INTRODUCTION — The need for sleep has long been assumed to be important for recovery from injury and sickness, and there is an emerging understanding of the restorative role of sleep in health and disease. Unfortunately, the hospital environment is often poorly conducive to sleep [1,2]. Pain, anxiety, medication effects, medical interventions, environmental noise and light, and the acute illness itself all contribute to decreased quality and quantity of sleep in hospitalized patients. As a result, issues related to sleep and sleep disorders are important to inpatient care.
This review will discuss the evaluation, consequences, and management of sleep disturbances in hospitalized adult patients. The evaluation and management of obstructive sleep apnea and other sleep disorders in the inpatient setting is reviewed separately. (See "Obstructive sleep apnea and other sleep disorders in hospitalized adults".)
SLEEP DISTURBANCES IN THE HOSPITAL — Hospitalized patients experience fragmented and poor quality sleep and are at risk for marked circadian rhythm disturbances. These changes become more pronounced with worsening severity of illness as well as in the immediate postoperative period. In some cases, sleep impairments persist well beyond the acute illness [3-6].
Poor sleep quality — Patients regularly report poor quality sleep during hospitalization across a range of inpatient settings [7-10]. Impaired sleep quality has been cited by patients as one of the main stressors during hospital admission [11,12]
Patients in the intensive care unit (ICU) exhibit significant alterations in sleep architecture. Sleep is highly fragmented, with prolonged sleep latencies, frequent arousals, poor nocturnal sleep efficiency, an increase in stage 2 (N2) sleep, a reduction or absence of deep or slow wave (N3) sleep, and a reduction or absence of rapid eye movement (REM) sleep [13-16].
Sleep changes in the ICU may occur to such an extent that traditional rules for scoring sleep stages may not be applicable [17-20]. Nearly half of intubated, difficult to wean patients in one prospective study had atypical sleep that could not be classified by standard criteria, and 20 out of 45 patients had no observable REM sleep [21]. REM sleep can also be reduced or absent on the night following surgery, which may be followed by a significant rebound of REM sleep on subsequent nights in some patients [22-24].
Circadian rhythms can be altered during hospitalization, especially in ICU patients [25]. Severity of illness, including encephalopathy, as well as certain medications, appear to be associated with the degree of circadian disturbances [26]. Patients in the ICU may have circadian rhythms that are significantly delayed or even absent, particularly in the setting of severe illness or sepsis [26-30]. The phase delay seen in some patients is likely related to a free-running circadian pacemaker that may not be responsive to bright light, the strongest external cue that aligns endogenous circadian rhythms with the environment [29,30].
Older adults are also vulnerable to loss of circadian rhythms, which may be associated with the development of delirium [31,32]. Circadian rhythm disturbances also contribute to fragmented and inadequate sleep. (See "Overview of circadian sleep-wake rhythm disorders".)
Reduced total sleep time — Average sleep duration over a 24-hour period may be reduced during hospitalization, although there is considerable variability in reports. A cross-sectional survey study involving 39 hospitals in the Netherlands and nearly 1500 patients found that subjectively reported total sleep time was 83 minutes shorter among inpatients compared with their habitual sleep times at home over the prior month [33]. For many acutely ill patients, sleep appears to be spread around the clock, with 40 to 60 percent of sleep occurring during typical waking daytime hours (figure 1) [13,14,17].
Acute and chronic insomnia — Extrapolating from the general outpatient adult population, a relatively high percentage of patients admitted to the hospital will have chronic insomnia, whether previously diagnosed or not. In particular, older adults and psychiatric patients have very high rates of chronic insomnia. Using standardized diagnostic tools, approximately 40 to 60 percent of patients older than 65 years of age [34-36] and 80 percent or more of psychiatric inpatients [37,38] suffer from some form of chronic insomnia. Patients with chronic insomnia admitted to the hospital may be expected to continue to have problems during admission, and the insomnia may be exacerbated by the factors discussed below. (See 'Contributing factors' below.)
