Мелатонин для улучшения сна у взрослых в отделении интенсивной терапии
Referencias
References to studies included in this review
References to studies excluded from this review
References to ongoing studies
Additional references
References to other published versions of this review
Characteristics of studies
Characteristics of included studies [ordered by study ID]
| Methods | RCT, parallel design, single‐centre study | |
| Participants | Total number of randomized participants: 25 Inclusion criteria: adult ICU patients with acute respiratory failure requiring MV and tracheostomy to assist weaning Exclusion criteria: expected length of stay < 5 days; pre‐admission treatment of sleep disturbances; contraindications to enteral feeding; a history of convulsions, psychiatric or neurological disease; alcohol consumption ≥ 50 units per week or drug use; sleep apnoea; severe heart failure (NYHA classification III/IV); and low levels of consciousness (defined as values < 4 on the SAS) Primary diagnoses: severe sepsis, postoperative respiratory failure, or pneumonia Baseline characteristics Intervention group
Control group
Country: UK Setting: adult general ICU | |
| Interventions | Intervention group
Control group
| |
| Outcomes | Measures of sleep (measured using BIS); sleep efficiency (measured using SEI), defined as ratio of a participant's total sleep time over the time available for 'nocturnal' sleep (9 hours, 10 p.m. to 7 p.m.); and AUC Use of actigraphy, nurse assessment (direct nurse observation using hourly epochs), and patient assessment (measured using RCSQ) | |
| Notes | Funding/declarations of interest: Sheffield Teaching Hospitals Department of Pharmacy and Medicines Management and Small Grants Scheme Study dates: not reported Notes
| |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Quote: "Patients were randomly assigned to melatonin or placebo by the pharmacy, using random assignment in blocks of 4." Study authors reported by email that participants (quote) "were randomly assigned to melatonin or placebo by the pharmacy, using random assignment in blocks of four" |
| Allocation concealment (selection bias) | Unclear risk | No details |
| Blinding of personnel (performance bias) | Low risk | Study authors reported that "only the pharmacy clinical trials staff were unblinded to study allocation to enable them to supply the placebo/melatonin liquid used in the study. The patient, all clinical nursing, medical and pharmacy staff as well as the investigators were all blinded to study allocation" (information supplied by email) |
| Blinding of participants (performance bias) | Low risk | Study authors reported that "only the pharmacy clinical trials staff were unblinded to study allocation to enable them to supply the placebo/melatonin liquid used in the study. The patient, all clinical nursing, medical and pharmacy staff as well as the investigators were all blinded to study allocation" (information supplied by email) |
| Blinding: subjective measures (detection bias) | Low risk | Study authors reported that "only the pharmacy clinical trials staff were unblinded to study allocation to enable them to supply the placebo/melatonin liquid used in the study. The patient, all clinical nursing, medical and pharmacy staff as well as the investigators were all blinded to study allocation" (information supplied by email) |
| Blinding: objective measures (detection bias) | Low risk | Study authors reported that "only the pharmacy clinical trials staff were unblinded to study allocation to enable them to supply the placebo/melatonin liquid used in the study. The patient, all clinical nursing, medical and pharmacy staff as well as the investigators were all blinded to study allocation" (information supplied by email) |
| Incomplete outcome data (attrition bias) | Unclear risk | Study authors report that 12 participants in each group were analysed. However, study authors also report that data were not available for 4 patients (3 in placebo group, 1 in melatonin group) for nights 3 and 4. This represents a high loss of data for a small sample size. We have assumed that ITT principle was used for this loss of participant data, but this is not stated in the paper |
| Selective reporting (reporting bias) | Unclear risk | Retrospective clinical trials registration: ISRCTN47578325. Not feasible to judge risk of selective outcome reporting |
