The ED floor admit that later requires ICU care: what are we missing and how can we improve?

Author: Brit Long, MD (@long_brit, EM Chief Resident at SAUSHEC, USAF) // Edited by: Alex Koyfman, MD (@EMHighAK, EM Attending Physician, UT Southwestern Medical Center / Parkland Memorial Hospital) and Stephen Alerhand, MD (@SAlerhand)

A 72 year-old male presents to the ED via ambulance with respiratory distress and fever. Your assessment reveals a cachectic-appearing male with significant retractions. His vital signs are RR 27, HR 116, BP 142/82, and T 99.7 F. His saturations are holding at 92% on a NRB at 5L. His history is pertinent for worsening cough and fevers to 103 F at home over three days. Today, his wife noted the patient was working harder to breath. He has crackles in the right lung base, and his mucosa are dry. You continue supplemental oxygen and provide IV antibiotics and fluids. His chest x-ray reveals multilobar pneumonia. His VS improve to HR 92 and RR 23, and he appears to be breathing easier. He is admitted to the floor, but one hour later, a rapid response is called for confusion, hypoxia, and hypotension. The patient requires intubation with vasopressors and is ultimately taken to the intensive care unit (ICU).

Emergency physicians care for a variety of patients. The phrase “You remember that patient you saw…?” invokes trepidation and anxiety in every physician. One of the key aspects of emergency medicine is disposition. We know how to resuscitate and provide top of the line care to the patient who is dying. A great deal of time the disposition is admit versus discharge home. However, this decision it not always so black and white. Many times we know the patient will be admitted, but at times it can be difficult deciding whether the patient is appropriate for the floor, step down unit, or ICU.

Are there factors associated with clinical decompensation after admission? If there are, what can emergency physicians do to ensure proper patient disposition?


First, appropriate disposition is important, as decompensation on the floor is associated with worse outcomes. Goldhill found transferred patients (floor to ICU) experienced a mortality rate of 53% compared to 30% of direct admissions to an ICU from the ED.1 Clinical deterioration on the hospital floor is an independent predictor of mortality.2-4 Patients admitted to the ICU directly have lower mortality and shorter length of stay when compared with patients with unplanned transfer from the floor/ward to the ICU. In patients with septic shock diagnosed in the ward versus ICU, patients on the ward had higher mortality, greater time to receipt of IV fluid, and greater time to vasopressor administration.2,3

Factors associated with decompensation

Several studies have demonstrated that the conditions with greatest decompensation risk are respiratory diseases, particularly patients with tachypnea and hypoxia. Kennedy et al. in 2010 evaluated 5300 patients, focusing on those patients who required transfer to ICU from a floor. One in three patients had pneumonia, and ultimately half required intubation! Predictors of decompensation included respiratory compromise, CHF, vascular disease, hypotension, tachycardia, or Cr > 2.0. These factors are shown in Figure 1 from the Kennedy study.5

floor to ICU 1

Figure 1 – Risk factors for ICU transfer from Kennedy et al.

Renaud et al. evaluated patients with community acquired pneumonia and found tachycardia, hypoxia, and tachypnea to be associated with need for ICU transfer, as well as the presence of one comorbidity such as CHF, COPD, and renal disease. Other factors with early ICU transfer were male gender, multilobar infiltrates, pleural effusion, acidosis, elevated BUN, and hyponatremia.6

In a Kaiser database of 178,000 patients with non-ICU admissions, 3% required transfer to an ICU, and the most common diagnoses associated with need for transfer were pneumonia, COPD, MI, or sepsis. More than half of these transfers occurred within the initial 24 hours of hospital admission, and this study found that mortality increased with unplanned transfer to the ICU.2 Predictors of unplanned ICU admission included older age, male gender, higher triage category, and history of comorbidities. Diagnoses including sepsis, renal failure, new cancer diagnosis, pneumonia, bowel obstruction, and COPD were associated with increased ICU admissions as well.2,7

An abstract evaluating sepsis patients found that albumin < 3.0 (OR 6), temperature > 38 C (OR 7.9), and an APACHE II score > 9 (OR 6.4) predicted clinical decompensation and ICU transfer. Interestingly, only one of 29 patients in the study population received a lactate measurement.8

What can emergency providers do?

