Hello, everyone. Thank you so much for your attention today. Today's presentation will be on barriers to improving sleep quality. A little bit about myself before we get started. My name is Reema Bhojram. I'm the lead medical ICU pharmacist at the UCSF Medical Center, as well as a pharmacist in the post-ICU, post-COVID, also known as Optimal Clinic at UCSF. I also am a health sciences assistant clinical professor at the UCSF School of Pharmacy, co-chair of the UCSF Hospital-Wide Sleep Group, as well as the UCSF ICU Liberation Sleep Subcommittee. My research areas include pain, sedation, sleep, delirium prevention, optimizing medication-related outcomes with a focus specifically on quality improvement in the hospital setting. I have no financial disclosures or conflicts of interest today. My objectives for today are to review endogenous and exogenous factors influencing sleep in the ICU, discuss modifiable and non-modifiable barriers to sleep, as well as describe patient environmental and cultural barriers to sleep optimization in the ICU setting. Before I delve into the basics, I wanted to take a minute to reiterate the importance of sleep. The importance of sleep, as we know, was highlighted not so long ago in the 2018 Clinical Practice Guidelines for the Prevention and Management of Pain, Agitation, Sedation, Delirium, and Immobility and Sleep Disruption in adult ICU patients. We care about sleep because we know it's common, and unlike many things in the ICU, it's a modifiable risk factor. We also know that poor sleep has been associated with worse clinical outcomes, including increased distress, delirium, duration of mechanical ventilation, dysregulation of immune function and neurocognitive function. Sleep disturbances are defined as disorders of initiating and maintaining sleep, disorders of excessive somnolence, sleep-wake cycles, sleep stages, or partial arousals. While there are no standardized methods of measuring sleep disturbances in the ICU, we know that sleep disturbances are very common, occurring in about 60% of patients discharged from the ICU. We also know that sleep disturbances are generally characterized by an increase in fragmentation, abnormal circadian rhythms, light sleep, also known as N1 and N2 sleep, as well as a decrease in slow-wave sleep, also known as N3 sleep, and rapid eye movement, or REM sleep. Definitions of sleep quality largely depend on the study. Measurement modalities in studies are somewhat variable. However, a common objective assessment for sleep quality is the percent of time in stage 3 or 4 non-REM sleep and REM sleep. A common subjective assessment for sleep quality can be performed using what we call the Richard Campbell Sleep Questionnaire, which is a validated scale in both intubated and non-intubated ICU patients. Risk factors for poor sleep include female gender, older age, reported poor quality of sleep at home, regular use of sleep aid medications at home, and specific pre-morbid medical conditions including hypertension, diabetes, cancer, and thyroid disease. Some of this may be a review in light of the last session focused on pathophysiology of sleep. However, just to reiterate that there are various endogenous and exogenous influencers that exist for sleep, and we know that these are commonly disrupted in critically ill patients. Endogenous changes that occur in the setting of critical illness include circadian rhythm changes, which we believe impact REM sleep and potentially cognitive processes, as well as neurohormonal changes. Exogenous factors include fluctuations in stress, nutrition, sleep deprivation, which impacts the homeostatic drive to sleep, and ultimately slow-wave activity tied to metabolic growth and recovery. Exercise, temperature, sensorial stimulation, or what we call Zeitgebers, such as light, which serve as sensorial cues to simulate endogenous changes, and exogenous supplementation of hormones through medications such as melatonin. Collectively, these endogenous and exogenous factors, when impaired, contribute to increased inflammation, sympathetic tone, arrhythmia, insulin resistance, and glucose tolerance, which may impair cardiovascular, immune, and metabolic responses in the setting of impaired sleep. We also know that there are many modifiable and non-modifiable influencers for sleep. Summarized here are the findings from a recent observational prospective single-center study of 137 patients in a mixed ICU setting, published by Martinez and colleagues in June 2022. In this study, they evaluated the modified versus non-modified risk influencers for sleep. What is notable in their findings is that there truly is a vast array of influencers of sleep overnight for ICU patients. While some influences relating to underlying pathophysiology or required procedures may not be modifiable, a significant number of influencers overnight are either modifiable or potentially modifiable for most patients. While the array of influencers to sleep is remarkable, the question then becomes, how often are there actual interruptions at play? The answer we know from the two studies summarized here is, unfortunately, very often. In a study by Friedman and colleagues in 2001, polysognography was utilized to evaluate sleep architecture in 22 mechanically ventilated patients. They found a total sleep time that widely varied. Fragmented sleep-wake cycles in the daytime and at nighttime were common. Noise caused about 11.5% of arousals and 17% of awakenings. And there was, on average, about 1.9 plus or minus 2.1 arousals per hour of sleep. This was also associated with daytime sleepiness, nighttime fragmented sleep, in addition to short durations in the light sleep stages, as well as a decrease in the restorative sleep stages. In a subsequent study in 2019 looking at MICU patients utilizing a sleep protocol versus usual care, they found the utilization of sleep protocol was associated with fewer room