Summary of study findings, by outcome
| Author, year | Treatment group | Method of sleep evaluation | Outcome by treatment group | Measure of effect |
|---|---|---|---|---|
| Mechanically ventilated patients in an ICU | ||||
| Alexopoulou et al,32 2014 | Dexmedetomidine iv infusion (0.6 (0.4–0.7) µg/kg/hour)* (n=16 enrolled, 13 analysed) | Polysomnography (no treatment on day 1, dexmedetomidine on day 2) | Sleep efficiency* 77.9% (65.6–80.2%) Sleep stage distribution(%)* Stage I=16.1 (6.2–21.3) Stage II=78.7 (69.2–92.5) Stage III=0.0 (0–0) REM=0.0 (0–0.4) Sleep fragmentations* 2.2 (1.6–4.5) arousals/hour | Sleep efficiency p=0.002 between first and second nights; p=0.001 between second and third nights Sleep fragmentation p=0.0023 Sleep stagedistribution Stage I: p=0.006 Stage II: p=0.006 Stage III: p=0.180 REM: p=0.173 |
| No treatment crossover (n=16 enrolled, 10 analysed) | Sleep efficiency* 15.8% (6.4–51.6%) Sleep stage distribution (%)* Stage I=56.2 (24.7–79.3) Stage II=39.2 (20.7–66.4) Stage III=0.0 (0–0) REM=0.0 (0–0.4) Sleep fragmentation* 7.1 (6.1–13.4) arousals/hour | |||
| Kim et al,39 2014 | Midazolam iv 0.02 (0.015–0.04) mg/kg/hour* (n=9 enrolled, 5 analysed) | Polysomnography | Sleep efficiency* 34.3% (8–59.7) Sleep stage distribution (min)* Stage I=82.0 (60.5–372.5) Stage II=88.0 (19.0–621.0) Stage III=0.0 (0–0) REM=10.0 (6.0–50.5) Sleep fragmentation* 16.1 (7.6–28.6) arousals/hour | Correlation between dose of midazolam and total sleep time (r=0.975, p=0.005) Correlation between midazolam dose and stage II (r=0.975, p=0.005) |
| Oto et al,40 2012 | Dexmedetomidine iv infusion 0.2–0.7 mcg/kg/hour (n=10 enrolled, 10 analysed) | Polysomnography | Sleep efficiency* 52.3% (47–89.7) Sleep fragmentation* 9.3 (3–19.5) arousals/hour Sleep stage distribution (min)* Stage I=76 (32–145) Stage II=188 (136–449) Stage III=0 (0.0–1.3) REM=0.0 (0–0) | None reported |
| Kondili et al,29 2012 | Propofol infusion 0.86 (0.67–1.25) mg/kg/hour (n=7 randomised, 6 analysed) | Polysomnography | Sleep efficiency* 76.3% (28.4–96.9) Sleep fragmentation* 4.8 (1.3–14.6) arousals/hour Sleep stage distribution* Stage I=20.8% (5.6–80.6) Stage II=48.9% (4.8–84) Stage III/IV=0% (0–5.8) REM=0% (0–0) | Sleep efficiency p=0.37 Sleep fragmentation p=0.33 Sleep stagedistribution Stage I: p=1.0 Stage II: p=0.66 Stage III/IV: p=0.75 REM: p=0.04 |
| No propofol crossover (n=6 randomised, 6 analysed) | Sleep efficiency* 62.6% (13.1–85.9) Sleep fragmentation* 8.1 (2.9–16.2) arousals/hour Sleep stage distribution* Stage I=30.7% (4.6–66.7) Stage II=46.1% (3.0–80.4) Stage III/IV=0% (0–00) REM=1.4% (0–13) | |||
| Oto et al,31 2011 | Midazolam infusion 0.6 mg/kg (0.5–0.9)* with daytime sedation interruption (n=11 randomised, 11 analysed) | Polysomnography | Sleep efficiency* 81% (73.3–91.1%) Sleep fragmentation* 4.4 (2.5–8.3) arousals/hour Sleep stage distribution (min)* Stage I =96 (60–144) Stage II=264 (222–360) Stage III/IV=1.8 (0.6–4.2) REM=30 (12–48) | Sleep efficiency p=0.047 Sleep fragmentation p=0.03 Sleep stage distribution Stage I: p=0.46 Stage II: p=0.55 Stage III/IV: p=0.03 REM: p=0.01 |
| Midazolam infusion 2.7 mg/kg (2.0–3.4)* continuous infusion (n=11 randomised, 11 analysed) | Sleep efficiency* 97% (91.1–97.8%) Sleep fragmentation* 2.2 (0–2.6) arousals/hour Sleep stage distribution (min)* Stage I =180 (54–342) Stage II=282 (66–468) Stage III/IV=0 (0–0) REM=0 (0–9) | |||
| Bourne et al,27 2008 | Melatonin 10 mg orally (n=13 randomised, 12 analysed) | BIS Actigraphy Nurse observation Patient self-report questionnaire | Sleep efficiency† BIS: 39% (27–51%) Actigraphy: 73% (53–93%) Nurse observation: 45% (26–64%) Patient self-report: 41% (24–59%) | Sleep efficiency BIS: p=0.09 Actigraphy: p=0.84 Nurse observation: p=0.58 Patient self-report: p=0.32 |