The prevalence of acute inpatient insomnia (distinct from sleep deprivation) has not been well studied. One study found that approximately one-third of general medical inpatients reported acute insomnia (mostly trouble maintaining sleep), but only 3 percent met criteria for severe insomnia, and 75 percent of cases resolved within two weeks of discharge [39].
Consequences of poor sleep — The consequences of acute and chronic sleep deprivation in the general population are well documented. Data on the impact of poor sleep quality and loss of sleep during hospitalization are more limited, but many of the same effects observed in the general population are relevant to the inpatient setting. Emerging data support the potential impact these can have on clinical outcomes. (See "Insufficient sleep: Definition, epidemiology, and adverse outcomes".)
●Cognitive and neurobehavioral consequences – Sleep deprivation leading to cognitive impairment and neurobehavioral changes is well documented in the outpatient setting [40,41]. These consequences have not been systematically studied in inpatients, where it is often more difficult to discern a cause and effect relationship. While it is postulated that a lack of sleep and loss of circadian rhythms are risk factors for delirium in older adults and in the ICU setting [31,32,42], the association between these conditions remains to be better defined.
●Respiratory function – Acute sleep deprivation under experimental conditions impairs several parameters of respiratory function, including hypoxic and hypercapnic chemosensitivity, control of ventilation, genioglossus muscle activity, inspiratory muscle endurance, and spirometry (in patients with chronic obstructive lung disease) [43-47]. In the ICU setting, early sleep disturbances and "atypical sleep" patterns on polysomnography have been associated with adverse outcomes including late noninvasive ventilation failure and prolonged weaning times [21,48].
●Immune function and inflammation – The role of sleep deprivation in immune function has been studied in animal models and otherwise healthy humans, suggesting a link between sleep loss and impaired immune function [49-53]. Whether physiologic changes in immune function and inflammation related to sleep deprivation lead to changes in outcomes in hospitalized patients with poor or insufficient sleep is not yet known.
●Metabolism – Sleep loss may affect insulin resistance and alter glucose metabolism. A study of 212 general medical ward patients found an inverse relationship between duration of sleep in the hospital and glucose control in both diabetics and nondiabetics [54].
●Anxiety and pain – Acute sleep deprivation is well recognized to increase anxiety and reduce the pain threshold [55,56].
CONTRIBUTING FACTORS
Environmental factors — Many aspects of the acute care hospital environment can adversely impact sleep, especially noise, patient care interruptions, and alterations in the light-dark cycle [33,57]. These factors only add to the inherent challenges of sleeping in an unfamiliar environment.
Noise — Noise in the hospital comes from a variety of sources, including talking, phones, televisions, music, pagers, overhead paging, medical equipment, and alarms on intravenous pumps, cardiac monitors, oximeters, and ventilators, and is often cited as the most common cause of sleep disruption [57]. Noise levels in the hospital routinely exceed those recommended by the World Health Organization for a good night's sleep [58,59]. Studies using polysomnography in the intensive care unit (ICU) have found that about 10 to 20 percent of arousals can be attributed to noise [13,17,60]. Louder sounds such as alarms are the most disruptive [61].
Patient care interventions — Caring for acute illness requires numerous patient interactions with the potential to disrupt sleep, and the frequency of patient care interactions for vital signs, blood draws, diagnostic testing, and treatments increases with the severity of illness [62]. These care interactions are commonly cited reasons for poor sleep in hospitalized patients [33,57,63].
Prospective studies in the ICU setting indicate that between 5 and 15 percent of nocturnal interventions are potentially modifiable causes of arousals [60,64]. In a study of seven mechanically ventilated patients monitored with polysomnography, 7 percent of arousals and awakenings were recognized as being from patient care interruptions [60]. Rearrangement of workflows and novel decision-support tools could conceivably reduce some of these interruptions, depending on the location of patient care. (See 'Interventions to improve sleep' below.)
