DOI: 10.1097/CCM.0b013e31828c25d8

,  

,  

PMID: 23685576

Issn Print: 0090-3493

Publication Date: 2013/06/01


Print

Palliative Care in the ICU: Lots of Questions, Few Answers*

Robert D. Truog

    loading  Checking for direct PDF access through Ovid

Excerpt

Given that one in five Americans will die in an ICU (1), intensivists have an obligation to make sure that the quality of care they provide at the end of life is the best possible. In recent years, Curtis et al at the University of Washington have performed a series of investigations that have given us a better understanding of palliative care in the ICU, provided us with tools to measure the quality of this care, and made efforts to improve our practices. In this issue of Critical Care Medicine, DeCato et al (2) publish the latest addition to this corpus of work. The study reported here is a subanalysis of data collected from a cluster randomized trial at a dozen hospitals in the Seattle-Tacoma area designed to test a quality improvement intervention for end-of-life care in the ICU (3). Unfortunately, this multifaceted intervention did not result in improvement in the quality of end-of-life care or shorten the length of stay in the ICU before death.
The analysis reported here includes data from 3,065 decedents who were enrolled in the trial, which were collected over 4½ years. The investigators found significant differences between hospitals for all chart-based indicators of the quality of palliative care as well as significant differences for survey data of family satisfaction, family-rated quality of dying, and nurse-rated quality of dying. In addition to variation between hospitals, they also looked for temporal trends in the data over the duration of the study and found that although there was a significant increase in pain assessments in the last 24 hours of life, there was a significant decrease in the documentation of family conferences and discussion of prognosis in the first 72 hours after ICU admission.
What can we learn from these data? Perhaps the most striking finding is the variability seen between hospitals in this fairly small geographical area. In addition, the investigators found that the quality was inconsistent within hospitals, with individual hospitals scoring highly on some measures but poorly on others, such that none of the hospitals performed better than its peers overall. These perplexing data lead one to wonder what is driving this variability between and within hospitals. The most compelling driver would be patient preferences; if the intensity of care at the end of life correlated with patient preferences, we could at least be reassured that patients are getting what they want. As the authors discuss, however, the limited data available suggest that this is not the case (4), and indeed none of the research to date has been successful in identifying the factors that are driving this variability. Some have suggested that the source of this variation may lie in the more difficult-to-measure social norms that shape hospital practice patterns (5).
The authors did not find much evidence that palliative care in ICUs improved over the course of the study, but 5 years is a short time over which to see changes in practices that depend on so many organizational and cultural factors. Nevertheless, compared with the alarmingly poor care reported from the SUPPORT study in the early 1990s (6), I think that there is reason to believe that palliative care has improved over this longer time frame. For example, this study found that family meetings were documented for 73% of the patients within the first 72hours of admission.
    loading  Loading Related Articles