| Other bias | Unclear risk | Study authors noted differences in baseline characteristics; more participants in control group were older and had delirium, and differences in modes of ventilation between groups |
| Methods | RCT, parallel design, single‐centre study Described as a pilot study | |
| Participants | Total number of randomized participants: 12 Inclusion criteria: adult patients admitted to the neurological ICU undergoing continuous EEG (cEEG) monitoring Exclusion criteria: lack of reactivity on initial cEEG, expected mortality ≤ 24 hours, general anaesthesia ≤ 24 hours before study, continuous infusions of benzodiazepines or opiates, history of sleep apnoea, or random SpO₂ < 85% during sleep Primary diagnoses: acute brain injury, cardiac arrest, sepsis Baseline characteristics Intervention group
Control group
Country: USA Setting: neurosurgical ICU | |
| Interventions | Intervention group
Control group
| |
| Outcomes | Sleep scores (measured using PSG); hospital and ICU length of stay; clinical diagnosis of delirium during hospital stay; and discharge modified Rankin score | |
| Notes | Funding/declarations of interest: not reported Study dates: June 2011 to June 2012 | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Block randomized using computer‐generated random numbers |
| Allocation concealment (selection bias) | Unclear risk | No details |
| Blinding of personnel (performance bias) | High risk | Not feasible to blind participants and personnel to intervention |
| Blinding of participants (performance bias) | High risk | Not feasible to blind participants and personnel to intervention |
| Blinding: objective measures (detection bias) | Unclear risk | No details |
| Incomplete outcome data (attrition bias) | High risk | Large loss of participants relative to very small sample size. Data for primary outcome available for only 1 participant in each group |
| Selective reporting (reporting bias) | Unclear risk | Clinical trial registration not reported. Not feasible to assess risk of selective outcome reporting. We noted that length of ICU stay was collected but not reported. |
| Other bias | Unclear risk | Very small study, with limited data |
| Methods | RCT, parallel design, single‐centre study Described as a pilot study | |
| Participants | Total number of randomized participants: 32 Inclusion criteria: tracheostomy in situ, weaning from MV, GCS > 9, and sedative infusions or boluses stopped for > 12 hours Exclusion criteria: patient < 16 years of age, is pregnant or breastfeeding, has known allergy to melatonin, intestinal obstruction, ileus, gastroparesis or other conditions likely to affect enteral absorption of melatonin, or likelihood that the patient would die ≤ 24 hours Primary diagnoses: not reported Baseline characteristics Intervention group
Control group
Country: Australia Setting: ICU | |
| Interventions | Intervention group
Control given
| |
| Outcomes | Duration of observed nocturnal sleep (total number of hours between 10 p.m. and 6 a.m.; criteria were: eyes closed, decreased motor activity, lack of interaction with the environment, and lack of purposeful activity) compared with diurnal sleep; degree of agitation; need for extra sedation or haloperidol treatment; vital status on discharge from hospital; and time to ICU and hospital discharge | |
| Notes | Funding/declarations of interest: not reported Study dates: August 2003 to February 2005 | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Computer‐generated randomization |
| Allocation concealment (selection bias) | Low risk | Quote: "Patient allocation was known only to the hospital research pharmacist who dispensed the treatment." |
| Blinding of personnel (performance bias) | Low risk | Quote: "Patients, ICU staff and investigators were blinded to treatment allocation." |
| Blinding of participants (performance bias) | Low risk | Quote: "Patients, ICU staff and investigators were blinded to treatment allocation." |
| Blinding: subjective measures (detection bias) | Low risk | Nurses who assessed sleep through observation were blinded |
| Incomplete outcome data (attrition bias) | Unclear risk | Loss of some patient analysis (5 patients), reasons were reported but not clear whether losses were balanced between groups |
| Selective reporting (reporting bias) | Unclear risk | Clinical trial registration not reported. Some outcomes from Methods section are not reported in the Results (time to discharge from the ICU) |