Light exposure — Patients in the ICU often report nocturnal light exposure as one of the more important factors disrupting their sleep [10,57,63,65]. Just as important may be inadequate levels of daytime light.
Several studies have found relatively low levels of nocturnal light exposure in the ICU (<2 to 4 lux) that would be in the range conducive to sleep [15,26,30,66]. However, the same studies also found relatively low levels of average daytime light exposure (70 to 80 lux). These levels are equivalent to a dim light environment in which entrainment of circadian rhythms may be lost. Low daytime light levels have also been found on the medical ward [2].
Nonenvironmental factors — There are a number of issues related to being acutely sick, aside from the hospital environment, which may significantly impact sleep quality and duration. Many of these factors are potentially modifiable.
Acute illness — Serious illnesses as well as the postoperative period are often associated with pain and anxiety. Acute pain may lead to prolonged sleep latency, fragmented sleep, and reductions in slow wave (N3) and rapid eye movement (REM) sleep [67-69].
Acute illness itself may adversely impact both sleep quality and quantity. Severe circadian rhythm disturbances may occur with illness, especially sepsis [27-30]. Fever also has variable and complex effects on sleep [70]. (See 'Poor sleep quality' above.)
Symptoms such as acute cough or severe diarrhea can make consolidated sleep nearly impossible. A systematic review of risk factors for poor sleep in the ICU found that pain, shortness of breath, hunger, and nausea were more commonly reported in patients with poor sleep [57]. Chronic obstructive pulmonary disease, asthma, congestive heart failure, and gastroesophageal reflux all have the potential to worsen at night, leading to fragmentation of sleep, prolonged awakenings, and loss of sleep [9]. Unrecognized depression may be associated with poor sleep, particularly in older inpatients [71]. Insomnia may also be a symptom of alcohol withdrawal or substance intoxication.
Medications — Many of the medications used in the ICU and other inpatient settings can disturb sleep and alter sleep architecture (table 1). Examples include vasopressors, sedatives and hypnotics, opioids, glucocorticoids, beta blockers, some respiratory medications, and specific classes of antibiotics [72]. (See "The effects of medications on sleep quality and sleep architecture" and "Risk factors, comorbidities, and consequences of insomnia in adults", section on 'Medication and substances'.)
Ventilators and other equipment — A variety of mechanical interventions used to care for severely ill patients (eg, endotracheal tubes, nasogastric tubes, urinary catheters, external orthopedic hardware) can impact sleep quality and duration due to their physical presence and associated alarms. Changes in physiology induced by certain ventilator settings can also affect sleep efficiency [73]. (See 'Ventilator modes' below.)
INTERVENTIONS TO IMPROVE SLEEP — There is no single intervention that will reliably optimize sleep in the inpatient setting. Incremental improvement is probably best achieved through enhanced awareness of the problem, multiple small interventions at an individual patient level, and hospital-wide efforts to make the inpatient environment as conducive as possible to obtaining good sleep.
Assess and treat underlying sleep disorders — Sleep disorders such as obstructive sleep apnea, restless legs syndrome, and chronic insomnia are prevalent in inpatient populations and frequently unrecognized. Evaluating for these conditions and implementing targeted therapies has the potential to improve sleep during hospitalization. (See "Obstructive sleep apnea and other sleep disorders in hospitalized adults".)
Sleep apnea is a common cause of poor quality sleep and requires a high index of suspicion to diagnose. Inpatient populations with a particularly high prevalence of sleep apnea include patients with heart failure, cardiovascular disease, and stroke. (See "Sleep-disordered breathing in heart failure" and "Obstructive sleep apnea and cardiovascular disease in adults" and "Sleep-related breathing disorders and stroke".)
Treatment of the acute illness — Acute illness results in physiologic derangements and symptoms that impair sleep. As such, treatment of the underlying medical or surgical condition, including treating pain, dyspnea, and other symptoms, should improve sleep quality and duration.
