ArticlesConsequences of discharges from intensive care at night
Introduction
It is generally believed that pressure for beds on intensive care units (ICUs) has increased. Although no rigorous research evidence exists, the increasing number of reports and correspondence related to this subject does suggest that the pressure on bed availability in ICUs may be greater now than before.
Increasing numbers of admissions with a concomitant decrease in length of stay in the ICU have been reported.1, 2, 3 Occupancy rates have been described as very high.4 The shortage of available ICU beds in London, UK, and elsewhere2 has been detailed and a wide variation in provision of facilities has been recorded.4, 5, 6 Cancelled operations due to the lack of available ICU beds4 and high rates of refused admissions have been reported, both regionally7 and nationally.5, 6 The transfer of patients over long distances in search of an ICU bed has also been reported8, 9, 10 and the potential dangers of transferring critically ill patients has been highlighted.11 Premature discharge of patients has been described.3, 12, 13
A national bed register covering England was introduced in 1996 by the Department of Health to ease the problem of finding a suitable bed for a critically ill patient.14 In addition, the Intensive Care Society, the professional organisation of intensive-care doctors, has called for both an increase in ICU facilities in areas of low provision and a formal transport system.15
High dependency units (HDUs) have been proposed for low-risk, short-stay ICU admissions to ease the pressure.1, 3, 12, 16, 17 However, a shortage of such intermediate facilities also seems to exist.
With the results of the Intensive Care Society's UK APACHE II study from July, 1988 to September, 1990, which indicated a high death rate on the ward after discharge from ICU (variation 6–16% across ICUs),18 and the belief that discharging patients from ICUs at night does not constitute good quality care and, where possible, is to be avoided, our study used discharge at night as a proxy measure to investigate pressure on ICUs. The aim was to investigate the change, over time, in the rate of discharge at night from ICUs and to find whether there were any adverse consequences following discharge at night.
By the use of two high-quality clinical databases, the UK APACHE II study database,18, 19 and the Intensive Care National Audit & Research Centre's Case Mix Programme Database (CMPD),20 the rate of discharge at night for 1988–90 was compared with the rate for December, 1995, to April, 1998. We used data solely from the CMPD to compare the consequences of discharge at night with discharge during the day.
Section snippets
Database
Data were extracted for the 10 806 admissions to 26 ICUs in the UK APACHE II study database, covering 1988–90, and for the 22 059 admissions to 62 ICUs in the CMPD, covering 1995–98. All these data had been collected prospectively. Age exclusion criteria (age <16 years) used in the UK APACHE II study were applied to the CMPD resulting in 21 295 adult admissions. Deaths in ICU were excluded from the analyses.
Data
Data were extracted for “time of discharge from ICU”. We consulted with ICU colleagues
Trend in night discharges
There were 2269 (21·0%) admissions who did not survive the ICU during 1988–90 compared with 4487 (21·1%) during 1995–98. Overall, 2·7% of discharges occurred at night in 1988–90 compared with 6·0% in 1995–98 (table 1), a 2·2-fold increase. The proportion of discharges at night varied 16-fold across ICUs in 1988–90 compared with 25-fold in 1995–98. Similar results were seen when the definition for discharge at night was restricted to discharge from ICU between 0000 h and 0459 h. When the ICUs
Discussion
Night discharges from ICU doubled in the UK over the past decade—a worrying trend because patients discharged at night fare significantly worse than those discharged during the day. Before considering possible explanations for these findings, it is important to recognise a potential methodological limitation—the adequacy of the UK APACHE II model for case-mix adjustment. While we can never be certain that all potential risk factors have been taken into account, the model used was developed and
References (24)
- et al.
Intensive care in the ailing UK health care system
Lancet
(1995) - et al.
Mortality among appropriately referred patients refused admission to intensive-care units
Lancet
(1997) - et al.
A changing role for intensive therapy: is there a case for high dependency care?
Anaesthesia
(1994) Rationing intensive care—crisis reactions to shortages should be replaced by analysis of strategic services
BMJ
(1995)- et al.
Rationing intensive care—intensive care provision varies widely in Britain
BMJ
(1995) - et al.
Study of provision of intensive care in England, 1993
(1995) Review of intensive care services in the West Midlands
(1994)ICU patient moved three times a day
Health Serv J
(1995)Intensive care units are overloaded, inquiry finds
Health Serv J
(1995)Bed register could be set up after patient's death
BMJ
(1995)
Rationing intensive care: preventing critical illness is better, and cheaper, than cure
BMJ
A prospective two-month audit of the lack of a high-dependency unit and its impact on intensive care
Anaethesia
Cited by (251)
Off-Hours Intensive Care Unit Transfer Is Associated With Increased Mortality and Failure to Rescue
2023, Annals of Thoracic SurgeryStochastic optimization approaches for elective surgery scheduling with downstream capacity constraints: Models, challenges, and opportunities
2022, Computers and Operations ResearchA distributionally robust optimization approach for stochastic elective surgery scheduling with limited intensive care unit capacity
2021, European Journal of Operational ResearchThe problem of the last bed: Contextualization and a new simulation framework for analyzing physician decisions
2020, Omega (United Kingdom)Associations between hospital occupancy, intensive care unit transfer delay and hospital mortality
2020, Journal of Critical CareCitation Excerpt :Extended ICU stay may also indirectly impact patient outcomes through other processes of care. For example, patients who experience delays in outward transfers from intensive care may have been discharged prematurely at periods of high ICU census [24], and often end up leaving the ICU after-hours and weekends, [13,14] — periods of increased vulnerability for ICU patients [25–29]. We hypothesized that higher hospital inpatient occupancy rates above a critical threshold, are associated with increasing transfer delays from intensive care and that intensive care patients with longer transfer delays will have higher rates of in-hospital death and unplanned readmissions to the ICU, than patients with shorter transfer delays.