| Other bias | Low risk | Baseline characteristics appear comparable. No other sources of bias identified |
| Methods | RCT, parallel design, single‐centre study | |
| Participants | Total number of randomized participants: 82 Inclusion criteria: high‐risk patients: acute respiratory failure requiring invasive or non‐invasive respiratory assistance with an expected length of ICU stay > 2 days, mortality predicted at ICU admission over 13% (using SAPS II) Exclusion criteria: aged < 18 years, gastro‐intestinal tract impracticability, status asthmaticus, intoxication, renal replacement therapy for chronic renal failure, child C hepatopathy, HIV, home mechanical ventilation, expected GCS at discharge < 12 points, neuropsychiatric disability, pregnancy/breast feeding, too ill (not defined by study authors) Primary diagnoses: pneumonia, pancreatic diseases, gastrointestinal diseases, cardiorespiratory arrest, acute myocardial infarction Baseline characteristics Intervention group
Control group
Country: Italy Setting: medical and surgical ICU | |
| Interventions | Intervention group
Control group
| |
| Outcomes | Use of sedative agents (hydroxyzine), overall amount of neuroactive drugs, RASS levels, sleep hours observed by nurses, duration of agitation, anxiety, pain, use of restraints, adequacy of sedative therapy, PTSD, time to wean from neuroactive drugs and from MV, costs of sedatives, ICU length of stay, ICU mortality, and hospital mortality | |
| Notes | Funding/declarations of interest: no external funding Study dates: July 2007 to December 2009 Notes:
| |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Computer‐generated 8‐patient block randomization procedure |
| Allocation concealment (selection bias) | Low risk | Quote: "each patient received a sealed brown envelope containing the random melatonin or placebo treatment: it was then opened just before evening of the third ICU day." |
| Blinding of personnel (performance bias) | Low risk | Use of placebo with identical appearance to melatonin. Physicians and nurses were unaware of treatment groups |
| Blinding of participants (performance bias) | Low risk | Participants were blinded to treatment groups |
| Blinding: subjective measures (detection bias) | Low risk | Nurses who assessed sleep time were blinded to treatment groups |
| Blinding: objective measures (detection bias) | Low risk | Physicians who assessed adverse events and made decisions to discontinue treatment were blinded to treatment groups. Assessment of mortality data unlikely to be influenced by blinding |
| Incomplete outcome data (attrition bias) | Low risk | Some participants did not complete study until final endpoint, but these were few. Reported by study authors and use of ITT analysis |
| Selective reporting (reporting bias) | Low risk | Prospective clinical trials registration NCT00470821. Some differences in focus of primary outcome (but not relevant for this review). Protocol had planned to use wrist actigraphy for sleep measures. However, decision not to use objective measures is reported in supplementary appendix; wrist actigraphy did not appear to provide consistent measures and was therefore abandoned because of lack of reliability |
| Other bias | Unclear risk | Study authors note differences between groups in pre‐treatment with opiates ‐ more participants in the control group received opiates and had been sedated with enteral hydroxyzine and lorazepam |
APACHE II: Acute Physiology and Chronic Health Evaluation II
AUC: area under the curve
BIS: bispectral index
BiPAP: biphasic positive airway pressure
cEEG: continuous electroencephalogram
CI: confidence interval
CPAP: continuous positive airway pressure
CPAP‐ASB: continuous positive airway pressure with assisted spontaneous breathing
GCS: Glasgow Coma Score
HIV: human immunodeficiency virus
ICU: intensive care unit
IQR: interquartile range
ITT: intention to treat
M/F: male/female
MV: mechanical ventilation
NYHA: New York Heart Association
PSG: polysomnography
PTSD: post‐traumatic stress disorder
RCT: randomized control trial
RCSQ: Richards‐Campbell Sleep Questionnaire
SAPS II: Simplified Acute Physiology Score II
SAS: Sedation Agitation Scale
SD: standard deviation
SEI: Sleep Efficiency Index
SOFA: Sequential Organ Failure Assessment
SpO2: oxygen saturation
Y: yes
Characteristics of excluded studies [ordered by study ID]
