Background Intensive care unit (ICU) beds, a scarce resource, may require prioritization of admissions when demand exceeds supply. We evaluated the effect of ICU bed availability on processes and outcomes of care for hospitalized patients with sudden clinical deterioration.
Methods We identified consecutive hospitalized adults in Calgary, Alberta, Canada, with sudden clinical deterioration triggering medical emergency team activation between January 1, 2007, and December 31, 2009. We compared ICU admission rates (within 2 hours of medical emergency team activation), patient goals of care (resuscitative, medical, and comfort), and hospital mortality according to the number of ICU beds available (0, 1, 2, or >2), adjusting for patient, physician, and hospital characteristics (using data from clinical and administrative databases).
Results The cohort consisted of 3494 patients. Reduced ICU bed availability was associated with a decreased likelihood of patient admission within 2 hours of medical emergency team activation (P = .03) and with an increased likelihood of change in patient goals of care (P < .01). Patients with sudden clinical deterioration when zero ICU beds were available were 33.0% (95% CI, −5.1% to 57.3%) less likely to be admitted to the ICU and 89.6% (95% CI, 24.9% to 188.0%) more likely to have their goals of care changed compared with when more than 2 ICU beds were available. Hospital mortality did not vary significantly by ICU bed availability (P = .82).
Conclusion Among hospitalized patients with sudden clinical deterioration, we noted a significant association between the number of ICU beds available and ICU admission and patient goals of care but not hospital mortality.
The Institute of Medicine's report To Err Is Human1 raised awareness about the high risk of medical errors, adverse events, and preventable death among hospitalized vulnerable populations with sudden clinical deterioration. In response, strategies to identify, evaluate, and quickly treat these patients using rapid response systems have been prioritized by prominent health care organizations.2,3
Intensive care unit (ICU) beds are an integral structural component of rapid response systems and provide life-saving support for critically ill patients.4 However, ICU beds are a scarce resource for which demand periodically exceeds supply.5 Aging populations are expected to steadily increase demand for critical care services,6,7 raising concerns about mechanisms for resource allocation when demand exceeds supply.5
Physicians routinely make decisions in the setting of resource constraints, despite limited training in resource allocation.5,8 Available evidence suggests that notions of equity often guide physician care9,10 and that patients who are perceived to be unlikely to benefit from critical care are more often declined ICU admission.5 Although the tenets of biomedical ethics and international law indicate that protocols should be used to guide resource allocation when demand for ICU resources exceeds supply,11,12 formal triage protocols are not routinely part of rapid response systems or decision making surrounding patient ICU admission.8 For these reasons, we undertook a study to examine the association between ICU bed availability (0, 1, 2, or >2 beds) and the processes and outcomes of care for a population-based cohort of hospitalized patients with sudden clinical deterioration.
We identified consecutive hospitalized adults (excluding cardiac surgery and coronary care units) in Calgary, Alberta, Canada (with a 1.5-million catchment population) having sudden clinical deterioration identified by the rapid response system and triggering medical emergency team (MET) activation between January 1, 2007, and December 31, 2009. A rapid response system is a hospital-based capability to detect patients with sudden clinical deterioration and mechanism to quickly respond to patients' needs, while a MET is a group of clinicians activated by the rapid response system.4 Alberta Health Services manages all METs (each composed of 1 ICU physician, nurse, and respiratory therapist 24 h/d and 7 d/wk) and medical-surgical (general system) ICUs (53 funded beds) in Calgary (3 hospitals, with a total of 2040 beds, each with a MET and an ICU). The ICUs are closed units staffed by fully trained intensivists. To trigger MET activation, the Alberta Health Services rapid response system used the following physiologically based criteria13: (1) respiratory (threatened airway, respiratory rate of <8 or >30 breaths/min, or acute change in oxygen saturation of <90%, despite oxygen >5 L/min), (2) heart rate (pulse rate of <40 or >140 beats/min), (3) blood pressure (systolic blood pressure of <90 mm Hg), (4) neurological (sudden decrease in Glasgow Coma Scale score of ≥2 points or prolonged or repeated seizures), and (5) worried (“any patient that you are seriously worried about”). Decisions to admit patients to the ICU are made by the attending physician on a case-by-case basis, without a triage protocol or decision-making support. Hospital wards activating the MET are not generally aware of ICU bed availability.