Is there a way to decrease this risk of unplanned ICU transfer? The most important factor is knowing what risk factors are associated with decompensation and need for transfer.2 Emergency physicians provide great care, and patients will often improve with your resuscitation measures in the ED. The patient who was initially on the cusp of intubation or death may turn around quickly with measures including IV fluids, antibiotics, pain treatment, and/or noninvasive positive pressure ventilation. You might be thinking that because the patient looks so good now, they will look just as good several hours later.

If admitting the patient to a normal floor or ward, be sure to re-evaluate the patient’s vital signs and clinical appearance. Go through the laboratory and imaging results and look for those risk factors discussed above. If something is bothering you about the patient, speak with the admitting physician about your concerns. Closely look at the vital signs to ensure they truly are normal. Be wary of the patient who experienced a transient drop in blood pressure but then returns to normal, which is associated with increased risk of death. A recent 2015 study evaluated patients presenting with non-traumatic hypotension, and this study showed an in-hospital mortality of 12% for hypotension in the ED and an in-hospital mortality of 33-52% for pre-hospital shock!9


Fortunately, several studies have evaluated decision tools for disposition. In patients with pneumonia, one study evaluated the Risk of Early Admission to ICU index (REA-ICU index), which comprises 11 criteria independently associated with ICU admission: male gender, age younger than 80 years, comorbid conditions, respiratory rate 30, heart rate > 125bpm, multilobar infiltrate or pleural effusion, WBC < 3 or > 20K, hypoxemia (oxygen saturation < 90% or arterial partial pressure of oxygen (PaO2) < 60 mmHg), BUN > 11 mmol/L, pH < 7.35 and sodium < 130 mEq/L. The REA-ICU index stratified patients into four risk classes with a risk of ICU admission on days 1 to 3 ranging from 0.7% to 31%. The figure below demonstrates their risk classification.6

 floor to ICU 2

Figure 2 – REA-ICU index stratification

For pneumonia, another score is SMART-COP, which is a score for predicting need for respiratory or vasopressor support and ICU admission. The score consists of low systolic blood pressure (2 points), multilobar chest radiography involvement (1 point), low albumin level (1 point), high respiratory rate (1 point), tachycardia (1 point), confusion (1 point), poor oxygenation (2 points), and low arterial pH (2 points). A score of > 3 points predicts need for higher level of care.2,10


When caring for the patient with diagnosed or suspected sepsis, the presence of three or more of the following warrants ICU admission: Cr > 2.0, Plt < 100,000, Bili > 2.0, INR > 1.5, PaO2/FiO2 < 250, AKI, blood pressure < 90/60, MAP < 65, lactate > 2.0, cardiac dysfunction, and presence of DIC. This comes down to the involvement of multiple organ systems and potential for development of shock.2

Other scoring systems include the Early Warning Score (EWS) and modified Early Warning Score (MEWS), which are physiologic scoring systems designed to identify patients at risk for deterioration. The MEWS scoring system is shown below in Figure 3. Scores of five and greater in one study demonstrated an increased risk of death (OR 5.4), ICU admission (OR 10.9), and step down unit admission (OR 3.3). The authors advocate for this score due to its ease of use for nurses and physicians.11 A Cochrane review in 2007 evaluated the use of these scores and found significant study heterogeneity, calling for multi-site randomized clinical trials. The trials included either showed no difference in mortality or a decrease in mortality with score use. 12

 floor to ICU 3

Figure 3 – MEWS scoring system from Subbe et al.