| Placebo (n=12 randomised, 12 analysed) | Sleep efficiency† BIS: 26% (17–36%) Actigraphy: 75% (67–83%) Nurse observation: 51% (35–68) Patient self-report: 50% (43–58%) | |||
| Ibrahim et al,28 2006 | Melatonin 3 mg orally (n=14 randomised, 14 analysed) | Observation by nurse | Sleep efficiency‡ 50% (16–69) | Sleep efficiency p=0.98 |
| Placebo orally (n=18 randomised, 18 analysed) | Sleep efficiency‡ 50% (0–72%) | |||
| Hardin et al,33 2006 | Intermittent lorazepam iv: 0.04±0.04 mg/kg/day§ (n=6 randomised, 6 analysed) | Polysomnography | Sleep efficiency§ 42±26.7% Sleep fragmentation§ 4.24±2.4 arousals/hour Sleep stage distribution (%)§ Stage I=5.8±6.6 Stage II=57.7±23.8 Stage III/IV=31.9±24.6 REM=3.6±5.7 | Sleep efficiency p=0.07 Sleep fragmentation p=0.90 Sleep stagedistribution Stage I: p=0.40 Stage II: p=0.20 Stage III/IV: p=0.93 REM: not reported |
| Continuous lorazepam iv 0.72±0.39 mg/kg/day§ (n=6 randomised, 6 analysed) | Sleep efficiency§ 58.8±26.3% Sleep fragmentation§ 4.79±3.7 arousals/hour Sleep stage distribution (%)§ Stage I=20.8±35.5 Stage II=28.34±35 Stage III/IV=49.6±49 REM=NA | |||
| Continuous lorazepam iv with neuromuscular blockade 0.62±0.20 mg/kg/day§ (n=6 randomised, 6 analysed) | Sleep efficiency§ 78.8±19.6% Sleep fragmentation§ 4.4±3.7 arousals/hour Sleep stage distribution (%)§ Stage I=7.4±5.9 Stage II=48.8±22.3 Stage III/IV=38.9±28.1 REM=NA | |||
| Hospitalised patients not in an ICU | ||||
| Engelmann et al,30 2014 | Propofol 2 mg/kg/hour (n=34 randomised, 34 analysed) | BIS (analysis only in 31 propofol patients and 25 flunitrazepam patients) Sleep diary (PgShD) used to evaluate sleep quality, sleep efficiency and sleep fragmentation Sleep quality recorded on 5-point scale | Sleep efficiency‡ BIS: 94.7% (30.7–100) Sleep diary (median): 86% Sleep fragmentation BIS (median): 0.6 arousals/hour Sleep stage distribution‡ BIS values=74.05 (65.68–79.38) Stage A=7.4% (1.1–98) Stage B=36.6% (5.4–82.3) Stage C=5.0% (0–50.7) Sleep quality Sleep diary (median)=2/5 | Sleep efficiency BIS: p=0.777 Sleep diary: p=0.623 Sleep fragmentation BIS: p=0.041 Sleep stagedistribution BIS values: p=0.016 Stage A: p=0.044 Stage B: p=0.004 Stage C: p=0.69 Sleep quality Sleep diary: p<0.001 |
| Flunitrazepam bolus 0.015 mg/kg (32 randomised, 32 analysed) | Sleep efficiency‡ BIS: 92.7% (16.7–100) Sleep diary (median): 71% Sleep fragmentation BIS (median): 1.1 Sleep stage distribution‡ BIS values=78.7 (72.05–81) Stage A=30% (4.0–81.3) Stage B=55.4% (96–90) Stage C=7.6% (0–83.7) Sleep quality Sleep diary (median): 3/5 | |||
| Li Pi Shan et al,41 2004 | Lorazepam 0.5–1.0 mg (n=10 randomised, 9 analysed) | Nurse observation Patient questionnaire (10-point scale for quality of sleep, alertness in the morning) | Total sleep time§ 469±46.2 min Sleep quality* 8.5 (7.5–10) Conditioning in the morning* Alertness=9 (8–10) Tiredness=7.5 (5–10) | Total sleep time p=0.09 Sleep quality p=0.17 Conditioning in the morning Alertness: p=0.60 Tiredness: p=0.29 |
| Zopiclone 3.75–7.5 mg (n=10 randomised, 9 analysed) | Total sleep time§ 443±37.8 min Sleep quality* 8 (5–9) Conditioning in the morning* Alertness=9 (6.5–10) Tiredness=8 (5.5–8.5) | |||
| Morgan et al,37 1997 | Triazolam 0.25 mg (n=119 randomised, 119 analysed) | Nurse observation (first 2 hours post dose only) Sleep quality: patient self-report 1=excellent 2=good 3=fair 4=poor Conditioning in the morning (patient self-report) Visual analogue scale (0–100 where 0=very sleepy and 100=not at all sleepy) | Sleep latency Median: 30 min <30 min=45% <60 min=76% <90 min=87% <120 min=90% Sleep quality (mean±SD) 2.0±0.1 Conditioning in the morning (mean±SD)=62.8±2.5 | Sleep latency Triazolam and zolpidem groups were significantly better than placebo (p<0.001) but not significantly different between each other. Sleep quality Triazolam and zolpidem groups were significantly better than placebo (p<0.001) but not significantly different between each other. Morning sleepiness No significant difference between groups (p=0.443) |