| Study | Reason for exclusion |
| RCT. Participants were critically ill patients in the ICU who were randomized to receive melatonin. This is a pharmacokinetics study that administered enteral melatonin in order to simulate endogenous release. Melatonin was not given with the specific aim of promoting sleep and study authors do not measure or report any sleep‐related outcomes. | |
| Wrong study design. Commentary on an observational study of sleep‐wake regulation in mechanically‐ventilated patients receiving sedation | |
| RCT. Study authors compared melatonin with ear plugs and eye masks, and with placebo. Participants were healthy volunteers and slept in a simulated ICU environment with noise and light | |
| Non‐randomized cohort study assessing pharmacokinetics after melatonin administration in critically ill ICU patients | |
| Wrong study design. Commentary on Mistraletti 2015 | |
| Wrong study design. Literature review on pharmacological agents for sleep promotion in the ICU | |
| Non‐randomized study assessing melatonin use in ICU patient with COPD. Participants in the control group were not in the ICU |
COPD: chronic obstructive pulmonary disease
ICU: intensive care unit
RCT: randomized control trial
Characteristics of ongoing studies [ordered by study ID]
| Trial name or title | Exogenous melatonin to improve sleep in critically ill patients |
| Methods | RCT |
| Participants | Target number of randomized participants: 60 Inclusion criteria: patients ≥ 18 years of age admitted to ICU, requiring tracheostomy, GCS ≥ 12, with difficulty sleeping Exclusion criteria: patients on sedative infusions, with contraindications to enteral feeding or administration of medications, pregnant, or breastfeeding |
| Interventions | Intervention group
Control group
|
| Outcomes | Quantity of nocturnal sleep, quantity of diurnal sleep, number of awakenings, episodes of agitation, length of stay |
| Starting date | 22 June 2010 |
| Contact information | |
| Notes | Funding/declarations of interest: pharmacy department Notes
|
| Trial name or title | Impact of oral melatonin on critically ill adult patients with ICU sleep deprivation: study protocol for a randomized controlled trial |
| Methods | RCT |
| Participants | Target number of randomized participants: 198 Inclusion criteria: patients ≥ 18 years of age; with GCS score ≥ 10; sedation with propofol, morphine, alfentanil, and dexmedetomidine discontinued ≥ 36 hours (48 hours for lorazepam and midazolam); expected mechanical ventilation ≥ 5 days; and clinically and biologically stable Exclusion criteria: patients pregnant or breastfeeding; preadmission treatment of sleep disturbances; history of convulsions, psychiatric or neurological disease, sleep apnoea, deafness or blindness; alcohol consumption ≥ 50 units per week or drug use; liver insufficiency; renal insufficiency; severe heart failure; intestinal obstruction, ileus, gastroparesis or other conditions likely to affect enteral absorption of melatonin; use of drugs that might alter melatonin secretion; known allergy to melatonin; readmitted to ICU after randomization to study; enrolled in another trial Country: China Setting: comprehensive ICU, Fuxing Hospital, Capital Medical University, Beijing |
| Interventions | Intervention group
Control group
|
| Outcomes | Sleep cycle and sleep quality: night and day sleep time, total sleep time, percentages of NREM stage 1/2, slow wave sleep and REM, incidence of arousals per hour, duration of sleeps without waking, and number of sleep periods Objective sleep quality (measured using RCSQ), anxiety (measured using VAS‐A), stress levels, oxidative stress, inflammation, delirium‐free days, ventilator‐free days, antibiotic‐free days, length of ICU stay, mortality, melatonin side effects |
| Starting date | February 2014 |
| Contact information | |
| Notes | Funding/declarations of interest: supported by the State Science and Technology Support Program. The sponsors have no role in the study design, data collection, data analysis, data interpretation, or writing of the report |
| Trial name or title | The effect of melatonin tablet on sleep disorders in patients in cardiac surgery intensive care unit |
| Methods | RCT |
