As the aim of this quality improvement study was process
improvement, the primary outcome measure was hours of
discharge delay per patient discharged alive per month.
Patients were deemed ready for discharge after discharge
paperwork, including discharge summary and orders and
associated nursing carewere completed. Patientswere considered
to have no delay if this time period was less than
1 hour (Chaboyer et al. 2006). Secondary outcomes included
in-hospital mortality after discharge from ICU and
ICU readmission within 72 hours. These outcomes were
chosen because several previous studies have shown that
other discharge practices can lead to these negative outcomes
(Goldfrad & Rowan 2000; Beck et al. 2002; Bell &
Sales 2004; Duke et al. 2005; Priestap&Martin 2006; Laupland
et al. 2008). All outcome datawere accessed from the
routinely collected Australia and New Zealand Intensive
Care Society Centre for Outcome and Resource Evaluation
(CORE) database for 15 months prior to (January 2007
to March 2008) and 12 months after (May 2008 to April
2009) the new discharge process was implemented, which
began in April 2008. This database, or registry, captures a
large amount of data about patients who are admitted to
ICUs in Australia. Demographic and clinical data as well
as some patient outcomes such as mortality and ICU readmission
are recorded in CORE. An experienced and trained