Intensive care units (ICU) may be associated with poor-quality end-of-life care. A recent study’s findings call that into question — at least from the family’s perspective.1

“ICUs have traditionally been viewed as the worst place in the world to die, even worse than regular hospital wards,” says Josh Rolnick, MD, JD, the study’s lead author and a clinical scholar in the National Clinician Scholars Program at the University of Pennsylvania. “You see that assumption among clinicians, policymakers, and researchers. We wanted to understand if this assumption was correct.”

Rolnick and colleagues examined the association between ICU care during terminal hospitalization and family ratings of end-of-life care for patients who died in more than 100 Veterans Affairs hospitals between 2010 and 2016. The authors created four categories: patients who received no ICU care, those who received only ICU care, those who received ICU care but died outside the unit, and patients who received ICU care and died in the unit.

Of 57,550 decedents, a family member or close contact had completed a survey for 28,062 decedents. Before looking at the survey results, Rolnick and colleagues hypothesized that ICU use would be linked to lower ratings of care. However, they discovered that the opposite was true: More ICU time was linked to higher ratings.

That was true not just for patients treated solely in ICUs, but also for patients who were treated in both levels of care (ICUs and general wards). “Unfortunately, another surprising finding was the high level of uncontrolled pain reported across all hospital settings, even in ICUs,” Rolnick says. About half of families reported “some” or “significant” pain during the patient’s last month of life.

“If it is true that ICUs lead to better end-of-life care, it may be for a few reasons,” Rolnick offers. For one thing, the staffing ratios are different. There are more clinical staff available per patient, and the nursing staff ratio also is lower. “With only one or two patients rather than four or more, the nurse has more time available to devote to the needs of the patient and family,” Rolnick says.

People also may sense that “everything” was handled in the ICU setting. Additionally, says Rolnick, “ICU staff are more experienced with end-of-life care than ward staff because they treat more critically ill patients.”

The study’s findings do not imply that people should be admitted to ICUs for end-of-life care, Rolnick stresses. ICU ratings were worse than nursing homes and hospice units. “While we don’t know if this association is causal, it does suggest that we should make efforts to avoid having patients die in acute hospitals if feasible,” Rolnick says.

It is not as simple as just making sure people stay out of the ICU. Doing so, says Rolnick, “may even worsen their experience. Instead, we need to find ways to get them out of acute care hospitals entirely.”

The findings carry important implications for discussions on end-of-life decision-making. “It may not be appropriate to counsel people that the end-of-life experience will be worse in the ICU than on regular hospital wards,” says Rolnick. “That is not so clear.”

What is clear is the complexity of improving end-of-life care. “It may be a matter of having better staffing ratios for patients who need intensive palliative efforts,” Rolnick offers.


  1. Rolnick JA, Ersek M, Wachterman MW, Halpern SD. The quality of end-of-life care among intensive care unit versus ward decedents. Am J Respir Crit Care Med 2020;201:832-839.