We used data from Alberta Health Services clinical and administrative databases. The MET database captures reason for assessment, vital signs, diagnostic and therapeutic interventions, and patient disposition, with data acquired at the time of patient assessment. The ICU database is an electronic patient information system that captures demographic, clinical, and outcome data for all patients admitted to the ICU (excluding coronary care units). Alberta Health Services administrative databases capture data on all hospitalized patients, including vital status at discharge, dates of admission and discharge, up to 25 Canadian Enhancement of International Statistical Classification of Diseases, 10th Revision, diagnostic codes, and up to 20 Canadian Classification of Health Interventions procedure codes. Hospital electronic medical records are used by the study hospitals and contain detailed patient information, including physician orders and patient goals of care. These 4 databases are used for clinical care and administrative data capture and have been used for program evaluation and research.14
Patient, physician, and hospital factors
We identified factors that may affect the likelihood of ICU admission after a MET activation. Hospital factors included the number of ICU beds available (0, 1, 2, or >2) and the day of the week (weekday [Monday 8 AM to Friday 5 PM] vs weekend) and time of day (daytime [8 AM to 5 PM] vs nighttime) of the MET activation.15 Available ICU beds were defined as the number of beds (funded to be operational) without an assigned patient at the time of MET activation. Physician factors included the type of physician attending the MET activation (ICU attending physician, ICU fellow, resident, or ICU physician extender [defined as a licensed nonintensivist physician providing in-hospital support to the ICU attending physician]). Patient factors included demographic variables, socioeconomic status, race/ethnicity, comorbidities, reason for hospital admission (medical, surgical, or neurological), and baseline patient goals of care (resuscitative [life-sustaining care that includes ICU admission], medical [life-sustaining care that does not include ICU admission], and comfort [care focused on symptom control]). Socioeconomic status was defined using the median neighborhood household income quintile.16 The presence of comorbidities was derived using the Deyo classification of Charlson comorbidities and validated Canadian Enhancement of International Statistical Classification of Diseases, 10th Revision, coding algorithms (summarized as a single comorbidity score for multivariable analyses).17 First Nations populations were defined using the registry file from Alberta Health Services administrative databases.18 Chinese race/ethnicity was determined using the Chinese surname list algorithm,19 and South Asian race/ethnicity was defined using the Nam Pehchan classification program.20
Process and outcome measures
The primary outcome was ICU admission within 2 hours of MET activation,21 identified using the MET and ICU databases. We chose ICU admission within 2 hours because local policy is to decide on patient disposition within 30 minutes of MET arrival. An additional 90 minutes was provided to enable ICU transfer. We examined the following 4 secondary outcome measures: (1) change in patient goals of care (resuscitative, medical, or comfort) within 24 hours of MET activation; (2) hospital mortality; (3) health care resource use, evaluated by investigations and interventions performed during the initial MET activation or a new MET activation (including the use of cardiopulmonary resuscitation); and (4) ICU admission during the remainder of the hospitalization (ie, >2 hours following the initial MET activation).
The primary analysis tested associations between the number of ICU beds available (0, 1, 2, or >2) and ICU admission within 2 hours of MET activation using a generalized estimating equations model. The unit of analysis was the initial MET activation during the hospitalization (subsequent MET activations were evaluated as an outcome). We selected a generalized estimating equations model because it is an extension of standard logistic regression analysis that adjusts for correlation among observations (ie, patients with >1 hospitalization having a MET activation during the study period or patients clustered within hospitals)22,23 and provides population average effect (average response for observations sharing the same covariates), which has familiar interpretation and is more useful when estimating effects at a population level.24,25 We adjusted for patient, physician, and hospital covariates (all baseline variables measured) when analyzing outcome measures. Health care resource use was tested across the 4 defined levels of ICU bed availability using Pearson product moment correlation χ2 test for differences in proportions and Kruskal-Wallis test for differences in medians. Sensitivity analyses were performed to explore the effect on the primary outcome of study hospital, time of MET activation, physician attending MET activation, and patient goals of care. Statistical analyses were performed using commercially available software (SAS version 9.2; SAS Institute, Inc), and 2-sided P < .05 was considered significant. The Conjoint Health Research Ethics Board at the University of Calgary approved this study and waived the need for informed consent from patients and physicians.
