnosis have not been supported by most studies. However, loss of circadian alignment and amplitude have been variably associated with severity of illness, brain injury, and sepsis in different ICU cohorts. And then in terms of ICU care itself, additional factors associated with sleep and circadian disruption that have been hypothesized would include not only the need for mechanical ventilation, but the mode of mechanical ventilation and medication exposures, a big one being sedatives. And lastly, it's worth noting that cues to circadian entrainment that are likely blunted by ICU care may also include immobility as well as feeding, which is typically continuous in the ICU setting. Given the multifaceted nature of ICU sleep and circadian disruption, and the diversity of hypothesized and identified risk factors, interventions to promote sleep or circadian function tend to be complex and include multiple components. And this can lead to challenges for implementation and sustainability. Furthermore, little is known regarding intervention targets, appropriate timing of interventions, measurement of associated sleep and circadian outcomes, and which ICU sleep and circadian disruption domains are most closely linked to specific ICU important outcomes. So let's talk about non-pharmacologic interventions to improve sleep and circadian disruption. Now, control of the ICU environment has historically been a major focus when attempting to reduce sleep disruption. More recently, circadian principles have also been applied. There's been numerous studies that have explored interventions to control the environment, to cluster care delivery patterns, or combine elements of both. And in fact, multi-component sleep promotion bundles are guideline recommended for ICU patients. But let's quickly look at dynamic light therapy, the use of eye masks and ear plugs, and then look a little bit at these care bundles. More recently, some ICU sleep and circadian disruption interventions have really focused on reestablishing normal diurnal light variation. Effective circadian entrainment really depends on exposure to light that has a sufficient intensity and duration, and also has the correct spectral characteristics, specifically spectral characteristics mimicking natural sunlight. And though sample sizes are small, daytime light interventions have demonstrated some benefits, including increased subjective patient satisfaction, improved early postoperative mobility, reduced perioperative delirium, and normalization of circadian phase as determined by melatonin metabolites. Other daytime light studies have failed to show a benefit. However, high lighting levels in control groups and potentially inappropriate timing, duration, and spectral characteristics of light interventions may have limited findings in these studies. What I show here on the slide is one of the largest randomized control trials to date in the ICU, looking at dynamic light therapy as compared to control and its effect on cumulative incidence of delirium. This was a negative study in the intervention group for dynamic light therapy. Patients were exposed to bluish white light up to 1,700 lux between nine and 4 p.m., with the light being dimmed around noon for approximately an hour and a half. And the control group was exposed to approximately 300 lux of light during similar time periods. And the primary outcome was delirium incidence. The odds ratio adjusted was 1.24, with a 95% confidence interval that of course included one. So no significant difference in cumulative incidence of delirium with dynamic light therapy alone. And again, this is the largest study to date looking at dynamic light therapy. We can compare that to a more recent study. Again, a single center study that just looked at consecutive admissions into single rooms with either natural light exposure or what were considered dark rooms. You can see in the graph on the left-hand side that patients who were exposed to light, natural sunlight in the rooms that they were admitted to, typically the maximum amount of light that they were exposed to was 150 to 175 lux at a peak of two in the afternoon. Whereas in the dark rooms, patients rarely were exposed to light intensity that was higher than 25 lux. And again, here in a natural light experiment, delirium again was the primary outcome and there was no significant differences in incidence of delirium. However, in secondary outcomes, patients in the natural light group were less likely to require antipsychotics for agitation and less frequently reporting hallucinations. So this is kind of the best evidence that we have currently for light as a monotherapy. With regards to eye masks and or earplugs, Karimi and colleagues published last year a meta-analysis of 21 studies. 