Institutional members access full text with Ovid®

Do Intensivist Staffing Patterns Influence Hospital Mortality Following ICU Admission? A Systematic Review and Meta-Analyses*

Wilcox, M. Elizabeth MD, MSc1; Chong, Christopher A. K. Y. MD2,3; Niven, Daniel J. MD, MSc4; Rubenfeld, Gordon D. MD, MSc5; Rowan, Kathryn M. DPhil6; Wunsch, Hannah MD, MSc7,8; Fan, Eddy MD1

doi: 10.1097/CCM.0b013e318292313a
Feature Articles

Objective: To determine the effect of different intensivist staffing models on clinical outcomes for critically ill patients.

Data Sources: A sensitive search of electronic databases and hand-search of major critical care journals and conference proceedings was completed in October 2012.

Study Selection: Comparative observational studies examining intensivist staffing patterns and reporting hospital or ICU mortality were included.

Data Extraction: Of 16,774 citations, 52 studies met the inclusion criteria. We used random-effects meta-analytic models unadjusted for case-mix or cluster effects and quantified between-study heterogeneity using I2. Study quality was assessed using the Newcastle-Ottawa Score for cohort studies.

Data Synthesis: High-intensity staffing (i.e., transfer of care to an intensivist-led team or mandatory consultation of an intensivist), compared to low-intensity staffing, was associated with lower hospital mortality (risk ratio, 0.83; 95% CI, 0.70–0.99) and ICU mortality (pooled risk ratio, 0.81; 95% CI, 0.68–0.96). Significant reductions in hospital and ICU length of stay were seen (–0.17 d, 95% CI, –0.31 to –0.03 d and –0.38 d, 95% CI, –0.55 to –0.20 d, respectively). Within high-intensity staffing models, 24-hour in-hospital intensivist coverage, compared to daytime only coverage, did not improved hospital or ICU mortality (risk ratio, 0.97; 95% CI, 0.89–1.1 and risk ratio, 0.88; 95% CI, 0.70–1.1). The benefit of high-intensity staffing was concentrated in surgical (risk ratio, 0.84; 95% CI, 0.44–1.6) and combined medical-surgical (risk ratio, 0.76; 95% CI, 0.66–0.83) ICUs, as compared to medical (risk ratio, 1.1; 95% CI, 0.83–1.5) ICUs. The effect on hospital mortality varied throughout different decades; pooled risk ratios were 0.74 (95% CI, 0.63–0.87) from 1980 to 1989, 0.96 (95% CI, 0.69–1.3) from 1990 to 1999, 0.70 (95% CI, 0.54–0.90) from 2000 to 2009, and 1.2 (95% CI, 0.84–1.8) from 2010 to 2012. These findings were similar for ICU mortality.

Conclusions: High-intensity staffing is associated with reduced ICU and hospital mortality. Within a high-intensity model, 24-hour in-hospital intensivist coverage did not reduce hospital, or ICU, mortality. Benefits seen in mortality were dependent on the type of ICU and decade of publication.

1Interdepartmental Division of Critical Care Medicine, University of Toronto, University Health Network and Mount Sinai Hospital, Toronto, ON, Canada.

2Section of General Internal Medicine, Lakeridge Health Oshawa, Oshawa, ON, Canada.

3Faculty of Health Sciences, Queen’s University, Kingston, ON, Canada.

4Department of Critical Care Medicine, University of Calgary, Calgary, AB, Canada.

5Interdepartmental Division of Critical Care Medicine, University of Toronto, Sunnybrook Health Sciences Centre, Toronto, ON, Canada.

6Intensive Care National Audit and Research Centre, London, United Kingdom.

7Department of Anesthesiology, College of Physicians and Surgeons, Columbia University, New York, NY.

8Department of Epidemiology, Mailman School of Public Health, Columbia University, New York, NY.

* See also p. 2433.

Dr. Wilcox conceived the study, searched the literature, selected studies for inclusion, abstracted data, analyzed data, wrote the first draft of the manuscript, and revised the manuscript. Drs. Chong and Fan selected studies for inclusion, abstracted data, verified analyses, and revised the manuscript. Drs. Niven and Wunsch abstracted data and revised the manuscript. Drs. Rubenfeld and Rowan provided feedback on study design and revised the manuscript. All authors approved the final manuscript.

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s website (

The authors have disclosed that they do not have any potential conflicts of interest.

For information regarding this article, E-mail:

© 2013 by the Society of Critical Care Medicine and Lippincott Williams & Wilkins