| Participants | Number of randomized participants: 40 Inclusion criteria: patients > 18 years of age, GCS ≥ 10, stable haemodynamic state, not dependent on ventilator Exclusion criteria: pregnancy, sleep disorders, convulsions, neuropsychiatric disorders, blindness and deafness, liver and kidney disorders, CHF, gastroparesis, taking drugs that reduce melatonin effect, sensitivity to melatonin, worsening of overall health, earlier discharge from the 6th day |
| Interventions | Intervention group
Control group
|
| Outcomes | Sleep disorders (measured on 1st and 4th night using SSSQ), serum leptin levels, serum melatonin levels, serum cortisol levels |
| Starting date | 20 April 2016 |
| Contact information | |
| Notes | Funding/declarations of interest: Zanjan University of Medical Sciences |
| Trial name or title | Prophylactic melatonin for delirium in intensive care (Pro‐MEDIC): study protocol for a randomized controlled trial |
| Methods | RCT |
| Participants | Target number of randomized participants: 850 Inclusion criteria: patients expected ICU length of stay of ≥ 72 hours Exclusion criteria: patients < 18 years of age, already receiving melatonin therapy before ICU admission, prior hypersensitivity to study drug components, expected discharge within 72 hours of admission, expected death within 48 hours of enrolment, pregnancy or breastfeeding, non‐English speaking, condition not expected to improve sufficiently to use CAM‐ICU, neurological problems that would affect CAM‐ICU assessment, no enteral route/nil by mouth, hepatic impairment Country: Australia Setting: ICUs, multicentre |
| Interventions | Intervention group
Control group
|
| Outcomes | Prevalence of delirium, duration of delirium, severity of delirium, sleep quality, length of ICU and hospital stay, morbidity and mortality, hospital costs |
| Starting date | 4 June 2016 |
| Contact information | |
| Notes | Funding/declarations of interest: State Health Research Advisory Council Grant of Western Australia, and from John Hunter Hospital Charitable Trust Research Grant Notes
|
| Trial name or title | Melatonin for prevention of delirium in critically ill patients (MELLOW‐1) |
| Methods | RCT |
| Participants | Target number of randomized participants: 69 Inclusion criteria: critically ill patients ≥ 18 years of age, anticipated ICU stay of > 48 hours, able to receive enteral administration of study drug (i.e. by mouth or naso‐ or oro‐ or percutaneous gastric or post‐pyloric feeding tube), consent to participate Exclusion criteria: ICU admission of > 48 hours prior to screening; unable to assess for delirium (e.g. comatose defined as SAS 1 or 2 or either 'No Response' score A or B on ICDSC, chemically paralysed with neuromuscular blocking drugs); screened delirium positive prior to randomization (ICDSC score ≥ 4 out of 8); anticipated withdrawal in next 48 hours; history of severe cognitive or neurodegenerative disease (e.g. dementia, Parkinson's disease) or severe structural brain injury (e.g. traumatic brain injury, intracranial haemorrhage); unable to communicate in English or French; contraindications to receiving any enteral medication (defined as absolute contraindication to enteral nutrition such as gastrointestinal obstruction, perforation, recent upper GI surgery, no enteral access); active seizures; pregnancy; blindness; known allergy to melatonin Country: Canada Setting: ICUs, multicentre |
| Interventions | Intervention group 1
Intervention group 2
Control group
|
| Outcomes | Study adherence, ICU mortality, length of ICU stay, duration of mechanical ventilation, sleep (using RCSQ), incidence and duration of delirium, adverse events, feasibility (time in motion, and enrolment rate) pharmacokinetic outcomes, hospital mortality, and incidence of subsyndromal delirium |
| Starting date | April 2016 |
| Contact information | |
| Notes | Funding/declarations of interest: Mount Sinai Hospital |
CAM‐ICU: Confusion Assessment Method in the intensive care unit
CHF: congestive heart failure
GCS: Glasgow Coma Scale
GI: gastrointestinal
ICDSC: Intensive Care Delirium Screening Checklist
ICU: intensive care unit
NREM: non‐rapid eye movement sleep
RCSQ: Richards‐Campbell Sleep Questionnaire
RCT: randomized control trial
REM: rapid eye movement sleep
SAS: sedation agitation scale
SSSQ: Stanford Sleep Scale Questionnaire
VAS‐A: Visual Analogue Scale‐Anxiety
Study flow diagram
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study. Blank spaces in figure indicate that outcomes were not reported by study authors, and therefore risk of bias was not completed for these domains.
