Hospitals are dangerous places

Adrian G Barnett, Nicholas Graves, Ben S Cooper, Rahul Batra and Jonathan D Edgeworth
Med J Aust 2008; 189 (11): 672. || doi: 10.5694/j.1326-5377.2008.tb02239.x
Published online: 1 December 2008

Length of stay in the intensive care unit (ICU) is a complex variable with many contributing factors, including primary diagnosis, comorbidities and adverse events such as health care-acquired infection. The risk of an adverse event increases on weekends when staff availability is lower. We aimed to assess the effect of day of the week on patients’ odds of being discharged alive from the ICU, while controlling for other determinants of length of stay.


We studied 4569 patients admitted to the ICU of St Thomas’ Hospital, London, from 2002 to 2006. For each patient, we had the date of admission to and discharge from the ICU. We modelled the probability of discharge using a longitudinal survival model.1 Patients who died were censored. We modelled day of the week as a categorical variable, with Wednesday as the reference category. The model controlled for age, APACHE II (Acute Physiology and Chronic Health Evaluation) score, whether the patient was being filtrated or ventilated, and presence of health care-acquired infection. Software from the BUGS (Bayesian inference Using Gibbs Sampling) Project2 was used for the analysis.


The odds of being discharged by day of the week are shown in the Box. The odds of being discharged alive were significantly lower on Sunday and significantly higher on Friday (the 95% posterior interval for these 2 days does not include 1).


Our results show a frightening pattern of discharge from an ICU ward. The causes are most likely a complex web of specialist availability and patient requests to be home for the weekend. During the week, discharges run as smooth as silk, but things spin out of control at the start of the weekend. This problem has crept undetected into the St Thomas’ Hospital system; removing it may require sweeping changes to staff rosters.

Odds of being discharged alive from intensive care unit, by day of week

Black line represents the mean; red and green lines show the 95% posterior interval. The reference day of Wednesday is represented by a large dot.

  • Adrian G Barnett1
  • Nicholas Graves1
  • Ben S Cooper2
  • Rahul Batra3
  • Jonathan D Edgeworth3

  • 1 Institute of Health and Biomedical Innovation, Queensland University of Technology, Brisbane, QLD.
  • 2 Statistics, Modelling, and Bioinformatics Department, Centre for Infections, Health Protection Agency, London, UK.
  • 3 Guy’s and St Thomas’ National Health Service Foundation Trust, London, UK.


Competing interests:

None identified.

  • 1. Hedeker D, Gibbons RD. Longitudinal data analysis. Hoboken, NJ: Wiley, 2006.
  • 2. Spiegelhalter DJ, Thomas A, Best NG, et al. WinBUGS version 1.4.3 user manual. Cambridge: MRC Biostatistics Unit, 2007.


remove_circle_outline Delete Author
add_circle_outline Add Author

Do you have any competing interests to declare? *

I/we agree to assign copyright to the Medical Journal of Australia and agree to the Conditions of publication *
I/we agree to the Terms of use of the Medical Journal of Australia *
Email me when people comment on this article

Online responses are no longer available. Please refer to our instructions for authors page for more information.