ReviewSeverity of illness and risk of readmission to intensive care: A meta-analysis*
Introduction
Despite more patients surviving intensive care, approximately 10% require readmission to the intensive care unit (ICU).1 It is not clear whether the decision to discharge patients from the ICU or the level of care given to these patients on the general wards or a combination results in readmission to ICU. This event, associated with adverse health events has been highlighted as a potential marker of quality of care.2 Therefore, the ability to identify patients at high risk of readmission to ICU during the same hospitalisation could allow objective decisions to be made by clinicians related to the timing of discharge from ICU, the level of care required by patients on the ward and the need for follow-up by ICU staff.
In an attempt to address this problem several authors have either collected specific data to identify risk factors for readmission to ICU, or have analyzed routinely collected data to identify predictors of readmission.3, 4, 5, 6, 7, 8, 9 Some papers have specifically addressed the ability of predictors to discriminate between patients who are readmitted to ICU and those who are not.4, 10 Importantly, a index of severity of illness (APACHE II)11 routinely collected on admission to the ICU is by itself as accurate as a more complex predictive model that also uses characteristics of the ICU stay, such as length of stay and mechanical ventilation days.4 While severity of illness is routinely calculated in patients admitted to ICU to predict in-hospital mortality in many settings throughout the world,12 the association between severity of illness and the risk of readmission to ICU has not been systematically summarized.
For this reason, a meta-analysis was designed to combine information from published studies to assess the relationship between level of severity of illness in ICU patients and the risk of readmission to ICU during the same hospitalisation.
Section snippets
Methods
This article was prepared in accordance with the Meta-analysis of Observational Studies in Epidemiology (MOOSE) statement.13 Briefly the MOOSE statement outlines a prospective and systematic method for reporting meta-analyses of observational studies including considering the background, search strategy, methods, results, discussion, and conclusion. The rationale for choosing this method is that observational studies potentially reflect the ‘usual care’ of contemporary clinical practice, in
Results
We identified a total of 195 potential studies to be included in the meta-analysis. After excluding review papers, including only adult populations, readmissions to ICU during the same hospitalisation, and studies that reported a valid severity of illness score: 11 studies were retained for analysis. The characteristics of these studies are presented in Table 1. Year of publication ranged from 1993 to 2008. The paper by Chen at al.5 separated results based on teaching or community hospitals.
Discussion
This meta-analysis has shown a relationship between increasing intensive care severity of illness score and risk of readmission to ICU in patients who survive ICU and are discharged to the ward. The effect was consistent, regardless of the timing of measurement of severity of illness (at admission to the ICU or at the time of discharge from ICU). For each standard deviation increase in severity of illness score the risk of readmission to ICU during the same hospitalisation increased by 43%.
Conflict of interest
No potential conflict of interest declared by any of the authors.
References (30)
- et al.
Patients readmitted to ICUs: a systematic review of risk factors and outcomes
Chest
(2000) - et al.
Predicting death and readmission after intensive care discharge
Br J Anaesth
(2008) - et al.
The effect of comorbidities on risk of intensive care readmission during the same hospitalisation: a linked data cohort study
J Crit Care
(2009) - et al.
Reliability of a measure of severity of illness: acute physiology of chronic health evaluation—II
J Clin Epidemiol
(1992) - et al.
Meta-analysis in clinical trials
Control Clin Trials
(1986) - et al.
The APACHE III prognostic system Risk prediction of hospital mortality for critically ill hospitalized adults
Chest
(1991) - et al.
Long-term functional outcome and performance status after intensive care unit re-admission: a prospective survey
Br J Anaesth
(2008) - et al.
Can early re-admission rates accurately detect poor-quality hospitals?
Med Care
(1995) - et al.
Readmission to surgical intensive care increases severity-adjusted patient mortality
J Trauma
(2006) - et al.
Patients readmitted to the intensive care unit during the same hospitalisation: clinical features and outcomes
Crit Care Med
(1998)
A case–control analysis of readmissions to the cardiac surgical intensive care unit
Eur J Cardiothorac Surg
Critically ill patients readmitted to intensive care units—lessons to learn?
Intensive Care Med
Who bounces back? Physiologic and other predictors of intensive care unit readmission
Crit Care Med
The Stability and Workload Index for transfer score predicts unplanned intensive care unit patient readmission: initial development and validation
Crit Care Med
The use of severity scores in the intensive care unit
Intensive Care Med
Cited by (81)
Risk factors for readmission to ICU and analysis of intra-hospital mortality
2022, Medicina ClinicaThe ISTH DIC score predicts outcome in non-septic patients admitted to a cardiovascular intensive care unit
2020, European Journal of Internal MedicineRisk prediction models for intensive care unit readmission: A systematic review of methodology and applicability
2020, Australian Critical CareCitation Excerpt :Hence, there has been a high interest in the ICU readmission rate as a quality indicator of critical care.1,8 Nevertheless, current studies have shown that ICU readmission rates are influenced by factors other than quality of care, such as patient characteristics and length of stay.1,9,10 Furthermore, inconsistent results have been reported about the negative impact of ICU readmissions on patient outcomes.
Prospective validation and refinement of the APPROACH cardiovascular surgical intensive care unit readmission score
2019, Journal of Critical CareCitation Excerpt :The inclusion of respiratory variables, including percent 02 requirements at discharge, tracheostomy, or re-intubation during initial CVICU stay had little impact on model performance. In contrast, the Stability and Workload Index for Transfer (SWIFT) score which is used to assess readmission risk after discharge from general intensive care unit found Pa02/Fi02 ratio was utilized in the final model [20,21]. Similarly, a meta-analysis found that SAPS 2 score, which incorporates the lowest Pa02/Fi02 ratio in a 24-hour period before discharge, was a predictor of general intensive care readmission [20].
- *
A Spanish translated version of the summary of this article appears as Appendix in the final online version at doi:10.1016/j.resuscitation.2009.02.015.