Review of concomitant medications — Careful consideration should be given to medication side effects, especially when patients voice sleep-related complaints. A variety of medications used in the acute care setting can alter sleep architecture and impair sleep quality and duration (table 1). Medications with activating properties should be given in the morning and sedating medications in the evening, when possible. (See 'Medications' above.)
In some cases, medications within the same class of drugs or used for a common effect can have differential effects on sleep. Beta blockers that do not cross the blood-brain barrier (eg, atenolol) have less risk of causing sleep disturbances than those that do (eg, propranolol, metoprolol). (See "The effects of medications on sleep quality and sleep architecture", section on 'Beta blockers'.)
In the intensive care unit (ICU), data suggest that sedation with dexmedetomidine mimics stage N2 non-rapid eye movement (NREM) sleep, with little associated N3 sleep and no REM sleep [74]. Benzodiazepines and propofol have variable effects and may contribute to sleep disruption in some cases [75-77]. A systematic review found the data insufficient to determine whether propofol improves sleep quality or duration in ICU patients [77].
Ventilator modes — Further studies are needed to determine whether certain ventilator modes or types improve sleep-related outcomes. Maneuvers that have been suggested to be helpful include adding dead space to nocturnal pressure support ventilation or changing to proportional assist ventilation at night, with the goals of reducing central apneas and ventilator asynchrony [73,78]. One study found that nocturnal ventilation via pressure support, compared with spontaneous breathing, was associated with longer sleep duration (51 minutes per night) but no change in sleep quality in difficult-to-wean tracheostomy patients [79], suggesting that nocturnal support may help with sleep in some populations.
A 2015 systematic review that included six studies evaluating ventilator modes and sleep concluded that certain ventilator settings might offer benefits over others (eg, fewer central apneas with assist control mode, less patient-ventilator asynchrony with proportional assist mode), but the quality of the evidence was low and results were not consistent across all trials [80]. In addition, data on outcomes other than sleep and sleep-related respiratory parameters are lacking.
Nonpharmacologic strategies — A variety of nonpharmacologic interventions have been proposed to help enhance sleep quality and duration in hospitalized patients [80]. Given the often small resource allocation required and the relative safety of these interventions, they should generally be considered as first-line maneuvers to improve sleep.
●Noise reduction – Several methods have been used to reduce nighttime noise exposure in the inpatient setting, including earmuffs or ear plugs for patients, sound masking (white noise), installing sound proofing acoustic materials, closing doors, reducing volume levels of medical equipment, and behavioral modifications (eg, "quiet time" protocols). While the overall quality of evidence is low and the primary outcomes are mostly subjective, studies consistently find that patients report modest improvements in sleep with these relatively simple interventions [81]. Even though environmental noise reduction protocols typically only achieve small decreases in noise levels, subjective and observed sleep quality and duration appear to be improved.
●Light therapy – A few small trials have found that use of eye masks in conjunction with ear plugs improves sleep in the hospital [82-85]. As noted above, however, nocturnal light appears to be less of an issue in modern hospital settings compared with inadequate daytime light exposure. (See 'Light exposure' above.)
Other studies have evaluated enhanced daytime light exposure in hospitals to improve circadian rhythms and nocturnal sleep. Light exposure typically consists of two to five hours of daytime to early evening light in the range of 2000 to 5000 lux [86-88]. Most have shown improved total sleep time with daytime light exposure, although the effects are modest, sleep duration has primarily been subjectively reported, and the overall quality of the evidence is low. In addition, this may not be effective in certain patient populations (eg, septic ICU patients, patients with decompensated cirrhosis) [29,89]. A pilot study utilizing morning bright light and evening short-wavelength filter glasses found less daytime sleepiness and improved moods in medical inpatients, but no effect on other sleep measures [90].
