Background

Lack of sleep affects a person’s physical and mental health and, for people who are critically ill, sleep is thought to improve healing and survival. People in the intensive care unit (ICU) experience poor sleep. Many factors contribute to poor sleep including high noise levels, 24-hour lighting and intrusive patient care activities. Propofol is an anaesthetic agent given by infusion into a vein that is sometimes used to sedate people who are in the ICU. In this review, we looked at studies in which propofol was given to adults at night-time to improve the quality and quantity of sleep.

Study characteristics

The evidence is current to October 2017. We included four randomized controlled studies (clinical studies where people are randomly put into one of two or more treatment groups) with 149 participants in the review. Two studies are awaiting classification (because we could not assess their eligibility) and one study is ongoing. All participants were critically ill and were in the ICU.

Key results

We did not combine the results from the studies because of differences in comparison (called control) treatments and study design. One study compared propofol with no agent. This study used polysomnography (which records brain waves, oxygen level in blood, heart rate, breathing, and eye and leg movements) to measure sleep quality and quantity. It reported no improvement in duration of sleep with propofol but participants woke up less often and for shorter lengths of time and described their sleep quality as being improved with propofol. One study compared a higher dose of propofol at night described as additional night sedation, with a constant day-time and night-time dose. This study used the Ramsay Sedation Scale (which is normally used by anaesthetists to assess how easily a person is roused) and reported that participants appeared to have an improved sleep rhythm. Two studies compared propofol with benzodiazepines (a tranquilizing medicine; flunitrazepam in one study and midazolam in one study). These studies used the Pittsburgh Sleep Diary and the Hospital Anxiety and Depression Scale to measure quantity and quality of sleep. The study with flunitrazepam reported fewer awakenings of reduced duration with propofol but similar total sleep time in each group and the study with midazolam reported no difference in sleep quality. The study with flunitrazepam also measured sleep with Bispectral Index (used by anaesthetists to assess depth of anaesthesia) and reported longer time in deep sleep, with fewer awakenings. The study with midazolam reported higher levels of anxiety and depression in both groups, and no difference when participants were given propofol. No study reported on side effects.

Quality of evidence

We judged the evidence to be very low quality. We found only four small randomized controlled studies and the results of the studies were not consistent. We noted differences in illness severity of participants and the medicines that were compared with propofol in the included studies. Measuring quality of sleep using diaries, questionnaires and scoring systems is based on, or is influenced by, personal feelings or opinions, and we were concerned that staff and participants were aware which medicine they had been given; we believed that this could have influenced the results. Only one study used polysomnography, which is the most appropriate unbiased measurement tool for sleep.

Conclusions

We were unable to collect sufficient evidence to determine whether propofol given at night to adults in the ICU improves quality and quantity of their sleep, as a way of helping recovery.