Table 2.

Studies included in the review.

Authors, publication year, and country	Aim (A), study design (D), and intervention (I)	Setting (S) and sample size (SS)	Findings of interest in this review	Generalisation	Suggested future research
Chaboyer et al. [26], 2006, Australia	A: to examine the impact of an ICU liaison nurse on discharge delay D: prospective block intervention study I: liaison nurse in discharge process	S: 580-bed tertiary hospital with a 13-bed ICU SS: n = 186 patients (n = 101 control group, n = 85 intervention group)	The liaison nurse assessed patients for transfer to the ward and coordinated patient transfers, including communication with ward staff prior to and after discharge. Patients whose discharge did not involve the liaison nurse were 2.5 times more likely to have a delay of 4 h or longer in comparison to the ones that did	Single-centre study	A hospital-wide perspective on service enhancements Evaluate the liaison nurse's role Explore factors influencing ICU discharge delay
Chaboyer et al. [25], 2012, Australia	A: to evaluate the impact of an ICU nursing discharge process redesign D: time series design I: redesigned discharge process	S: 12-bed general ICU incl. HDU beds in a 580-bed hospital SS: n = 1,787 discharges (n = 1,001 before, n = 786 after implementation)	A redesigned ICU discharge process demonstrated a 3.2 h reduction in the average patient discharge delay time (from 4.6 h to 1 h). Both ward and ICU staff were involved in the process and this may have contributed to mutual situational awareness leading to more timely and effective discharge processes. The changes decreased discharge delays without increasing mortality or readmission rates	Single-centre study	Demonstrate the effect of different changes in the ICU discharge process on outcomes
Gillman et al. [28], 2006, Australia	A: to determine the incidence and nature of adverse events occurring during transfer from the ED to the ICU D: prospective observational study	S: ED and ICU in a tertiary hospital SS: n = 290 patient transfers	38% (n = 54) of patient transfers from the ED to the ICU were delayed. Delay times were less than 20 min (68%), 21–60 min (24%), and 61–120 min (9%) and more than 121 min (5%). Delays were caused by a lack of ICU beds (30%), a lack of staff (15%), busy workload in the ICU (9%), delay in the availability of transfer orderlies (11%), computed tomography (CT) not ready (9%), OR not ready for transfer (7%), ICU busy (5%), no anaesthetist (6%), speciality rereview (7%), and unknown reasons (6%).	Single-centre study	Establish benchmark indicators for adverse events and transfer delays
Johnson et al. [29], 2013, USA	A: to analyse the incidence, causes, and costs of delayed transfer from a surgical intensive care unit (SICU) D: prospective observational study	S: 900-bed tertiary hospital with a 20-bed surgical ICU SS: n = 731 patient transfers	22% (n = 160) of ICU transfers to the wards were delayed. Delay times varied from 1 to 6 days (mean 1.5, median 2 days). Delays were caused by lack of an available ward bed (71%), lack of an appropriate room for infectious patients (18%), change of medical speciality (7%), and lack of an available bedside nurse for the patient (3%). A positive association existed between the number of patients in the hospital and the number of ICU beds occupied by delayed patient transfers (Spearman rho = 0.27, p < 0.0001). Delayed patients were more likely to be transferred between 7 p.m. and 6.59 a.m. (21% versus 12%, p < 0.005). The delay-related costs were estimated to be US $21,547/week	Single-centre study	Calculate cost data for ICU discharge delays Analyse census of wards for delayed patients, while emphasising bed availability Interventions to reduce beds occupied by discharge-ready patients Study further ICU transfer delays
Lin et al. [24], 2013, Australia	A: to explore the factors influencing the ICU patient discharge process D: ethnographic study	S: 580-bed tertiary hospital with 14-bed, level 3 medical or surgical ICU SS: n = 28 discharges	43% of the ICU discharges were delayed. 11% of the delays lasted for 1 day, while 32% of delays lasted for 2-3 days. Altogether, 33% of discharges were delayed due to the limited availability of ward beds. Three activity systems were identified in the discharge process: the ICU discharge activity, the ward accepting the ICU patient, and the management of hospital beds. Better coordination, communication, and shared goals could decrease discharge delays	Single-centre study	Strategies to improve the efficiency of acute care beds Develop tools to support the discharge process Explore the role of the discharge guidelines
Williams and Leslie [32], 2004, Australia	A: to examine the prevalence and reasons for delayed discharges D: cross-sectional observational study	S: 955-bed tertiary hospital with a 22-bed mixed ICU SS: n = 652 discharges	27.3% (n = 176) of ICU discharges were delayed. The median delay time was 21.3 h with a variation ranging from 10 min to 26 days. Delays were caused by no available beds (75%), ward bed delayed (5.7%), medical complication (8.5%), the environment (0.6%), a lack of medical coverage (0.6%), transport (0.6%), unknown reasons (5.7%), the closure of a ward (1.7%), and other reasons (e.g., lack of an available room for infection control, no available nurses, and inadequate nursing skills on the ward) (1.7%). Most delays occurred on weekends	Single-centre study	Apt admission and discharge criteria Determine the effect of hospital occupancy on discharge and the factors causing care transition between medical specialties Determine the optimal number of ICU and intermediate care beds
Williams et al. [33], 2010, Australia	A: to examine whether the introduction of a critical care outreach role would decrease the frequency of ICU discharge delays D: comparison of observational data from 2000/2001 and 2008	S: 955-bed tertiary hospital with a 22-bed general ICU. SS: in 2000/2001 n = 607, in 2008 n = 516 discharges	31% (n = 488) of the ICU discharges were delayed by more than 8 h in 2008, with an increase of 6% from 2000/2001 (p < 0.001). In 2008, the reasons for delay included bed delay (17%), no bed (36%), staff shortage (2%), no accommodation (1%), other delays (20%), and medical concerns (24%). In 2000/2001, the reasons for delay included no bed (74%), bed delay (6%), medical reasons (9%), no accommodation (1%), and other reasons (10%). The mean delay time, when excluding medical reasons, was 21 h in 2000/2001 and 25 h in 2008. After-hour discharges were more frequent in delayed discharges and these occurred more often in 2008 when compared to 2000/2001. Patients also spent more time in the hospital when the discharge was delayed	Single-centre study	Examine the effects of bed management models on patient flow Study the associations with ICU discharge delays