entrances and decreased sound overall. However, notably, the length of time in between room entrances was still on average of 45 minutes, even after the sleep protocol, which ultimately lends itself to a very fragmented sleep cycle overall for ICU patients. This slide is somewhat busy, but what I'm showing you here is the summary of a single night of interruptions for a real patient in the ICU. The number of interruptions here is 26. This sounds extreme, but it's actually the average number of interruptions we discovered in our units at UCSF, which is actually nearly half of the interruptions cited in literature. Previous literature have cited averages up to 51 nocturnal care interactions per night. As you can imagine, these activities bring sound, pain, discomfort, in addition to an already physically and emotionally taxing evening for our patients. We have discussed many barriers to sleep in the setting of endogenous versus exogenous, modifiable versus non-modifiable influencers, and the exorbitant number of interruptions to sleep. I would be remiss if I didn't mention some of the practical limitations to sleep in the ICU, however. Much like peeling an onion, there are many layers to optimizing sleep in the ICU, including cultural, environmental, and patient-specific factors that I'll go over in the next couple of slides in more detail. Starting with cultural barriers. In order to implement change within any hospital system, change management is extremely important. Inadequate culture shift planning, lack of shareholder involvement, flawed communication and leadership strategies may result in transient changes to sleep optimization at best. Creating a safety culture within an organization is extremely important to ensure bi-directional feedback is obtained within any quality improvement project. Additionally, it's important to acknowledge that many of the interruptions that play overnight, whether it be blood pressure cuff cycling overnight every hour, frequent lab checks, scheduled pain assessments, medications, may be appropriate in some instances but may be influenced by emotional decisions in other instances. And shareholders weighing in on changes within a medical system may come in with their own biases and sensitivities, which can make changing practices to promote sleep a little bit more complicated. Automation, while it improves our efficiency, puts us in a new situation where it's easier for healthcare professionals to go on autopilot. Whether it's the nurse programming the blood pressure cuff to cycle every 15 minutes and then getting distracted with another patient or forgetting, or a provider scheduling Q4 hour blood sugar checks, Q6 hour medications, or labs, it takes someone actively reviewing interruptions, which may be present in different areas within the charting system, in efforts to adjust or bundle care sometimes. Automation also makes it such that it's almost too easy at times to order an additional data point. The extra lab check to ensure we're trending values without realizing that collective volume of influencers of sleep that may be present and the potential unintended downstream effects of impaired sleep. Environmental barriers are also plentiful. We know that lights, noise, bed comfort, positioning, patient monitoring interruptions, procedures, temperature, visitors, or the lack thereof, the patient next door that may be coding or altered, staffing interruptions, all can influence sleep significantly. And lastly, we have patient barriers. Illness itself can be associated with physiologic changes that impact sleep. Pain, discomfort, anxiety, restlessness, sadness, thirst, nausea, delirium, the presence of preexisting sleep disorder can impact sleep, and of course, trauma or depression. It's important for providers to investigate these things when they're occurring. And unfortunately, it's not as simple as saying this patient is not sleeping, therefore prescribing a sleep medication. At UCSF, we found the most common self-reported barriers to sleep within our hospital system were thirst, nausea, pain, and anxiety, most of which would not be treated by, let's say, melatonin or a traditional sleep medication. Prescribing a sleep medication in the instance where someone may have untreated pain may actually cause harm for the patient. While we saw earlier in the presentation that staff interruptions overnight are common, unfortunately, these are not consistently providers necessarily who may have the ability to adjust the orders per se. In fact, at most hospitals, cross-cover staff may be covering several units, therefore investigating and troubleshooting these patient barriers to lack of sleep is not consistently performed. So in summary, we've talked about many endogenous and exogenous alterations in sleep architecture that serve as barriers. While there are many non-modifiable barriers to sleep, there are also many modifiable or partially modifiable barriers that exist. We've talked about sleep fragmentation, and we know that there are many patient, environmental, and cultural barriers to sleep that should be considered with any attempts to optimize sleep in the ICU setting. While this portion of the presentation was focused on barriers to sleep, I do want to leave you with a glimmer of hope prior to the next presentation, which focuses on interventions for good sleep. This is a quote from the Dalai Lama that says, if you think you're too small to make a difference, try sleeping with a mosquito. We did a lot of quality improvement at UCSF relating to sleep, and we're always amazed that sometimes it's the smallest things that make the biggest differences. So I wanted to thank you very much for your attention today, and please do not hesitate to reach out if I can answer any questions. Thank you. Here are my references. Thank you again for your attention.

sleep quality

ICU

barriers

clinical outcomes

sleep disturbances