Acute Care Research

Project Director/Lead Investigator:  Mitchelle Cohen, MD

 The Use of Tissue Oxygen Monitoring in Critically Injured Patients


Brief Summary of Project:   Our assessment of resuscitation status in trauma patients is limited by a lack of precise markers that reflect perfusion of critical organ beds. The purpose of this prospective observational study of severely injured polytrauma patients is to investigate tissue oxygen tension and microcirculatory flow as markers for resuscitation status.  The study methodology is based on our previous laboratory and clinical work measuring brain and muscle tissue oxygen during hemorrhagic shock and resuscitation. Data collected from the initial pilot study will be used to develop hypotheses and interventional studies using tissue oxygen tension and microcirculatory flow as guides for resuscitation of trauma patients. nbsp;

Studies and Results:   In our sample-population analysis, we concluded: 1) that PmO2 correlates with base deficit and offers a minimally invasive, continuous guide for resuscitation, 2) that initial low values for either PmO2 or StO2 are associated with post-injury complications, 3) that PmO2 can be used to identify patients in the state of occult under-resuscitation who remain at risk for developing infection and Multiple Organ Failure (MOF), and 4) that the variability in deltoid StO2 readings make this reading impractical for use in the intensive care unit.

The data from 9 patients has been analyzed and an abstract describing determinants of microvascular ow (FLOW) in relation to peripheral muscle oxygen (PmO2), was submitted for presentation at the . The presentation was well received and the associated manuscript was submitted to the Journal of Trauma. We hypothesized that low flow would correlate with low PmO2 in under resuscitated patients. As with our prior findings, PmO2 was higher in resuscitated patients. Interestingly, FLOW was significantly decreased in resuscitated patients. Stepwise regression indicates that FLOW increases with inspired oxygen, phenylephrine dose, and presence of head injury and is decreased with higher MAP and PmO2 (all p<0.05).        

We continue to advance our unique research into the use of bioinformatic clustering analysis. We recently published an article showing that clustering methodologies from bioinformatics are applicable to continuous rapidly changing multivariate physiologic data in critically injured patients - the first article of this kind for this patient population, yielding important insight into patient physiology and outcomes (Cohen, 2010 & Grossman, 2009). We hypothesized that processing of multivariate data using hierarchical clustering techniques would allow identification of otherwise hidden patient physiologic patterns that would be predictive of outcome. We identified 10 clusters, which we defined as distinct patient states. While patients transitioned between states, they spent significant amounts of time in each. Clusters were enriched for our outcome measures: 2 of the 10 states were enriched for infection, 6 of 10 were enriched for MOF, and 3 of 10 were enriched for death. Further analysis of correlations between pairs of variables within each cluster reveals significant differences in physiology between clusters.

We now have a dataset that includes patients enrolled in the tissue oxygen monitoring study as well as other patients for a total of approximately 200 patients for bioinformatic analysis. In the next year, we intend to focus on enrolling patients who have significant torso/extremity injuries (they may or may not have head injuries).  We would like to confirm the important relationship between low muscle oxygen in the first 24 hours of ICU care and subsequent development of multiple organ failure.  We would also like to examine the relationship between peripheral muscle microvascular blood flow and peripheral muscle oxygen.

Significance: Contrary to our hypothesis, muscle microvascular blood flow was significantly higher in under resuscitated patients and declines as PmO2 rises. This suggests that tissue oxygen content may be the primary driving force for peripheral perfusion and thus a worthy target to monitor during resuscitation.

Our groundbreaking work in the field of complex systems bioinformatics has significant potential to reduce injury severity, disability, and death on a large scale. We show for the first time the feasibility of clustering physiological measurements to identify clinically relevant patient states after trauma. These results demonstrate that hierarchical clustering techniques can be useful for visualizing complex multivariate data and may provide new insights for the care of critically injured patients.

Glucose Control in the ICU at San Francisco General Hospital (SFGH)

 Brief Summary of Project: A recent paper published in the New England Journal of Medicine reported that tight glucose control in a heterogeneous group of ICU patients resulted in greater morbidity and mortality. (The NICE-SUGAR Study Investigators.  Intensive versus Conventional Glucose Control in Critically Ill Patients.  NEJM. 2009;360:13-1283-97) Unfortunately, no data exists on the glucose control parameters maintained by our ICU patient protocol, the rate of hyper or hypoglycemia, or the outcome in a primarily surgical population. This retrospective trial will examine the level and effectiveness of glucose control protocols currently applied to our severely injured and ill ICU population at San Francisco General Hospital.