| Melatonin compared with no agent for the promotion of sleep in adult patients in the ICU | ||||
| Patient or population: adult patients in the ICU Settings: ICUs, in Australia, Italy, UK, and US Intervention: melatonin Comparison: no agent | ||||
| Outcomes | Impacts | No of participants | Quality of the evidence | Comments |
| Quantity and quality of sleep as measured through reports of participants or of family members or by personnel assessments Data collected at end of follow‐up | In 1 study, participants completed the RCSQ and study authors reported no difference in SEI scores between groups. This was consistent with nurse assessment for which study authors also reported no difference in SEI scores between groups 2 studies reported no difference in duration of sleep observed by nurses | 139 (3 studies) | ⊕⊝⊝⊝ | We did not conduct meta‐analysis because studies used different methods to report data |
| Quantity and quality of sleep as measured by PSG, actigraphy, BIS, or EEG Data collected at end of follow‐up | In 1 study, investigators used BIS and actigraphy to record sleep. Study authors reported no difference in SEI scores with both tools. Study authors also reported longer sleep in participants given melatonin which was not statistically significantly different, and also reported evidence of improved sleep in participants given melatonin from analysis of AUC using BIS data 1 study used PSG, with a large loss of participant data at follow‐up, which prevented analysis of sleep data | 37 (2 studies) | ⊕⊝⊝⊝ | We did not conduct meta‐analysis because studies differed in types of measurement tools |
| Anxiety or depression, or both Data collected at end of follow‐up | 1 study (using VNR ≥ 3) reported no evidence of a difference in anxiety scores between groups | 82 (1 study) | ⊕⊝⊝⊝ | We identified only one study and could not conduct meta‐analysis |
| Mortality Study authors did not report final timepoint for data collection. | 1 study reported one‐third of participants had died; number of deaths per group was not reported 1 study reported no evidence of a difference between groups in hospital mortality | 94 (2 studies) | ⊕⊝⊝⊝ | We did not conduct meta‐analysis because studies different in methods of reporting data |
| Length of stay in the ICU | One study reported no evidence of a difference in length of ICU stay between groups | 82 (1 study) | ⊕⊝⊝⊝ | We identified only 1 study and could not conduct meta‐analysis |
| Adverse events (such as nausea, dizziness and headache) Data collected at end of follow‐up | 1 study reported headache in one participant who was given melatonin 1 study reported a cutaneous rash in one participant in the control group, and 2 participants (1 participant in both groups) with excessive sleepiness | 107 (2 studies) | ⊕⊝⊝⊝ | We could not conduct meta‐analysis because studies different in reported types of adverse events |
| Acronyms and abbreviations AUC: area under the curve; BIS: bispectral index; EEG: electroencephalogram; ICU: intensive care unit; PSG: polysomnography; RCSQ: Richards‐Campbell Sleep Questionnaire; SEI: sleep efficiency index; VNR: verbal numeric range | ||||
| GRADE Working Group Grades of Evidence | ||||
| aWe downgraded by 2 levels for imprecision; there were few studies with few participants, and outcome measures included both personnel and participant reports. We downgraded by 1 level for study limitations; we noted differences in baseline characteristics in two studies which may have influenced results. We downgraded by 1 level for inconsistency; we noted differences in doses of melatonin given in each study. We downgraded by 1 level for indirectness; we could not be certain that ICU sedation protocols in one study were generalizable to most ICUs bWe downgraded by 2 levels for imprecision; there were few studies very few participants, and outcome measures used different assessment tools which may not effectively measure sleep in the ICU patient. We downgraded by 1 level for study limitations; we noted differences in baselines characteristics in one study which may have influenced results. We downgraded 1 level for inconsistency; we noted differences in doses of melatonin given in each study. We downgraded by 1 level