PURPOSE: To describe the observed sequence of withdrawal of eight different
forms of life-sustaining treatment and to determine whether aspects of tho
se treatments determine the order of withdrawal.
SUBJECTS AND METHODS: We observed 211 consecutive patients dying in four mi
dwestern US hospitals from whom at least one of eight specific life-sustain
ing treatments was or could have been withdrawn. We used a parametric stati
stical technique to explain the order of withdrawal based on selected chara
cteristics of the forms of life support, including cost, scarcity, and disc
omfort.
RESULTS: The eight forms of life support were withdrawn in a distinct seque
nce. From earliest to latest, the order was blood products, hemodialysis, v
asopressors, mechanical ventilation, total parenteral nutrition, antibiotic
s, intravenous fluids, and tube feedings (P <0.0001). The sequence was almo
st identical to that observed in a previous study based on hypothetical sce
narios. Forms of life support that were perceived as more artificial, scarc
e, or expensive were withdrawn earlier.
CONCLUSION: The preference for withdrawing some forms of life-sustaining tr
eatments more than others is associated with intrinsic characteristics of t
hese treatments. Once the decision has been made to forgo life-sustaining t
reatment, the process remains complex and appears to target many different
goals simultaneously. (C) 1999 by Excerpta Medica, Inc.