The use of 'do not resuscitate' (DNR) orders in hospitals has been the
subject of considerable comment in both the medical and the lay press
. Guidelines have been produced to help make DNR decisions but, as yet
, there have been no published accounts of these in practice. We have
used audit to develop DNR policy in our hospital, and have reviewed pr
actice after the introduction of guidelines. This led to early consult
ant involvement in making decisions in 55 of 80 patients (69%) who wer
e assessed as DNR at the time of death or discharge, documentation of
reasons for DNR in all 55 of these and documentation of discussion wit
h nurses in 49 (89%). Consultants agreed with DNR decisions made by th
eir juniors in 31 of 34 cases (91 %) and changed 'for CPR' decisions t
o DNR in 24 of 108 (22%). We have demonstrated that audit is an approp
riate way to change and develop practice in sensitive areas such as th
is.