Testing the Effectiveness of an Evening Blue-depleted Light Environment in an Acute Psychiatric Ward

In recent decades there has been increased attention to the impact of disturbed sleep on general health. For example, sleep-wake cycle abnormalities linked with circadian dysrhythmias are associated with physical disorders such as diabetes, metabolic abnormalities, obesity, impaired functioning of the immune system, and a greater risk of cancer. Given that light is a central zeitgeber of the circadian system, this had led to some researchers exploring the benefits of photo- or chrono-therapies for selected medical illnesses, especially in those individuals who have a concurrent comorbid mental disorder. The latter is noteworthy as sleep problems are uniquely important in the field of mental health. For instance, sleep abnormalities may be prodromal symptoms that precede the onset of a first episode of a major mental disorder, sleep-wake cycle disruptions are criterion symptoms of unipolar and bipolar disorders and circadian dysrhythmias may exacerbate suicidal behaviours. In addition, research demonstrates that day-to-day variability in the sleep-wake cycle is associated with longer duration of admission to an acute psychiatric inpatient unit and frequency of aggressive behavior and violent incidents. Lastly, and importantly, sleep problems are often the last clinical symptoms to resolve during recovery from an acute episode of a mood or psychotic disorder. Overall, experimental and clinical research emphasize the reciprocal relationship between sleep-wake disruptions and mental disorders showing that they perpetuate and exacerbate each other and that improved sleep is associated with improvements in mental state.

The observations noted above have increased awareness of the need to stabilize sleep-wake cycles in individuals with major mental disorders and highlighted the importance of incorporating therapeutic interventions targeted at circadian dysrhythmias as an adjunct to other treatments offered for acute illness episodes. Psychological and pharmacological interventions are efficacious approaches for sleep-wake cycle disturbances in adults without comorbid mental disorders. However, their use in individuals with an acute exacerbation of a major mental illness can be problematic, including attenuation of the benefit-to-risk ratio for therapies or contra-indications to the use of some medications. Partly as a response to these concerns, but also because of new research on circadian rhythms, attention has shifted to the potential role of chronotherapeutic interventions based on controlled exposures to environmental stimuli that act on biological rhythms. These strategies initially focused on e.g. bright light therapy for seasonal affective disorders and for some sub-types of depression; but more recently they have been extended to e.g. dark therapy (spending 14 hours in darkness per day) or the use of blue blocking glasses in for patients who are hospitalized for the treatment of acute mania. To date, clinical trials of all these interventions have been targeted at small and/or homogeneous samples with a specified sleep or mood disorder (e.g. delayed sleep phase syndrome or bipolar depression, etc.) and have required the study participants to adhere to a protocol for the repeated use of equipment at specified times of the day (sitting by a lamp, resting in forced darkness or wearing glasses, etc.).

The above represent interesting treatment advances. However, given the prevalence of sleep-wake cycle disturbances in individuals with mental disorders, it is logical to extend trials of the use of these types of interventions to broader trans-diagnostic populations. Furthermore, to enhance the generalizability of interventions, it would be helpful to avoid giving personal responsibility for following protocols regarding exposure to different intensities or spectra of light to individuals who are acutely unwell. A pragmatic alternative is to create a therapeutic environment for patients with major mental disorders where changes in light exposure are regulated automatically and where programable lighting conditions form an integral part of a hospital unit. This is an intriguing option as little consideration has been given to how contemporary technology might be employed to augment any benefits of acute treatment in an inpatient facility. Historically, acute psychiatric admission units have offered asylum and a place of safety, and it is assumed that ward routines and structured activities may reduce arousal, regularize sleep-wake cycle patterns and improve confidence and self-esteem, etc. However, the focus is primarily on physical and pharmacological treatments that reduce symptoms and suicidality, enhance social functioning and sufficiently improves the individuals mental state to allow a timely return to outpatient or community care. Less attention has been given to the creation of a state-of-the-art inpatient milieu.

The investigators have been involved in the planning and design of a newly-built psychiatric unit and this process has allowed the investigators to consider how the inpatient environment might be modified to try to enhance recovery from acute illness. The unit comprises of two wards: one ward incorporates state-of-the-art lighting technology while the other ward has an identical layout and facilities but has normal lighting conditions. This unit offers a unique opportunity to explore how exposure to different lighting conditions may modify sleep-wake cycles and how any changes may impact on the clinical and functional outcomes of individuals experiencing an acute episode of a severe mental disorder that requires inpatient care. The findings could influence the future design of hospital units offering care to patients with mental or physical disorders.