for study limitations; we noted high risk of performance bias and high attrition which prevented adequate outcome reporting cWe downgraded by 2 levels for imprecision; data was from one study with few participants, and we could not be certain whether this outcome was measured with an appropriate tool for patients in the ICU. We downgraded by 1 level for study limitations; pre‐treatment use of opiates differed between study groups. We downgraded by 1 level for indirectness; we could not be certain that ICU sedation protocols in 1 study were generalizable to most ICUs dWe downgraded by 2 levels for imprecision; there were few studies with few participants. We downgraded by 1 level for study limitations; we noted a high risk of performance bias and attrition bias in one study and we noted differences between groups in pre‐treatment use of opiates. We downgraded 1 level for inconsistency; doses of melatonin varied between studies. We downgraded by 1 level for indirectness; we could not be certain that ICU sedation protocols in 1 study were generalizable to most ICUs eWe downgraded by 2 levels for imprecision; data was from one study. We downgraded by 1 level for study limitations; pre‐treatment use of opiates differed between study groups. We downgraded by 1 level for indirectness; we could not be certain that ICU sedation protocols in 1 study were generalizable to most ICUs fWe downgraded by 2 levels for imprecision; there were few studies with few participants. We downgraded by 1 level for study limitations; we noted differences in baseline characteristics in two studies which may have influenced results. We downgraded by 1 level for indirectness; we could not be certain that ICU sedation protocols in 1 study were generalizable to most ICUs | ||||
| Outcome: quantity and quality of sleep as measured through reports of participants or family members or by personnel assessments | |||||
| Study | Measurement (tool) | Data* Intervention | Data* Control | Mean difference (95% CI)* | P value* |
| SEI, patient assessment (RCSQ) | mean (95% CI): 0.41 (0.24 to 0.59); n: 12 | mean (95% CI): 0.50 (0.43 to 0.58); n: 12 | −0.09 (−0.28 to 0.09) | 0.32 | |
| SEI, nurse assessment (observations) | mean (95% CI): 0.45 (0.26 to 0.64); n: 12 | mean (95% CI): 0.51 (0.35 to 0.68); n: 12 | −0.06 (CI −0.29 to 0.17) | 0.58 | |
| Duration of sleep, nurse assessment (observations) | median (range): 240 minutes (75 to 331.3); n: 14 | median (range): 243.4 minutes (0 to 344.1); n: 18 | not reported | 0.98 | |
| Duration of sleep, nurse assessment (observations) | 9 p.m. to midnight, mean (SD): 1.5 (± 1.6) hours; n: 41 midnight to 7 a.m., mean (SD): 4.5 (± 1.9) hours; n: 41 | 9 p.m. to midnight, mean (SD): 1.4 (± 1.3) hours; n: 41 midnight to 7 a.m., mean (SD): 4.3 (± 1.8) hours; n: 41 | not reported | 0.92 0.83 | |
| Outcome: quantity and quality of sleep as measured by PSG, actigraphy, BIS, or EEG | |||||
| Study | Measurement (tool) | Data* Intervention | Data* Control | Mean difference (95% CI)* | P value* |
| SEI (BIS) | Mean (95% CI): 0.39 (0.27 to 0.51); n: 12 | Mean (95% CI): 0.26 (0.17 to 0.36); n: 12 | 0.12 (CI −0.02 to 0.27) | 0.09 | |
| SEI (actigraphy) | Mean (95% CI): 0.73 (0.53 to 0.93); n: 12 | Mean (95% CI): 0.75 (0.67 to 0.83); n: 12 | −0.02 (CI −0.24 to 0.20) | 0.84 | |
| Quantity of nocturnal sleep (BIS) | Mean: 3.5 hours; n: 12 | Mean: 2.5 hours; n: 12 | Not reported | ‐ | |
| Outcome: anxiety or depression, or both | |||||
| Study | Measurement (tool) | Data* Intervention | Data* Control | Mean difference (95% CI)* | P value* |
| Anxiety (using VNR) | Participants with score ≥ 3: 12; n: 41 | Participants with score ≥ 3: 14; n: 41 | Not reported | 0.10 | |
| Outcome: mortality at 30 days | |||||
| Study | Measurement | Data* Intervention | Data* Control | Mean difference (95% CI)* | P value* |
| Mortality in hospital | 14; n: 41 | 15; n: 41 | Not reported | 0.82 | |
| Outcome: length of stay in the ICU | |||||
| Study | Measurement | Data* Intervention | Data* Control | Mean difference (95% CI)* | P value* |
| ´Number of days | Median (IQR): 14 (8 to 20); n: 41 | median (IQR): 12 (9 to 29); n: 41 | Not reported | 0.75 | |
| * as reported by study authors BIS: bispectral index | |||||
