Journal Article Annotations
2021, 4th Quarter
Annotations by Natalie Fedotova, MD, PhD and O. Joseph Bienvenu, MD, PhD
Using a stepped-wedge cluster-randomized controlled design, investigators implemented a multicomponent nonpharmacological nursing intervention program to reduce the frequency of delirium among high risk ICU patients. The intervention focuses on addressing delirium risk factors including hearing impairment, orientation loss, sleep deprivation, cognitive impairment, and immobility. Following implementation, the authors found no improvement in delirium-free and coma-free days alive at 28 days as well as only a limited increase in time spent on interventions per nursing shift.
Strength and weaknesses:
This was a large cluster-randomized controlled trial, across multiple ICUs, and included medical, surgical, and trauma patients. Waiving explicit patient consent likely allowed for enrollment of a more representative study sample. In addition, this investigation examined the impact of multi- rather than single-component interventions. Given the nature of the interventions, however, blinding was not possible. As the selected ICUs came from the Dutch ICU Delirium Consortium (a group committed to improving delirium care), the quality of their baseline delirium care (control) was likely quite high.
Nonpharmacological interventions to optimize modifiable delirium risk factors are one of the cornerstones of delirium prevention/management, and current recommendations focus on multicomponent approaches (for example, Devlin 2018). The present study describes a large-scale effort to implement such interventions systematically, with formation of local working groups, education, and ongoing communication. The lack of a significant effect post implementation is difficult to interpret in light of the limited increase in time spent on interventions; it is possible that given the commitment of these medical centers to improving delirium care, they were providing a high level of care at baseline. Further work is necessary to clarify this finding and explore which intervention bundles would be of particular value in the critically ill population.
Jo Ellen Wilson and colleagues at Vanderbilt have been studying catatonia in the context of critical illness for several years. In this study, the investigators measured catatonic symptoms carefully, attempting to avoid over-calling symptoms that could be part of the coma or critical illness phenotypes. Remarkably, more than 1 in 5 critically ill patients had catatonia, which was more common in older patients. Intriguingly, catatonia appeared to mediate the relationships between age and both delirium and coma in this population. Staring, mutism, and immobility were strongly intercorrelated.
Strength and weaknesses:
Strengths include rigorous assessment methods, well-characterized cohorts of patients with a variety of ages and critical illness types, and sophisticated statistical methods. The study employed a convenience cohort, nested within two clinical trials and two observational studies.
As the authors note, though it is now increasingly (and appropriately) common to measure level of arousal and delirium in critical care settings, it is not common to measure another form of acute brain dysfunction, catatonia. It would be useful to know whether catatonia had independent prognostic significance, and whether it influenced (or should have influenced) treatment in any way (e.g., reduction of antipsychotic medication doses).
As the authors note, we humans appear to process environmental stimuli even when in altered states of consciousness. They propose that an intervention to verbally positively (and appropriately) frame what is happening, even when patients are delirious, may improve long-term psychiatric outcomes. The staff intervening (“doulas”, borrowing from obstetric practice) also held patients’ hands, when appropriate, and continued to meet with patients once the patients were no longer sedated and able to communicate. In addition, intervening staff met with patients upon ICU discharge, to let patients know what to expect and normalize experiences. The investigators found that the intervention was feasible (the aim of this pilot study).
Strength and weaknesses:
The authors thoughtfully considered whether a purposeful use of extra staff (ICU doulas appear unique to this study) may improve long-term psychiatric outcomes. The design relied on historical controls to compare short-term psychiatric outcomes (within 96 hours of ICU discharge). Notably, patients were excluded if they had severe metabolic encephalopathy.
If, in a future randomized clinical trial, the ICU doula intervention is shown to meaningfully prevent post-ICU emotional distress, this intervention may spread. Unfortunately, no routine intervention is known to consistently improve critical illness survivors’ long-term mental health outcomes, so this seems worth trying.