The investigators aim to recruit 400 individuals who give written informed consent to participate in a two-arm pragmatic effectiveness randomized controlled clinical treatment trial (RCT). However, based on projected admission rates, the investigators believe that this sample size is at the lower limit of the estimated study population as they are permitted to continue recruitment for at least six consecutive months.

Eligible individuals will be allocated to a ward with a lighting system that produces an environment with blue depleted evening light or to a ward with the same layout and facilities but lacking the new lighting technology. The trial will test whether the environment with programmable lighting conditions offers any additional benefits beyond those associated with usual treatment in an acute psychiatric inpatient unit. The main objectives are to examine if there are any differences between groups in the mean duration of hospitalization. In addition, the investigators will explore whether level of functioning, day-to-day symptom ratings, episodes of suicidality or aggressive behavior, medication usage, and self-reported side-effects differ between groups and whether shorter duration of admission is associated with greater stability of sleep-wake cycles and circadian rhythms.

Given that this trail takes place in a unique setting, the investigators will undertake several ancillary investigations to determine the range of benefits or adverse effects for subpopulations of patients (e.g. different diagnostic subgroups) and examine any potential benefits or drawbacks to the use of this new lighting system, including actigraphic recordings of sleep-wake patterns of nurses and any self-reported effects on well-being that are recorded by professionals who experience working in both wards.

Update March 8th, 2019

As noted in the original protocol, we aimed to recruit a minimum of 400 individuals who give written informed consent to participate in a two-arm pragmatic effectiveness randomized controlled clinical treatment trial (RCT), with the expectation that we would actually include about 200-250 individuals per group. This recruitment target takes into account 'failure to complete randomization' (i.e. the patient is randomized to a specified lighting condition, but is not admitted to the bed in the ward with the allocated lighting condition; this may occur for a number of reasons that were listed in the original protocol) and the delayed consent procedure (we note that the delay between randomization and consent is, on average, about 14 days).

The recruitment period was about 6 months (Oct 23th 2018 until April 12th 2019), but we set a limit on randomization to March 31st and for obtaining delayed consent of April 12th. This time frame was selected as it corresponds to the winter period in this region of Norway. After March 31st, daylight saving time changes and also we experience very different environmental light exposure (long days, short nights, etc). This is relevant as it means that any extension to the recruitment period needs to be delayed until the environmental light exposure matches that experienced by individuals who entered the RCT at the start of recruitment. As such, we agreed (in advance of starting the RCT) with the ethics committee and trial steering group, that should we fall short of the target sample size after six months or recruitment, then the recruitment process would be suspended and would be re-opened in October 2019 (when the environmental light exposure would be the same as during the original winter recruitment period).

Although our projected randomization rate is as expected (and we have randomized >600 individuals by month 4-5 of the study), two issues have arisen with completion of randomization. The first was totally unpredictable, namely that a regional electricity power failure meant that we could not properly randomize admissions for a 72-hour weekend period (during with >15 admissions occurred). After discussion, it was agreed that the manual system of randomization that was introduced over that weekend could not be regarded as equivalent to the automated system, so these cases were immediately excluded from the RCT sample. Second, the bed occupancy rate in January and February has been running at about 100%, meaning that many randomizations cannot be completed (i.e. the patient is randomized but cannot then be admitted to the lighting condition to which they were allocated).

We undertook modelling of the projected number of randomizations that would be completed (i.e. admitted to the lighting condition allocated at randomization) and have given written informed consent by April 12th. the projected admission rates, we believe that this sample size is at the lower limit of the estimated study population (as we are permitted to continue recruitment for at least six consecutive months). We modelled the best- and worst-case scenario. The best-case scenario suggested we would achieve the target sample by April 12th (>400), but the worst-case scenario suggested we may only have randomized, completed correct allocation and obtained the delayed consent in about 350-360 cases.

Given the above, and the pre-trial agreement that recruitment could be suspended and re-opened in October 2019, this is course of action we have taken. As we do not break the codes for randomization (completed as allocated or failed) or match these to which individuals have given consent (via the deferred procedure), we cannot know the exact number of trial participants we need to recruit in October. However, the original protocol stated we would recruit 400-500 participants. With this in mind, we will re-open randomization for six weeks, starting October 2019, and close the consent procedure by the end of November 2019.

Added October 7t,h 2021:

The Statistical Analytical Plan (SAP) has been written before anyone have performed any analyses on data from the trial. Analyses of the data from the trial will start one week after the SAP has been uploaded to https://register.clinicaltrials.gov.