The cohort consisted of 3494 patients (3.2 patients per day with sudden clinical deterioration and MET activation). The median age was 72 years (interquartile range [IQR], 57-81 years), 46.8% were female, 46.2% had 1 or more comorbidities, and 10.3% had a prior ICU admission during their hospital stay. Fewer MET activations occurred when zero ICU beds (249 [7.1%]) or 1 ICU bed (470 [13.5%]) was available. Patient characteristics by ICU bed availability were similar, with a few notable exceptions (Table 1). MET activations at night and for respiratory reasons were more common when more ICU beds were available. Residents were more likely to be the physician attending MET activation when fewer ICU beds were available.
We examined MET processes of care according to the number of ICU beds available (Table 2). Although the duration of MET activations was similar across groups (median, 53 minutes; IQR, 33-80 minutes), there were significant positive associations between the number of ICU beds available and the use of investigations and interventions. Patients were more likely to have arterial blood gases measured and radiographs obtained of the chest or abdomen when more ICU beds were available (P < .01 for both). Noninvasive airway interventions, including airway suctioning (P < .01), oxygen (P = .08), and nebulized medications (P = .05), were more frequently provided when more ICU beds were available. In total during MET activations, 4.4% of patients were intubated and mechanically ventilated when zero ICU beds were available and 10.1% of patients when more than 2 ICU beds were available (P < .01).
Multivariable-adjusted analyses of patient outcomes of care are summarized in the Figure, Table 3, and Table 4. Hospitalized patients with sudden clinical deterioration were significantly more likely to be admitted to the ICU within 2 hours of MET activation when more ICU beds were available (P = .03) (Table 3). In total, 11.6% of patients evaluated by the MET when zero ICU beds were available and 21.4% of patients when more than 2 ICU beds were available were admitted to the ICU within 2 hours of MET activation. Among patients not admitted to the ICU within 2 hours of MET activation, subsequent ICU admissions were infrequent in all 4 groups (eTable). Few patients in all 4 groups (411 [11.8%]) experienced cardiac arrest or new MET activation following the initial patient MET activation. Length of ICU stay for patients admitted to the ICU within 2 hours of MET activation (median, 4 days; IQR, 2-8 days) and length of hospital stay after MET activation for patients discharged alive (median, 15 days; IQR, 7-35 days) were similar across the groups. More patients had their goals of care changed from resuscitative care to medical or comfort care when zero ICU beds were available (37 [14.9%]) compared with when 2 or more ICU beds were available (174 [8.5%]) (P < .01). Hospital mortality was similar across the 4 patients groups (P = .82), ranging from 32.1% to 34.7%.
The number of ICU beds available was associated with the likelihood of ICU admission within 2 hours of MET activation and with change in patient goals of care (Table 4 and Figure). Compared with when more than 2 ICU beds were available, ICU admission was 33.0% (95% CI, −5.1% to 57.3%) less likely when zero ICU beds were available and 32.4% (95% CI, 7.3% to 50.7%) less likely when 1 ICU bed was available. Conversely, change in patient goals of care from resuscitative care to medical or comfort care was 89.6% (95% CI, 24.9%-188.0%) more likely when zero ICU beds were available and 61.8% (95% CI, 15.5%-126.6%) more likely when 1 ICU bed was available. The adjusted odds ratios for hospital mortality were not significantly different across the 4 groups, although the 95% CI when zero ICU beds were available included a 25.6% increase in hospital mortality. Results were similar when analyses were stratified according to study hospital, time of MET activation, and physician attending MET activation or when restricted to patients with resuscitative goals of care at the time of MET activation.
In this population-based cohort study, we found a strong association between the number of ICU beds available and processes of care for hospitalized patients with sudden clinical deterioration. We noted a lower probability of patient ICU admission and a higher probability of change in patient goals of care from resuscitative care to medical or comfort care as the number of ICU beds available decreased. Hospital mortality was similar for patients regardless of the number of ICU beds available. This suggests that, for hospitalized patients with sudden clinical deterioration, the number of ICU beds available at the time of their deterioration affects processes of care but may not influence hospital mortality.