If the patient decompensates, immediate transfer should be completed, as delays are associated with increased patient mortality. Knott et al. found that patients who developed shock from sepsis had a mortality rate of 39%, compared to 70% of those who developed shock while on the ward.13 This speaks to the importance of prompt recognition of shock and initiation of resuscitation measures. Young et al. found for patients who decompensated on the floor, a longer time from onset of instability to ICU admission was a predictor of mortality.14

Ultimately, the patient requires evaluation for risk factors and multi-organ involvement of the disease process. The question of correct disposition often comes down to clinical gestalt. High-risk conditions like respiratory diseases and sepsis warrant close evaluation for ICU admission. The presence of multiple vital sign abnormalities and comorbidities suggests poor physiologic reserve and potential for decline. Utilizing scoring systems can assist providers, but they should not substitute for the clinician at the bedside. Vital sign abnormalities, respiratory conditions, and comorbidities should be closely evaluated. This can decrease the risk of patient decompensation and need for ICU transfer.

References/Further Reading:

  1. Goldhill DR, Sumner A. Outcome of intensive care patients in a group of British intensive care units. Crit Care Med. 1998 Aug;26(8):1337-45.
  2. Perkins J, Mattu A. Predictors of Clinical Decompensation on the Floor – EMCast. 2015 Feb. Accessed 20 Oct 2015.
  3. Lundberg JS, Perl TM, Wiblin T, Costigan MD, Dawson J, Nettleman MD, Wenzel RP Septic shock: an analysis of outcomes for patients with onset on hospital wards versus intensive care units. Crit Care Med. 1998 Jun; 26(6):1020-4.
  4. Simchen E, Sprung CL, Galai N, Zitser-Gurevich Y, Bar-Lavi Y, Levi L, Zveibil F, Mandel M, Mnatzaganian G, Goldschmidt N, Ekka-Zohar A, Weiss-Salz I . Survival of critically ill patients hospitalized in and out of intensive care. Crit Care Med. 2007 Feb; 35(2):449-57.
  5. Kennedy M, Joyce N, Howell MD, et al. Identifying Infected Emergency Department Patients Admitted to the Hospital Ward at Risk of Clinical Deterioration and Intensive Care Unit Transfer. Acad Emerg Med. 2010 October; 17(10): 1080–1085.
  6. Renaud B, Labarere J, Coma E, et al. Risk stratification of early admission to the intensive care unit of patients with no major criteria of severe community-acquired pneumonia: development of an international prediction rule. Crit Care. 2009; 13:R54.
  7. Frost SA, Alexandrou E, Bogdanovski T, Salamonson Y, Parr MJ, Hillman KM. Unplanned admission to intensive care after emergency hospitalisation: risk factors and development of a nomogram for individualising risk. Resuscitation. 2009 Feb;80(2):224-30.
  8. Drumheller BC, D’Amore JZ, Hussain E, Patel S. Emergency Department Sepsis Risk Factors Predictive of Clinical Decompensation and Intensive Care Unit Transfer Within 72 Hours of Hospital Admission. Ann Emerg Med 2007 Sep;50(3).
  9. Holler JG, Bech CN, Henriksen DP, Mikkelsen S, Pedersen C, Lassen AT Nontraumatic hypotension and shock in the emergency department and the prehospital setting, prevalence, etiology, and mortality: a systematic review. PLoS One. 2015 Mar 19;10(3):e0119331.
  10. Charles PGP, Wolfe R, Whitby M, Fine MJ, et al. SMART-COP: A Tool for Predicting the Need for Intensive Respiratory or Vasopressor Support in Community-Acquired Pneumonia. Clin Infect Dis. (2008) 47 (3): 375-384.
  11. Young MP, Gooder VJ, McBride K, James B, Fisher ES. Inpatient transfers to the intensive care unit: delays are associated with increased mortality and morbidity. J Gen Intern Med. 2003 Feb; 18(2):77-83.

Leave a Reply