A retrospective review of a large cohort of patients undergoing the process of withholding or withdrawal of life support.

OBJECTIVES

To determine the proportion of patients who died as a result of the withdrawal or withholding of life support (WD/WHLS) in the intensive care units (ICUs) of three teaching hospitals and to describe the process involved by determining: a) why the decision was made to withdraw support (WDLS); b) when WDLS took place; and c) how the WDLS process was conducted.

DESIGN

Retrospective cohort study.

SETTING

Three university-affiliated ICUs.

PATIENTS

Four hundred nineteen patients who died in one of three academic, tertiary care ICUs over a 1-yr period.

INTERVENTIONS

Retrospective chart review. Data collected included age, gender, admitting diagnoses, comorbid disease, Acute Physiology and Chronic Health Evaluation II score, and mode of death (brain death, death due to withholding of life support, death due to WDLS, or death despite ongoing therapy). For those patients dying due to WDLS, the reason for WDLS, person initiating discussion, timing of WDLS, degree of organ dysfunction, order of withdrawal of life support modalities, and the use of sedatives and analgesics were recorded.

MEASUREMENTS AND MAIN RESULTS

Seventy percent of patients died by WD/WHLS and 8.4% were brain dead. Patients undergoing WD/WHLS were older and had a longer length of stay than patients dying from other causes. Poor prognosis was the most common reason given for WDLS, reflected by significant organ dysfunction at the time of WDLS. Future quality of life was a less frequently cited reason. Most patients undergoing WDLS did so early in their ICU stay, although time to WDLS appeared to reflect diagnostic group. Few meetings occurred before WDLS and death occurred soon after initiating WDLS. There was a preference of withdrawing mechanical ventilation last and large amounts of morphine (mean 21 +/- 33 [SD] mg/hr) and benzodiazepines (mean 8.6 +/- 11 mg/hr) were used. Little variability in practice was apparent among the studied ICUs.

CONCLUSIONS

Similar to other studies, WD/WHLS was the most common cause of death in academic ICUs and poor patient prognosis was considered the most important factor in deciding on WDLS. However, in contrast to other studies, future quality of life was not as frequently cited a reason for WDLS and larger amounts of morphine were used during WDLS. These discrepancies suggest areas for potential future research.