Our study provides important information on how ICU bed availability can affect processes of care for hospitalized patients with sudden clinical deterioration. Results of previous studies5,26-30 have demonstrated that as hospitals close ICU beds, patients admitted to the ICU are sicker, are less likely to be admitted for monitoring, and have shorter stays but with no adverse effects. Patient age, illness severity, and medical diagnosis were noted to be associated with decisions not to admit patients to the ICU.5,31-35 Our study adds to the literature in several ways. First, we demonstrate that the decision to admit a hospitalized patient with sudden clinical deterioration to the ICU is influenced by the number of ICU beds available. Second, rationing of ICU beds is a routine occurrence. We found that during a 3-year period, there was no ICU bed available at the time of MET activation for 7.1% of patients with sudden clinical deterioration. This is equivalent to 1 or 2 patients per week experiencing sudden clinical deterioration at a time when no ICU beds are available. Third, the decision to admit a hospitalized patient with sudden clinical deterioration to the ICU is complex and is influenced by patient, physician, and hospital factors. For example, our results suggest that ICU admission may be influenced by the type of physician attending MET activation (ICU attending, trainee, or ICU physician extender), despite ICU attending medical oversight of the rapid response system. Fourth, with limited training in resource allocation, it is unclear how effective physicians are at identifying patients who are most likely to benefit from ICU admission. In our study, there was no difference in hospital mortality among patients with MET activation according to the availability of ICU beds. Furthermore, among those patients who were not admitted to ICU, the risk of new MET activation and subsequent ICU admission was similar. These results suggest that available resources may have been efficiently allocated to meet patient needs. Conversely, findings from other studies36-38 have suggested that prognostication for critically ill patients is difficult and that physicians can be overconfident in their ability to identify patients with poor prognoses who are unlikely to benefit from ICU care.
Our study also underscores some important challenges for ensuring optimal allocation of ICU beds. First, results from our study may also suggest that ICU physicians are more likely to admit patients who will not benefit from critical care resources when ICU beds are available. Having different clinical thresholds for patient ICU admission based on physician or health system factors may result in misallocation of patients to hospital units and may compromise the quality and efficiency of health care delivery.8,39 Intensive care unit admission and discharge guidelines could optimize patient allocation.40,41 Second, mechanisms are required to increase the flexibility of critical care resources in systems that operate near capacity. METs are one mechanism for providing ICU outreach. Alternative strategies could include creating temporary ICU resources (strategies not now used by the study hospitals) by transferring hospitalized patients with sudden clinical deterioration to other spaces in the hospital that are suitable for advanced resuscitation and monitoring (eg, coronary care unit, postoperative care unit, or emergency department), while securing access to an ICU bed or transferring patients between ICUs. Third, mechanisms are needed to explicitly manage resource allocation when demand exceeds supply. Development and implementation of valid and reliable triage instruments would make evaluation and decision-making processes for critically ill patients transparent.42
As a result of the study design, we acknowledge the following limitations. First, the definition of our primary outcome (ICU admission within 2 hours of MET activation) is somewhat subjective because patients may be monitored on the ward and admitted after 2 hours if clinical conditions change. Nevertheless, this measure has been proposed as an indicator of ICU quality,21 and sensitivity analyses using different time thresholds for ICU admission (within 1, 3, 4, or 5 hours) demonstrated similar findings. Second, even with adjustment, residual confounding is an inherent risk for all observational studies, and it is possible that factors other than the number of ICU beds available might explain differences in ICU admission rates. However, most MET activations are triggered by hospital ward nurses who are unaware of ICU bed availability; as such, we do not anticipate the case mix of patients (specifically, unmeasured characteristics) triggering MET activations to change with the number of ICU beds available. Third, despite a study cohort of more than 3000 patients with detailed hospitalization (but not posthospitalization) data, we are unable to exclude small but clinically important differences in patient outcomes. Fourth, our study was performed in a single publicly funded health region. Allocation of ICU resources and decision-making processes for patient goals of care may vary across health care jurisdictions; therefore, the results may not apply to other institutions. However, the challenges of scarce ICU resources presented herein are likely common in other health organizations.5
In summary, for hospitalized patients, the number of ICU beds available at the time of sudden clinical deterioration affects processes of care. As the number of available ICU beds decreases, patients are less likely to be admitted to the ICU and are more likely to have their goals of care changed, although this does not seem to be associated with hospital mortality. This suggests that the development of validated ICU admission and discharge guidelines might improve hospital efficiency, without affecting outcomes.