14 studies were included in the qualitative analysis, so underwent statistical aggregation, and that included over 2,500 patients. And what they showed was that separately or combined, either eye masks and or earplugs improve sleep in the ICU. The overall scores for either the Pittsburgh Sleep Quality Index and Richard Campbell Sleep Questionnaire improved with either eye mask and or earplug intervention. And so potentially a benefit with use of these instruments. This is just one example of ICU care bundles to improve sleep. In a before and after design, a multi-component sleep quality intervention that was led by Joseph Tona and Biren Kamdar attempted to reduce the environmental impact on sleep in an effort really to reduce delirium in their ICU. The intervention was a bundle of elements shown in the table here on the left. And as you can see, compliance with the different components of the sleep quality improvement interventions are also reported for 932 patient day slash nights. So 93% of the time, blinds were raised during the day. 59% of the time, caffeine was avoided after 3 p.m. 70% of the time, less than half of the day shift was spent napping. So again, trying to promote daytime wakefulness. 93% of patients had room lights dimmed later on in the day. And 93% had curtains closed and so on and so forth. Now, during the intervention period, again, 642 admissions were analyzed, over 300 either in the baseline or intervention period. And during the intervention period, patients experienced fewer days of delirium as compared with the pre-intervention period. And although the days delirious were decreased, the sleep quality ratings were unchanged. Sleep quality was measured here again by the Richard Campbell sleep questionnaire. So here, what we see is improved delirium without impacting perceived sleep. And this finding is not isolated to this study. Again, raising the important point of outcome selection in sleep and circadian dysfunction promoting interventional studies. Shifting gears to pharmacologic interventions to improve sleep and circadian disruption. To date, there is no ICU guideline recommendation supporting the use of pharmacotherapies to improve ICU sleep and circadian dysfunction. Nevertheless, medications are frequently prescribed for sleep in the ICU. Melatonin is the most commonly prescribed medication. And although the PADIS guidelines do not recommend medications to promote healthy sleep, a clear reality is that the critical care community has a desire to do something for their patients, even when evidence is weak, as all sorts of medications are used in the ICU and hospital for promoting sleep. These medications are typically given all sorts of times, both too early and too late. And importantly, many patients leave the ICU or hospital with prescriptions for sleeping pills. And although the best approach from a pharmacology standpoint is really to do less, less sedation during the day affords the opportunity for more participation in physio and occupational therapy and is more likely to promote sleep at night, but we're gonna quickly review the existing literature. More recent studies involving sleep promotion through pharmacotherapy have primarily involved melatonin agonists, alpha-2 agonists, and orexin antagonists so we're gonna quickly review the data for those. In this table, I've highlighted the literature on melatonin agonists. As you can see, the ICU populations are varied, as are the sample size, anywhere from 12 neurosurgical ICU patients to 172 medical surgical ICU patients. And in terms of sleep-related outcomes, the measures are also highly variable, looking at sleep duration, sleep efficiency, various sleep scores as comparators. And in addition, the intervention in terms of dosing of melatonin ranges anywhere from three milligrams to 10 milligrams as compared to placebo. And with all of these studies, there's really been no improvement in sleep-related outcomes and in fact, in the Abbasi study, which is the largest study before ProMedic, which I'm gonna speak about in a second, less sedation and physical restraint use was seen in the melatonin group, although delirium was actually found to be higher. But I think it's important to realize that in this study, the group that received melatonin actually were at higher risk for delirium as per their pre-derelict criteria. So importantly, this year, a very large multicenter randomized placebo-controlled double-blind trial came out. Patients were enrolled from 12 Australian intensive care units and patients were randomized to either receive four milligrams of liquid melatonin or placebo. They enrolled over 800 patients, more than 400 patients per group. And the intervention or control was administered entirely at nine o'clock at night for a total of 14 consecutive nights. And this study showed no significant difference in the average proportion of delirium-free assessments per patient. And importantly, ProMedic also showed no significant differences in any secondary endpoints, including ICU length of stay, hospital length of stay, mortality, and also quantity or quality of reported sleep. If we look at the data for alpha-2 agonists, dexmethomidine is one of the most recently introduced agents for sedation in the ICU. Dex has been shown to more closely create natural sleep than other GABA agonist agents. And when dex is used for sleep induction, overall, the results have been positive, but these studies have been small in number and therefore difficult to generalize. So as you can see, the number of patients enrolled ranges from 13 to 100. But in terms of sleep-related outcomes, there's increase in sleep efficiency, increased total sleep time pretty much across the board with varying doses of dex. I wanted to quickly highlight