  Specific Aims:

1. To determine the range of glucose control achieved in our surgical/trauma ICU population.
2. To determine the rate of hyperglycemic and hypoglycemic events in our surgical/trauma ICU population.
3. To determine the effect of glucose levels and insulin protocol on outcome in our surgical/trauma ICU population.
4. To implement an evidence-based recommendation of insulin infusion protocol for acutely injured patients that optimizes outcome. 

Studies and Results: Data collection, cleaning, and analysis has begun. We are collecting retrospective data on 3 distinct cohorts of patients that correspond to 3 distinct phases of insulin infusion protocols in the ICU at our trauma center (SFGH). Until June 2005, there was no insulin infusion protocol at SFGH, but between June 2005 and February 2006 there was a strict insulin infusion protocol, and from March 2006 until the present there has been a moderate insulin infusion protocol. Retrospective data is available for a total of 2,678 patients encompassing all three phases of glucose control measures in our ICU. Daily data is available for these trauma, surgical, and neurosurgical patients from the time of admission to the ICU until discharge from the ICU. Outcome measures will be hospital days, ICU days, ventilator-free days, mortality, number of hyper- and hypoglycemic events, infection, ventilator associated pneumonia, and multiple organ failure (MOF, Denver Score). We will calculate descriptive statistics regarding the cohort.

Continuous Hemoglobin Monitoring in Acute Care

 Brief Summary of Project: In trauma, efforts to diagnose and track bleeding are paramount to the restoration of circulatory homeostasis. Unfortunately, the best current measure of hemoglobin (the complete blood count CBC) is obtained by an invasive blood draw. This test suffers from being intermittent, expensive, and slow. We are testing the hypothesis that the use of continuous hemoglobin (SpHb) monitoring to measure hemoglobin (Hb) concentration will provide a clinically useful measure of bleeding and anemia that is superior to current techniques. We will further test the hypothesis that continuous monitoring of hemoglobin in conjunction with our multivariate bioinformatics ICU data collection and analysis will provide superior information about a patient's perfusion status and predictive information about a patient's physiologic state and clinical trajectory. Ultimately, we hypothesize that we can create a predictive model of clinical outcomes related to traumatic injury and acute anemia to design interventions to improve tissue hypoxia.

  Specific Aims:

1. To determine the ability of continuous, noninvasive Pulse-CO Oximetry to accurately provide Hb measurement and provide superior tracking of physiologic trajectory in severely injured trauma patients.

  • 1a: To determine if continuous Hb measurements directly correlate with standard intermittent laboratory Hb measurements in severely injured patients beginning immediately after trauma.
  • 1b: To determine if continuous Hb measurements will provide superior tracking of bleeding and hemostasis in post trauma patients who are being monitored for potential ongoing bleeding.
  • 1c: To determine if continuous Hb measurements will provide superior information to guide transfusions in trauma patients.
  • 1d: To determine if continuous Hb measurements will provide superior prognostic information about morbidity (infection, lung injury, multiple organ failure, increased length of ICU/hospital stay) and mortality after injury.

2. To determine if continuous Hb measurement can be combined with our already monitored measures of tissue oxygen and tissue blood flow to provide superior measurement of perfusion and patient physiology.

3. To determine if continuous Hb measurement will integrate with other continuous ICU data to provide superior characterization of physiologic state and state trajectory in a multivariate informatic model.

 Studies and Results: This prospective cohort study will follow the clinical course of 200 trauma patients on admission to the emergency department (who require ICU admission) and for the subsequent 28 days. The Masimo Radical-7 Pulse CO-Oximeters will be placed upon ICU arrival and will remain for 96 hours or until ICU discharge (whichever is first).

 We have created an extensive web-based relational database (Quesgen Systems Inc. Burlingame, CA) for the direct entry and management of relevant patients. The study coordinators will record clinical data including inclusion and exclusion criteria, daily clinical and physiologic data, relevant laboratory data, and the endpoints of this study. Primary outcome measures include incidence of bleeding, accuracy of hemoglobin measures, coagulation abnormalities, and total blood product transfusions. Secondary endpoints include incidence of organ injury, ventilator associated pneumonia, 28-day mortality, ventilator-free days, ICU stay, hospital stay, and incidence of nosocomial infections.