Correspondence: Henry T. Stelfox, MD, PhD, FRCPC, Department of Critical Care Medicine, University of Calgary, Teaching, Research and Wellness Building, 3280 Hospital Dr NW, Calgary, AB T2N 4Z6, Canada (tstelfox@ucalgary.ca).
Accepted for Publication: December 22, 2011.
Published Online: March 12, 2012. doi:10.1001/archinternmed.2011.2315
Author Contributions: Dr Stelfox and Mr Gao had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Stelfox, Hemmelgarn, Bagshaw, Doig, Nijssen-Jordan, and Manns. Acquisition of data: Stelfox, Bagshaw, and Doig. Analysis and interpretation of data: Stelfox, Hemmelgarn, Bagshaw, Gao, Nijssen-Jordan, and Manns. Drafting of the manuscript: Stelfox, Hemmelgarn, Gao, and Nijssen-Jordan. Critical revision of the manuscript for important intellectual content: Stelfox, Hemmelgarn, Bagshaw, Doig, Nijssen-Jordan, and Manns. Statistical analysis: Bagshaw and Gao. Obtained funding: Stelfox, Hemmelgarn, Doig, and Manns. Administrative, technical, and material support: Stelfox, Bagshaw, Doig, and Nijssen-Jordan.
Financial Disclosure: None reported.
Funding/Support: This project was supported by operating grant H1N-104065 from the Canadian Institutes of Health Research. Dr Stelfox is supported by a New Investigator Award from the Canadian Institutes of Health Research. Drs Stelfox and Hemmelgarn are supported by Population Health Investigator awards from Alberta Innovates. Dr Bagshaw is supported by a Clinical Investigator Award from Alberta Innovates. Dr Manns is supported by a Scholar Award from Alberta Innovates.
Additional Contributions: David Megran, MD, gave comments on early versions of the article, and Alberta Health Services provided access to clinical and administrative databases.
1.Institute of Medicine. To Err Is Human: Building a Safer Health System. Washington, DC: National Academies Press; 1999
2.Berwick DM, Calkins DR, McCannon CJ, Hackbarth AD. The 100,000 lives campaign: setting a goal and a deadline for improving health care quality.
JAMA. 2006;295(3):324-32716418469
PubMedGoogle ScholarCrossref 3. The Joint Commission announces the 2006 National Patient Safety Goals and Requirements.
Jt Comm Perspect. 2005;25(7):1-1016629275
PubMedGoogle Scholar 4.DeVita ME, ed, Bellomo R, ed. Textbook of Rapid Response Systems: Concept and Implementation. New York, NY: Springer; 2011
5.Sinuff T, Kahnamoui K, Cook DJ, Luce JM, Levy MM.Values Ethics and Rationing in Critical Care Task Force. Rationing critical care beds: a systematic review.
Crit Care Med. 2004;32(7):1588-159715241106
PubMedGoogle ScholarCrossref 6.Angus DC, Kelley MA, Schmitz RJ, White A, Popovich J Jr.Committee on Manpower for Pulmonary and Critical Care Societies (COMPACCS). Caring for the critically ill patient: current and projected workforce requirements for care of the critically ill and patients with pulmonary disease: can we meet the requirements of an aging population?
JAMA. 2000;284(21):2762-277011105183
PubMedGoogle ScholarCrossref 7.Needham DM, Bronskill SE, Calinawan JR, Sibbald WJ, Pronovost PJ, Laupacis A. Projected incidence of mechanical ventilation in Ontario to 2026: preparing for the aging baby boomers.