●Reducing nighttime interruptions – Studies suggest that many nocturnal interruptions could be avoided, and that altering workflows may potentially enhance patient sleep [64,91]. This might include delaying or minimizing nocturnal vital signs, using passive vital sign monitoring, rescheduling the timing of nocturnal medications, and delaying morning phlebotomy and x-rays.
For patients on the general medical service, one trial demonstrated that the mean number of nighttime vital signs checks could be safely decreased by approximately one-third through use of a medical record-based prediction algorithm that offered clinicians the choice to discontinue overnight vital signs in low-risk patients [92]. In a multicomponent protocol that included minimizing patient care interactions during an eight-hour nocturnal "quiet time" in patients on a medical surgical unit, there was a 49 percent reduction in as-needed sedatives, although overall sleep quality did not improve [93]. A pilot study examining restricting, nonurgent bedside care between midnight and 4 AM in ICU patients found that the sleep protocol resulted in 32 percent fewer room entries, nine fewer minutes of in-room activity, and lower noise levels during the restricted period compared with a control group [94]. Sleep-specific outcomes were not reported.
While a strategy focused on limiting night-time interruptions makes intuitive sense, further study is needed to determine how best to safely omit nighttime patient care interruptions and how this may help consolidate sleep, improve sleep quality, and impact patient outcomes.
●Relaxation techniques – There is limited, low quality evidence that relaxation techniques such as music, massage, guided imagery, and aromatherapy achieve minor improvements in subjective or nurse-determined sleep quality and duration [80,86,95]. Given the safety and relatively low cost of many of these interventions (eg, music), implementation is worth considering on a case-by-case basis.
●Multifaceted protocols – At an institutional level, multifaceted protocols aimed at improving sleep and decreasing use of sedative-hypnotic sleep aids have shown promise [96-98], although further work is needed to verify that such protocols consistently improve patient-centered sleep-related outcomes. Programs require broad cultural and behavioral shifts, and ensuring adherence to interventions may be challenging [99]. The most effective strategies likely combine efforts to create a more sleep-conducing inpatient environment with prescriber and caregiver education and feedback [98].
Pharmacotherapy — Patients may request sleep aids during hospitalization, and prescribing these medications appears to be a fairly common practice. A single-institution review found that 26 percent of inpatient admissions were provided with sleep aids [100]. Another study found that 16 percent of inpatients aged 65 years and older were given new prescriptions for benzodiazepines or sedative-hypnotics during admission, and the majority (77 percent) of these were deemed by the authors to be inappropriate [101]. While it is generally recommended that the nonpharmacologic interventions reviewed above be tried first [102], sleep aids may still be needed in some cases.
Medications used in the treatment of insomnia are reviewed separately (see "Pharmacotherapy for insomnia in adults"). The following considerations are relevant to use of these medications in the inpatient population. Selection of an agent should be individualized based on severity of symptoms, age and comorbidities, side effects, and drug-drug interactions.
Among the various drugs and classes, melatonin is often the best first-line choice, based on its mild side effect profile, low potential for drug-drug interactions, and potential to improve circadian rhythms. The sleep-inducing properties of melatonin are variable and may not be as robust as other sleep aids [103], although a study of inpatients receiving sleep aids found that patient perceptions of sleep disturbance and effectiveness were similar with melatonin and zolpidem [104]. Use of melatonin in inpatients appears to be increasing [105].
●Melatonin and melatonin receptor agonists – Several small studies in hospitalized patients and in simulated environments have found improvements in sleep and sleep quality (subjective and as measured by actigraphy) with 1 to 5 mg of melatonin at night [106-108], although an effect has not been consistently demonstrated in all studies [109]. Melatonin is well tolerated, and proper timing of the dose may help to regulate circadian rhythms.
While dosing is not standardized and there may be variability across different formulations, a typical dose is 1 to 3 mg scheduled at 9 to 10 PM, depending on the patient's habitual bedtime. Immediate release melatonin preparations should be given 30 to 60 minutes before bedtime, and sustained release preparations should be given one to two hours before bedtime.