 A Review of Spinal Cord Injury Patients Requiring Mechanical Ventilation, 2003-2007

  Brief Summary of Project: Spinal cord injury (SCI) is common after both blunt and penetrating trauma and often leaves a patient ventilator dependent in the ICU. This is a retrospective medical record review of patients who suffered a spinal cord injury and required mechanical ventilation. Whether these patients have the respiratory ability to warrant an attempt at extubation remains an important unanswered question. We hypothesized that we could identify the level of cord injury and factors associated with successful extubation after spinal cord injury.

  Specific Aims:

1. To identify predictors of mechanical ventilation requirements, including factors associated with successful extubation, in patients with SCI.

2. To identify outcomes associated with respiratory ability in patients with SCI.

  Studies/Results: The study design is a retrospective medical record review. The patient cohort comes from the trauma registry database at San Francisco General Hospital. Data collection is complete and analysis is nearing completion. Preliminary results indicate that SCI is associated with prolonged ventilator dependence and the level of cord injury correlates strongly with failure to wean and extubate. An abstract has been submitted for presentation to the American Association for the Surgery of Trauma (AAST) in September of 2010.

  Significance: Despite the correlation of cord injury severity with failure to extubate, there remains a small set of patients with high cord injury who can be safely weaned and extubated. Attention to detailed pulmonary care can maximize the likelihood of extubation in these severely injured patients.


Hemostatic Resuscitation and Traumatic Coagulopathy: Effect of Platelet Number and Function on Coagulopathy and Outcome after Trauma

  Brief Summary of Project: Coagulopathy is present in 25-38% of trauma patients on arrival to the hospital and these patients are four times more likely to die. Recently, a high Fresh Frozen Plasma (FFP):Packed Red Blood Cell (PRBC) ratio has been shown to decrease mortality in massively transfused trauma patients. Therefore, we hypothesized that patients with coagulopathy on arrival to the hospital may benefit more from transfusion with a high ratio of FFP:PRBC than those who are not coagulopathic.

 Recent studies have also shown that platelets play a central role in hemostasis after trauma. However, the platelet count of most trauma patients does not fall below the commonly taught "normal range" (150K-450K), and as a result, little attention has been paid to admission platelet count as a predictor of outcome. We examined the relationship between admission platelet count and outcome after trauma.

  Specific Aims:

1. To determine if coagulopathic patients benefit more than non-coagulopatic patients from transfusion with a high ratio of FFP:PRBC.
2. To examine the relationship between admission platelet count and outcome after trauma.
3. To suggest an optimal transfusion protocol for acutely injured patients.

 Studies/Results: We conducted two studies. The first was a retrospective multi-center cohort study of 452 massively transfused trauma patients to determine if the effect of the ratio of FFP:PRBC on death at 24 hours is modified by a patient's coagulation status on arrival to the hospital. Data collection is complete and data analysis is nearing completion. Preliminary analysis indicates that those who arrived to the hospital with coagulopahty (INR ³1.4) had a greater risk of death than those without coagulopathy (INR <1.4). However, as the ratio of FFP:PRBC transfused increased, mortality decreased similarly between the two groups. Most of the decrease occurred with transfusion of at least one half unit of FFP for every one unit of PRBCs transfused. 

The second was a retrospective cohort study of 389 massively transfused trauma patients to test the association between admission platelet count and 24 hour mortality and units of PRBCs transfused. Data collection is complete and data analysis is nearing completion. Preliminary results indicate that admission platelet count was inversely correlated with 24 hour mortality and transfusion of PRBCs.

Significance: The mortality benefit from a high FFP:PRBC ratio is similar for all massively transfused trauma patients. This is contrary to the current belief that coagulopathic trauma patients benefit most from a high FFP:PRBC ratio. Furthermore, it is unnecessary to determine whether coagulopathy is present before transfusing a high FFP:PRBC ratio. Further studies are needed to determine the mechanism by which a high FFP:PRBC ratio decreases mortality in all massively transfused trauma patients.

 Admission platelet count was inversely correlated with 24 hour mortality and transfusion of PRBCs. Our results suggest that a normal platelet count may be insufficient after severe trauma, and as a result, these patients may benefit from a lower platelet transfusion threshold.