Crit Care Med. 2005;33(3):574-57915753749
PubMedGoogle ScholarCrossref 9.Hurst SA, Forde R, Reiter-Theil S,
et al. Physicians' views on resource availability and equity in four European health care systems.
BMC Health Serv Res. 2007;7:13717764556
PubMedGoogle ScholarCrossref 10.Hurst SA, Reiter-Theil S, Slowther AM, Pegoraro R, Forde R, Danis M. Should ethics consultants help clinicians face scarcity in their practice?
J Med Ethics. 2008;34(4):241-24618375673
PubMedGoogle ScholarCrossref 12.Domres B, Koch M, Manger A, Becker HD. Ethics and triage.
Prehosp Disaster Med. 2001;16(1):53-5811367943
PubMedGoogle Scholar 13.Peberdy MA, Cretikos M, Abella BS,
et al; International Liaison Committee on Resuscitation; American Heart Association; Australian Resuscitation Council; European Resuscitation Council; Heart and Stroke Foundation of Canada; InterAmerican Heart Foundation; Resuscitation Council of Southern Africa; New Zealand Resuscitation Council; American Heart Association Emergency Cardiovascular Care Committee; American Heart Association Council on Cardiopulmonary, Perioperative, and Critical Care; Interdisciplinary Working Group on Quality of Care and Outcomes Research. Recommended guidelines for monitoring, reporting, and conducting research on medical emergency team, outreach, and rapid response systems: an Utstein-style scientific statement: a scientific statement from the International Liaison Committee on Resuscitation (American Heart Association, Australian Resuscitation Council, European Resuscitation Council, Heart and Stroke Foundation of Canada, InterAmerican Heart Foundation, Resuscitation Council of Southern Africa, and the New Zealand Resuscitation Council); the American Heart Association Emergency Cardiovascular Care Committee; the Council on Cardiopulmonary, Perioperative, and Critical Care; and the Interdisciplinary Working Group on Quality of Care and Outcomes Research.
Circulation. 2007;116(21):2481-250017993478
PubMedGoogle ScholarCrossref 15.Laupland KB, Shahpori R, Kirkpatrick AW, Stelfox HT. Hospital mortality among adults admitted to and discharged from intensive care on weekends and evenings.
J Crit Care. 2008;23(3):317-32418725035
PubMedGoogle ScholarCrossref 16.Alter DA, Naylor CD, Austin P, Tu JV. Effects of socioeconomic status on access to invasive cardiac procedures and on mortality after acute myocardial infarction.
N Engl J Med. 1999;341(18):1359-136710536129
PubMedGoogle ScholarCrossref 17.Quan H, Sundararajan V, Halfon P,
et al. Coding algorithms for defining comorbidities in
ICD-9-CM and
ICD-10 administrative data.
Med Care. 2005;43(11):1130-113916224307
PubMedGoogle ScholarCrossref 18.Gao S, Manns BJ, Culleton BF,
et al; Alberta Kidney Disease Network. Access to health care among status aboriginal people with chronic kidney disease.
CMAJ. 2008;179(10):1007-101218981441
PubMedGoogle ScholarCrossref 19.Quan H, Wang F, Schopflocher D,
et al. Development and validation of a surname list to define Chinese ethnicity.
Med Care. 2006;44(4):328-33316565633
PubMedGoogle ScholarCrossref 20.Cummins C, Winter H, Cheng KK, Maric R, Silcocks P, Varghese C. An assessment of the Nam Pehchan computer program for the identification of names of South Asian ethnic origin.
J Public Health Med. 1999;21(4):401-40611469361
PubMedGoogle ScholarCrossref 21.Berenholtz SM, Dorman T, Ngo K, Pronovost PJ. Qualitative review of intensive care unit quality indicators.
J Crit Care. 2002;17(1):1-1212040543
PubMedGoogle ScholarCrossref 23.Liang KY, Zeger SL. Longitudinal data analysis using generalized linear models.
Biometrika. 1986;73(1):13-22
Google ScholarCrossref 24.Carlin JB, Wolfe R, Brown CH, Gelman A. A case study on the choice, interpretation and checking of multilevel models for longitudinal binary outcomes.