Melatonin receptor agonists such as ramelteon have not been examined for their effects on inpatient sleep, although limited data suggest that they may play a role in preventing delirium in older adult inpatients [110]. (See "Delirium and acute confusional states: Prevention, treatment, and prognosis", section on 'Medications to prevent delirium'.)
●Trazodone – Trazodone, a serotonin modulator with significant sedation as a side effect, is generally better tolerated than benzodiazepines, without the risk for physiologic tolerance or dependence. In a small observational study of hospitalized psychiatric patients, trazodone appeared to be a more effective sleep aid than quetiapine based on patient-reported and nursing-observed sleep outcomes [111].
The most common side effects of trazodone are headache, dry mouth, and nausea. Additional potential side effects relevant to the inpatient population include orthostatic hypotension and infrequent atrial and ventricular arrhythmias, and thus caution should be used in patients at risk for these problems [102]. When trazodone is used, we suggest starting at a low dose (eg, 50 mg at bedtime) with close attention to drug-drug interactions.
●Nonbenzodiazepines – The nonbenzodiazepine benzodiazepine receptor agonists (eg, zolpidem, eszopiclone/zopiclone, zaleplon) are effective sleep aids in the outpatient setting and commonly used in the inpatient setting [100,101,112]. However, increasing reports of these medications being associated with cognitive dysfunction, delirium, and falls in hospitalized patients have tempered enthusiasm for their use in this setting [112-116]. This class of medications should probably be avoided in older adult inpatients, and use in other inpatient populations requires close monitoring for side effects.
●Benzodiazepines – While benzodiazepines are effective at reducing sleep latency and increasing total sleep time, they should be used with great caution in the inpatient setting. Benzodiazepines are associated with significant adverse effects, particularly in older adults, including respiratory depression, cognitive decline, delirium, daytime sleepiness, and falls [117]. In hospitalized adults, use of a benzodiazepine has been associated with increased risk for mechanical ventilation, delirium, and falls, independent of age and comorbidities [118]. Additional drawbacks to their use in the inpatient setting include long half-lives, drug-drug interactions, and the potential for both dependence and withdrawal insomnia.
●Antihistamines – Diphenhydramine is the most common ingredient in over-the-counter sleep aids, and the first-generation antihistamines have generally been felt to be safe due to their long history of use. However, trials evaluating their effectiveness as sleep aids are limited and show mixed results [119]. In addition, many potential side effects may be enhanced in the inpatient setting, including impaired cognition, anticholinergic effects (including constipation and urinary retention), and cardiac toxicity [120].
●Low-dose doxepin – Doxepin is a tricyclic antidepressant that at low doses (3 to 6 mg) has antihistaminergic effects and has been used as a sleep aid primarily for sleep maintenance insomnia in the outpatient setting [121]. Inpatient data is limited to a single small retrospective study of using low-dose doxepin for insomnia in depressed inpatients that did not show improvements in insomnia measures over a four-week course of treatment [122]. Similar precautions to other antihistamines are warranted with doxepin use.
●Other medications – A variety of other medication classes (anticonvulsants, antidepressants, antipsychotics, barbiturates) have been used in the management of insomnia and are reviewed separately (see "Pharmacotherapy for insomnia in adults"). These medications have not been well studied in the inpatient setting and all carry potential risks.
The dual orexin receptor antagonists (suvorexant, lemborexant, daridorexant) have not been well studied in the inpatient setting [123]. Limited data suggest that suvorexant may help to prevent delirium in the ICU [124], and a single retrospective study of post-stroke patients reported improved subjective sleep quality with suvorexant when combined with ramelteon [125]. However, caution with this class of agents is advised in hospitalized patients due to their longer half-lives, potential for daytime somnolence, and certain drug-drug interactions [126]. (See "Pharmacotherapy for insomnia in adults", section on 'Dual orexin receptor antagonists'.)
