DELIRIA-J

Published

Hatta K, et al. "Preventive effects of ramelteon on delirium: a randomized placebo-controlled trial". JAMA Psychiatry. 2014. 71(4):397-403.
PubMed • Full text

Clinical Question

In recently hospitalized elderly patients, does ramelteon reduce the risk of delirium compared to placebo?

Bottom Line

In this small single-center study, ramelteon reduced the risk of delirium in elderly inpatients.

Major Points

Delirium and other acute confusional states are common among hospitalized patients, and older individuals are disproportionately affected. Because treatment is challenging and often ineffective, the mainstay of management has been prevention. A 2007 Cochrane review concluded that sparse evidence existed supporting any intervention to prevent delirium, citing only early geriatric consultation as an evidence-based intervention in selected patients. There remains a sizable gap in the evidence for strategies that can prevent delirium.

The DELIRIA-J study randomized 67 hospitalized patients from 65-89 years of age to either the ramelteon or placebo, given daily for up to 7 days. Patients with pre-existing severe liver disease, dementia, and other disorders associated with fluctuating mental status difficult to discern from delirium were excluded. Patients were evaluated daily for the development of delirium using the Delirium Rating Scale-Revised-98 (DRS-98). By the end of the 7-day study period, delirium had developed in 1 of 33 patients receiving ramelteon (3%) and 11 of 34 patients receiving placebo (32%), representing a relative risk of 0.09 (95% CI 0.01-0.69; P=0.003). The authors reported no adverse events associated with either study drug.

This is a well designed study that suggests that the melatonin agonist ramelteon may reduce the risk of delirium in elderly patients hospitalized with an acute illness. Based on these data, three patients would need to be treated with daily ramelteon to prevent one episode of delirium, in line with a 2011 study comparing melatonin to placebo. Indeed, a 2019 network meta-analysis suggested that among the agents studied for delirium prevention, ramelteon may be the most effective. Caution is advised in generalizing the results of this study, however, since the evidence base for ramelteon and melatonin stems from small, underpowered studies such as DELIRIA-J. An editorial accompanying the publication of DELIRIA-J expounds on the limitations of this study. Data from DELIRIA-J were included in a 2016 Cochrane review which found no clear evidence that melatonin or its agonists reduce the risk of delirium in hospitalized patients.

Guidelines

As of September 2021, no guidelines have been published that reflect the results of this trial.

Design

• Multicenter, single-blind, randomized, controlled trial
• N=67 patients admitted to general medical or intensive care unit
  • Ramelteon (n=33)
  • Placebo (n=34)
• Setting: 5 centers in Japan
• Enrollment: 2011-2012
• Mean follow-up: 7 days
• Analysis: Exploratory
• Primary outcome: Incidence of delirium (by DSM IV criteria)

Population

Inclusion Criteria

• Age 65-89 years
• Newly admitted due to serious medical problems
• Expected inpatient stay >48 hours
• Expected life expectancy >48 hours

Exclusion Criteria

• Delirium present on admission
• Psychotic or mood disorder
• Alcohol or drug withdrawal
• Lewy body disease
• Severe liver dysfunction

Baseline Characteristics

• Mean age: 78 years
• 19% had dementia
• 21% used habitual sleeping pills
• Admission diagnoses in the ramelteon group: 36% stroke, 12% infection, 18% fracture, 9% HF/MI

Interventions

• Randomized to ramelteon 8 mg/night or non-matching placebo for 7 days
• Psychiatrists were trained and assigned to specific sites, and would make rounds every afternoon for 7 days to assess each patient for the onset of delirium after reading nursing reports and talking with each nurse.
• DRS-R98 (delirium screening), APACHE II, and ECOG status were recorded daily

Outcomes

Comparisons are ramelteon vs. placebo

Primary Outcomes

Incidence of delirium

3% vs. 32% (RR 0.09; 95% CI 0.01-0.69; P=0.003)

Secondary Outcomes

No statistical differences in sleep parameters, worst APACHE II, worst ECOG scores, or use of hydroxyzine.

Subgroup Analysis

Incidence of delirium in patients not using PRN hydroxyzine

4% vs 32% (P=0.007)

Adverse Events

No adverse events potentially attributed to the study drugs were observed.

Criticisms

• Small sample sizes and the risk of publication bias limit the reliability of this study, though its intervention appears to be low-risk
• Though the trial was not double blind, it was rater-blind, with concealed allocation, and not unmasked until after trial analysis, thus the study outcome of delirium was unlikely to have been affected by this limitation.
• Nurses were not blinded to treatment group, though this would be unlikely to make them provide better delirium preventive care in the melatonin group.
• Though the placebo arm received more hydroxyzine, primary outcome remained significant in subgroup analysis with these patients excluded

Funding

Funded by a grant from Japan Society for the Promotion of Science