Biostatistics. 2001;2(4):397-41612933632
PubMedGoogle ScholarCrossref 25.Hu FB, Goldberg J, Hedeker D, Flay BR, Pentz MA. Comparison of population-averaged and subject-specific approaches for analyzing repeated binary outcomes.
Am J Epidemiol. 1998;147(7):694-7039554609
PubMedGoogle ScholarCrossref 26.Byrick RJ, Mazer CD, Caskennette GM. Closure of an intermediate care unit: impact on critical care utilization.
Chest. 1993;104(3):876-8818365304
PubMedGoogle ScholarCrossref 27.Singer DE, Carr PL, Mulley AG, Thibault GE. Rationing intensive care—physician responses to a resource shortage.
N Engl J Med. 1983;309(19):1155-11606413862
PubMedGoogle ScholarCrossref 28.Walther SM, Jonasson U. A prospective cohort study of 6-month mortality in a community hospital experiencing a gradual reduction in critical care services.
Intensive Care Med. 2001;27(4):700-70511398696
PubMedGoogle ScholarCrossref 29.Azoulay E, Pochard F, Chevret S,
et al; PROTOCETIC Group. Compliance with triage to intensive care recommendations.
Crit Care Med. 2001;29(11):2132-213611700409
PubMedGoogle ScholarCrossref 30.Strauss MJ, LoGerfo JP, Yeltatzie JA, Temkin N, Hudson LD. Rationing of intensive care unit services: an everyday occurrence.
JAMA. 1986;255(9):1143-11463945032
PubMedGoogle ScholarCrossref 31.Sprung CL, Geber D, Eidelman LA,
et al. Evaluation of triage decisions for intensive care admission.
Crit Care Med. 1999;27(6):1073-107910397207
PubMedGoogle ScholarCrossref 32.Marshall MF, Schwenzer KJ, Orsina M, Fletcher JC, Durbin CG Jr. Influence of political power, medical provincialism, and economic incentives on the rationing of surgical intensive care unit beds.
Crit Care Med. 1992;20(3):387-3941541100
PubMedGoogle ScholarCrossref 33.Frisho-Lima P, Gurman G, Schapira A, Porath A. Rationing critical care: what happens to patients who are not admitted?
Theor Surg. 1994;9(4):208-21111660442
PubMedGoogle Scholar 34.Joynt GM, Gomersall CD, Tan P, Lee A, Cheng CA, Wong EL. Prospective evaluation of patients refused admission to an intensive care unit: triage, futility and outcome.
Intensive Care Med. 2001;27(9):1459-146511685338
PubMedGoogle ScholarCrossref 35.Metcalfe MA, Sloggett A, McPherson K. Mortality among appropriately referred patients refused admission to intensive-care units.
Lancet. 1997;350(9070):7-119217712
PubMedGoogle ScholarCrossref 36.Kruse JA, Thill-Baharozian MC, Carlson RW. Comparison of clinical assessment with APACHE II for predicting mortality risk in patients admitted to a medical intensive care unit.
JAMA. 1988;260(12):1739-17423137374
PubMedGoogle ScholarCrossref 37.Gusmão Vicente F, Polito Lomar F, Mélot C, Vincent JL. Can the experienced ICU physician predict ICU length of stay and outcome better than less experienced colleagues?
Intensive Care Med. 2004;30(4):655-65914735235
PubMedGoogle ScholarCrossref 39.Litvak E, ed. Managing Patient Flow in Hospitals: Strategies and Solutions. 2nd ed. Oak Brook, IL: Joint Commission Resources; 2009
40.Task Force of the American College of Critical Care Medicine, Society of Critical Care Medicine. Guidelines for intensive care unit admission, discharge, and triage.
Crit Care Med. 1999;27(3):633-63810199547
PubMedGoogle ScholarCrossref 41.Society of Critical Care Medicine Ethics Committee. Consensus statement on the triage of critically ill patients.
JAMA. 1994;271(15):1200-12037818629
PubMedGoogle ScholarCrossref 42.Shahpori R, Stelfox HT, Doig CJ, Boiteau PJ, Zygun DA. Sequential organ failure assessment in H1N1 pandemic planning.
Crit Care Med. 2011;39(4):827-83221263327
PubMedGoogle ScholarCrossref