this study by Joanna Skrobik and colleagues, where they enrolled 100 delirium-free ICU patients. And they administered dex at night to a target RAS of minus one, or to a maximum rate of 0.7 mics per kilogram per minute. And 80% of patients who received dex remained delirium-free during their ICU stay compared to 45 patients in the control group. This really begs for a very large well-done study with subjective sleep assessments because no difference was seen in the lead sleep evaluation questionnaire score between these two groups, and possibly even an exploration of the dose and outcome relationship for dex in sleep improvement, sleep and circadian dysfunction improvement as well as delirium rates. In terms of studies on the horizon, I know of the MIND study, which is minimizing ICU neurologic dysfunction with dex-induced sleep. It's a trial of 370 patients. And it is a single-site parallel arm superiority trial. Patients who are over the age of 60 and undergoing cardiac surgery with planned bypass, but will be randomized to receive a sleep-inducing dose of dex or placebo. And again, the primary outcome is incidence of delirium on post-op day one. So this is gonna be an important study. And also likely there are more studies on the horizon to investigate oral dexmethomidine and clonidine and their impact on sleep circadian dysfunction and delirium in the ICU. Lastly, I wanna touch on orexin antagonists. There was one study recently published in 2018 in Psychogeriatrics. It looked at a mixed ICU patient population. It was a retrospective cohort study where patients received either suvarexant or no suvarexant and delirium incidence was lower. This was a non-statistically significant difference, however, in the suvarexant group as compared to control with an odds ratio after adjustment for delirium incidence of 0.23. There were no sleep-related endpoints, so further investigations are required. The study that I'm aware of looking at suvarexant and sleep and delirium in ICU patients is ongoing currently at Beth Israel. The population again is a post-cardiac surgery population of patients over the age of 60, 60 and older. They plan to randomize 120 patients, again, to a double-blind placebo-controlled trial of suvarexant at a dose of 20 milligrams or placebo. And their primary outcome, however, is not delirium, but wakefulness after persistent sleep onset with a secondary endpoint of total sleep time and incident delirium as exploratory endpoints. So more studies on the horizon here. One last important point to make about pharmacotherapy in the ICU is that some of the many medications we use to induce sleep actually have adverse outcomes on sleep architecture. If you review this literature on medications and sleep in the ICU, there's not a lot of data overall and much of it is conflicting. As an example, benzodiazepines and propofol, both GABA agonists, are frequently used for sedation in the ICU. Propofol is in fact recommended as first-line agent by current guidelines. Benzodiazepine administration results in decreased sleep latency, but adversely affects sleep architecture, decreasing slow-wave sleep and REM stages of sleep. Propofol, a potent suppressor of slow-wave sleep and at high doses can induce EEG burst suppression. Opioids, also commonly administered in conjunction with sedatives in critically ill patients, bind the mu receptors of the panto-thalamic arousal pathway, a key pathway in REM generation, and in a dose-dependent manner, suppresses both slow-wave sleep and REM. So I think it's really important when you're at the bedside with your team and you're trying to use pharmacotherapy to improve sleep that you not only think about timing, but you also think about reducing medication that might actually be interfering with sleep architecture and potentially propagating circadian dysrhythmia. So in summary, try to promote daytime wakefulness. This can be done non-pharmacologically with care bundles. Optimize the timing of sedating medications. Try not to use them throughout the day. Try to promote people being awake during the day and potentially receive sedating medications only in the evening. Review other medications that may impact sleep and cluster those medications as is possible. If you are gonna adopt a pharmacologic strategy, I think with ProMedic now, there's sufficient evidence to not typically use melatonin. However, the data is not strong enough yet for orexin antagonists and alpha-2 agonists. So we wait to see the results of ongoing studies. And lastly, it's really important to review medications for discontinuation prior to ICU discharge. Because as I stated, a lot of these patients are going home with these sleep aid medications, which may impact their recovery. I did mention that I wanted to briefly talk about the future of interventional studies. I just wanna highlight this important paper that was led by Melissa Kennard and Najeeb Ayas, looking at causes, consequence, and treatment for sleep and circadian disruption in the intensive care unit. The results of which outline future directions for sleep and circadian dysfunction trials in the ICU. And with that, I'll say thank you.

sleep disruption

ICU

risk factors

non-pharmacologic interventions

pharmacologic interventions

ICU environment