Quality of death: assessing the importance placed on end-of-life treatment in the intensive-care unit

Med Care. 2004 May;42(5):423-31. doi: 10.1097/01.mlr.0000124245.62354.57.


Context: The value of good end-of-life (EOL) care could be underestimated if its effects are assessed using the standard metric of quality-adjusted survival, especially if the time horizon is limited to the duration of the EOL care. This issue is particularly problematic in the intensive-care unit (ICU) where death is frequent, care is difficult, and costs are high.

Objectives: The objectives of this study were to test whether people would trade healthy life expectancy for better EOL care, to understand how much life expectancy they would trade relative to domains of good care, and to determine the association of respondent characteristics to time traded.

Design and subjects: We used a computerized survey instrument describing hypothetical patient experiences in the ICU used to assess attitudes of a general population sample (n = 104) recruited in Pittsburgh, Pennsylvania.

Measures: We used life expectancy traded (from a baseline of 80 healthy years followed by a 1-month fatal ICU stay) for improving ICU care in 4 domains: pain and discomfort, daily surroundings, treatment decisions, and family support.

Results: Three fourths of respondents (n = 78) were prepared to shorten healthy life for better EOL care. Median time traded in individual domains ranged from 7.2 to 7.7 months overall and 9.6 to 11.4 months when restricted to those willing to trade. Median time traded for improvement in all domains was 8.3 months overall and 24.0 months by those willing to trade. In multivariable analyses, respondents who were older, nonwhite, or had children traded significantly less time, whereas those who did not perceive the ICU to be a caring environment traded more time.

Conclusions: Good EOL care is highly valued, both in terms of medical and nonmedical domains, as suggested by previous work and confirmed by our data showing respondents trading quantities of healthy life several times longer than the duration of the EOL period itself. The considerable interperson variation highlights the importance of soliciting individual preferences about EOL care.

Publication types

  • Research Support, Non-U.S. Gov't
  • Research Support, U.S. Gov't, P.H.S.

MeSH terms

  • Adult
  • Attitude to Death*
  • Critical Care / psychology
  • Critical Care / standards*
  • Female
  • Humans
  • Life Expectancy
  • Linear Models
  • Male
  • Pennsylvania
  • Quality of Health Care*
  • Quality of Life*
  • Statistics, Nonparametric
  • Terminal Care / psychology
  • Terminal Care / standards*