Patients with chronic insomnia — Patients admitted to the hospital with an established diagnosis of chronic insomnia who are on chronic pharmacotherapy should be reassessed to determine the need for ongoing treatment. If ongoing therapy is deemed necessary, important considerations include drug-drug interactions, the presence of new organ dysfunction that may affect drug metabolism or clearance, and altered susceptibility to side effects due to concurrent illness.
Depending on the clinical circumstances, dose adjustments, temporary discontinuation, or alternative therapies may be needed. As an example, a sedative hypnotic drug being taken for chronic insomnia may need to be held in a patient who is admitted with mental status changes or significant respiratory failure. However, clinicians should be alert to the potential for withdrawal syndromes and withdrawal insomnia with abrupt cessation of some insomnia medications (eg, benzodiazepines, benzodiazepine receptor agonists, some antidepressants, orexin receptor antagonists).
For patients with previously untreated or newly diagnosed chronic insomnia found during a hospital admission, the urgency with which to initiate therapy depends on multiple factors, including the severity of the insomnia, the current illness, comorbidities, and current medications. For most patients, treating acute medical and psychiatric conditions and then reassessing the insomnia as an outpatient, prior to initiating therapy, is prudent. The outpatient setting is also better suited for discussion and initiation of behavioral therapies as an alternative to first-line pharmacotherapy. (See "Overview of the treatment of insomnia in adults".)
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Insomnia in adults".)
SUMMARY AND RECOMMENDATIONS
●Background – Sleep quality is often impaired in patients hospitalized for acute illness. Common disturbances include fragmentation of sleep, altered sleep architecture with decreases in slow wave (N3) and rapid eye movement (REM) sleep, and disruption of circadian rhythms. The most severe disturbances are observed in the intensive care unit (ICU) setting. (See 'Sleep disturbances in the hospital' above.)
●Consequences of poor sleep – Indirect evidence from the general population suggests that substandard sleep in the hospital could lead to adverse effects on cognition, neurobehavioral outcomes, respiratory function, immune function, metabolic function, anxiety, and pain. In addition, sleep deprivation has been proposed as a risk factor for delirium. (See 'Consequences of poor sleep' above.)
●Causes of poor sleep in the hospital – Common factors implicated in poor sleep in the hospital include noise, patient care interventions, light, the illness itself (eg, pain, anxiety, fever), and treatments for disease (eg, medications, ventilators). (See 'Environmental factors' above and 'Nonenvironmental factors' above.)
●Interventions to improve sleep – A variety of interventions to improve sleep in the inpatient setting have been proposed, but most are lacking sufficient clinical evidence to support widespread use. Relatively simple, low-risk interventions that do not require intensive investment in new resources include the following (see 'Interventions to improve sleep' above):
•Evaluate for and treat underlying sleep disorders
•Address diseases and symptoms interfering with sleep
•Reduce noise at night
•Establish a clear day/night light exposure protocol consistent with environmental norms
•Use simple relaxation techniques (eg, music, relaxation tapes)
•Limit patient interactions during typical sleep hours to those that are truly required for patient care
•Review medication lists daily and eliminate or choose alternative medications in place of those that may interfere with sleep quality
●Role of pharmacologic sleep aids – Pharmacologic sleep aids should be used judiciously in the inpatient setting. When pharmacologic therapy is deemed necessary, selection of an agent should be individualized based on severity of symptoms, age and comorbidities, side effects, and drug-drug interactions.
We suggest melatonin as a first-line agent in most inpatients, based on its tolerability, low potential for drug-drug interactions, and potential to improve circadian rhythms (Grade 2C). However, the sleep-inducing effects of melatonin are variable and may not be as robust as other sleep aids. (See 'Pharmacotherapy' above.)
50 : Partial night sleep deprivation reduces natural killer and cellular immune responses in humans.
76 : Propofol versus flunitrazepam for inducing and maintaining sleep in postoperative ICU patients.