Rosuvastatin versus placebo for delirium in intensive care and subsequent cognitive impairment in patients with sepsis-associated acute respiratory distress syndrome: an ancillary study to a randomised controlled trial

Abstract

Background: Delirium is common in mechanically ventilated patients and is associated with cognitive impairment lasting at least 1 year after hospital discharge. Preclinical and observational studies suggest that the use of statins might reduce delirium in intensive care. We assessed whether the pleiotropic effects of statins can reduce delirium in intensive care and decrease subsequent cognitive impairment in a randomised controlled trial.

Methods: We did this ancillary study within the SAILS trial, a randomised controlled trial assessing mortality and ventilator-free days for rosuvastatin versus placebo for patients with sepsis-associated acute respiratory distress syndrome. This study was done at 35 hospitals in the USA. Patients were randomly assigned in permuted blocks of eight and stratified by hospital to receive either rosuvastatin (40 mg loading dose and then 20 mg daily until the earliest of 3 days after discharge from intensive care, study day 28, or death) or placebo. Patients and investigators were masked to treatment assignment. Delirium was assessed with the validated Confusion Assessment Method for intensive care. Cognitive function was assessed with tests for executive function, language, verbal reasoning and concept formation, and working, immediate, and delayed memory. We defined cognitive impairment as having one of these domains at least two SDs below population norms or at least two domains at least 1·5 SDs below norms. The primary endpoint was daily delirium status in intensive care up to 28 days in the intention-to-treat population and secondary endpoints were cognitive function at 6 months and 12 months. This trial is registered with ClinicalTrials.gov (NCT00979121 and NCT00719446).

Findings: 272 patients were assessed for delirium daily in intensive care. The mean proportion of days with delirium was 34% (SD 30%) in the rosuvastatin group versus 31% (29%) in the placebo group; hazard ratio 1·14, 95% CI 0·92-1·41, p=0·22. At 6 months, 19 (36%) of 53 patients in the rosuvastatin group versus 29 (38%) of 77 in the placebo group had cognitive impairment, with no significant difference between groups (treatment effect 0·93, 95% CI 0·39-2·22; p=0·87). At 12 months, 20 (30%) of 67 patients versus 23 (28%) of 81 patients had cognitive impairment, with no significant difference between groups (treatment effect 1·1, 95% CI 0·5-2·6; p=0·82).

Interpretation: Most patients had delirium, with around a third of survivors having cognitive impairment over 1 year of follow-up. Despite encouraging preclinical and observational studies, this trial shows no benefit of rosuvastatin in reducing delirium in intensive care or cognitive impairment during 12 months of follow-up although the study was not powered for superiority. Thus, there is continued need to evaluate interventions aimed at attenuating intensive care and post-intensive-care cognitive impairments commonly observed in this population.

Funding: National Heart, Lung and Blood Institute; Johns Hopkins Institute for Clinical and Translational Research; the SAILS trial was also supported by AstraZeneca.

Figure 1. Trial profile

*Less than the 745 patients in the SAILS trial because two sites did not participate and there was a temporary cessation of recruitment into this ancillary study. †57 (17%) of 329 patients were excluded from assessment of delirium, but still eligible for follow-up at 6 months and 12 months: 46 (14%) were in intensive care before the protocol revision expanding assessment of delirium and cognitive status to other hospitals, eight (2%) were at sites not participating in delirium assessment, and three (1%) had missed delirium assessment. ‡Because of the timing of approval by the institutional review boards for this ancillary study, but patients later provided consent and were eligible for follow-up at 12 months.

Figure 2

Proportion of patients with cognitive impairment

Figure 3. Treatment effect at 6 months

The effect size is the treatment effect (difference in means or proportions) divided by the pooled standard deviation from the rosuvastatin and placebo groups.^,