| Zolpidem 10 mg (n=120 randomised, 120 analysed) | Sleep latency Median: 25 min <30 min=42% <60 min=78% <90 min=92% <120 min=89% Sleep quality (mean±SD) 2.0±0.1 Conditioning in the morning (mean±SD)=58.4±2.7 | |||
| Placebo (n=118 randomised, 118 analysed) | Sleep latency Median: 60 min <30 min=22% <60 min=49% <90 min=64% <120 min=72% Sleep quality (mean±SD) 2.7±0.1 Conditioning in the morning (mean±SD)=59.0±2.9 | |||
| Feldmeier and Kapp,34 1983 | Midazolam 15 mg orally (n=19 randomised, 19 analysed) | Observation by nurse | Total sleep time§: 430 min Sleep latency: 30% asleep within 20 min 84% asleep within 40 min | Total sleep time Reported as no significant difference between groups Sleep latency (<20min) p<0.05 for midazolam and oxazepam vs placebo |
| Oxazepam 15 mg orally (n=15 randomised, 15 analysed) | Total sleep time§: 440 min Sleep latency: 16% asleep within 20 min 73% asleep within 40 min | |||
| Placebo (n=16 randomised, 16 analysed) | Total sleep time: NR Sleep latency: 0% asleep within 20 min 29% asleep within 40 min | |||
| Goetzke et al,38 1983 | Brotizolam 0.25 mg orally (n=86 randomised, 79 analysed but breakdown by group not reported) | Nurse observation Patient questionnaire Frequencies (%) are approximated from a bar graph | Sleep latency§ (<30 min): 55.5 patients (70%) Sleep quality Very good=17% Good=60%, Moderate=18% Poor=5% Conditioning in the morning Refreshed=47% Little tired=44% Very tired=9% | Not reported |
| Triazolam 0.25 mg orally | Sleep latency§ (<30 min): 55.5 patients (70%) Sleep quality Very good=18% Good=55% Moderate=21% Poor=6% Conditioning in the morning Refreshed=49% Little tired=43% Very tired=8% | |||
| Placebo | Sleep latency§ (<30 min): 36.5 patients (46%) Sleep quality Very good=10% Good=32% Moderate=30% Poor=28% Conditioning in the morning Refreshed=31% Little tired=45% Very tired=24% | |||
| Gallais et al,35 1983 | Midazolam 15 mg orally (n=19 randomised, 18 analysed) | Third-party observation (values are approximated from a graph) | Total sleep time <360 min=10% 360–420 min=53% >420 min=37% Sleep latency <30 min=63% 30–50 min=21% >50 min=16% | Total sleep time p<0.05 for both drugs when compared to placebo but no difference between drugs Sleep latency Sleep onset was shorter in the midazolam group when compared with placebo (p<0.05) and oxazepam (p,0.05). There was no statistical difference between oxazepam and placebo groups |
| Oxazepam 50 mg orally (n=20 randomised, 17 analysed) | Total sleep time <360 min=12% 360–420 min=52% >420 min=36% Sleep latency <30 min=17% 30–50 min=44% >50 min=39% | |||
| Placebo (n=20 randomised, 16 analysed) | Total sleep time <360 min=30% 360–420 min=54% >420 min=16% Sleep latency <30 min=10% 30–50 min=50% >50 min=40% | |||
| Lupolover et al,36 1983 | Midazolam 15 mg (40 randomised, 40 analysed) | Nurse observation | Total sleep time 360 min=13.8% 420 min=22.5% 480 min=30.8% 540 min=14.3% 600 min=3.6% Sleep latency <20 min=34.4% 20–40 min=48.7% 40–60 min=12.3% >60 min=4.7% Sleep fragmentation (arousals/night) None=15.9% 1–2=60.4% >2=23.7% | Total sleep time p>0.05 Sleep latency: p>0.05 Sleep fragmentation Not reported |
| Oxazepam 15 mg (38 randomised, 38 analysed) | Total sleep time 360 min=9.9% 420 min=41.6% 480 min=25.9% 540 min=10.9% 600 min=0.6% Sleep latency <20 min=15.2% 20–40 min=42.7% 40–60 min=22% >60 min=20.1% Sleep fragmentation (arousals/night) None=17.5% 1–2=55.4% >2=27.1% | |||
*Data presented as median (IQR).
†Data presented as mean (95% CI).
‡Data presented as median (range).
§Data presented as mean±SD.
ICU, intensive care unit; NR, not reported; PgShD, Pittsburgh Sleep Diary; REM, rapid eye movement.