Bridging the Critical Care Chasm
Written by Colonel Leopoldo Cancio
A new Army combat casualty program aims
to close the "technology gap" between
battlefield and U.S. trauma center care.
This gap in technology lies between what’s currently the state-of-the-art care on the battlefield and the best in care to be found at a well-equipped U.S. trauma center emergency department (ED) or intensive care unit (ICU). The gap is particularly notable at echelons 2 and higher and en route, including during inter-facility transport.
The impact of improved critical care capabilities on mortality, as well as the use of resources, was suggested by K. Grathwohl et al. in Critical Care Medicine in 2008. In that study, both ICU length of stay and ICU mortality decreased progressively at the combat support hospital in Baghdad as the model changed from “no intensivist” to “intensivist consult” to “intensivistdirected team.”
Similar convincing results have been seen in U.S. hospitals. The problem, however, is the scarcity of manpower on the battlefield: “Despite a rapidly mobile critical care platform, the U.S. military is unfortunately faced with ... shortages of critical care physicians and nurses to staff extended worldwide missions,” the study noted.
IS TECHNOLOGY THE ANSWER?
An important question is whether technology is the solution to this problem. The Institute of Medicine in its 2001 report “Crossing the Quality Chasm: A New Health System for the 21st Century” recognized that U.S. “health care delivery has been relatively untouched by the revolution in information technology” and that “information technology must play a central role in the redesign of the health care system.” Indeed, the vision of the C3E program is to employ technology—including better monitors, automated care systems and other critical care devices—to bridge the quality gap in battlefield health care.
Better patient monitors are needed for earlier diagnosis, targeted interventions, error reduction and improved documentation. As we continue to develop therapeutic improvements in trauma medicine, such as plasma in the ED, blood substitutes and drugs to stop bleeding, we also need to give providers objective data about when to use these capabilities.
Furthermore, costly errors of over- or under-resuscitation, or over- and under-triage, could be avoided by better monitors. In particular, traditional vital signs—which are ubiquitous in clinical care—nevertheless fail to accurately diagnose trauma patients in need of lifesaving interventions. This has been well documented in medical literature such as the Journal of Trauma.
This is probably because the body’s defense mechanisms tend to compensate for blood loss until it’s too late. Another limitation is that existing monitors take a long time to hook up to a patient, are big and bulky, employ wires that get in the way and provide lots of data but no actionable information—thus, military medics may be disinclined to use them.
Technology is also needed to reduce the errors of under- and over-triage. With undertriage, a provider fails to detect that a patient is critically injured and fails to diagnose in a timely fashion. In a recent study, deaths at a combat support hospital with high preventability scores were primarily related to delays in bleeding control. They occurred during the transportation (47 percent) or resuscitation phases (43 percent) of care, and were attributed to the system (63 percent) and/ or provider (70 percent) most frequently, as noted by M. Martin et al. in an April 2009 Journal of Trauma article.
With over-triage, a provider diagnoses too many patients as critically ill who in fact are not. This is common (50 percent of patients in urban U.S. trauma centers) and is considered the price paid for missing patients with severe injuries. But in a resource-scarce environment such as a mass casualty event or on the battlefield, over-triage causes an increase in the mortality of the critically wounded by using up scarce resources such as emergency department beds.
AVOIDING ERRORS: OVER- AND UNDER-RESUSCITATION
The need for automated critical care is evident from studies of over- and under-resuscitation of combat casualties. Our experience with combat casualties points to the high risk of abdominal and extremity compartment syndromes when burn shock resuscitation, for example, is performed by non-burn experts. Decreased fluid infusion can be achieved and complication rates reduced when resuscitation is performed using decision support software, even in a burn center.
Complications of over-resuscitation and of persistent coagulopathy are also common in patients undergoing massive transfusion following hemorrhagic shock. Decision Support algorithms have been successfully employed at a handful of busy civilian trauma centers, permit the implementation and periodic modification of evidence-based medicine and clinical practice guidelines, and reduce the occurrence of these complications. A principal goal of C3E is to adopt this approach so that it can be effectively implemented in combat support hospitals and other military treatment facilities (MTFs).
The need for “better effectors” and, in particular, mechanical ventilators and other therapies for intubated combat casualties, was perhaps best exemplified by a study by D. Davis et al., published in Critical Care Medicine in 2006, of patients with severe head injury, in which errors in over- or underventilation caused increased mortality.
ADDRESSING THE NEED
The goal of C3E is to develop new systems- based technology that creatively fills the gap between what is available for care in continental United States (CONUS) trauma centers and what is provided to combat casualties on the battlefield. Technology in this sense means hardware and software systems incorporating sensors, processors and effectors, including monitors, pumps, computers, ventilators and the like. The goal is not merely delivering existing CONUS technology to the combat zone. The battlefield is different from CONUS trauma centers in several key respects, including:
• constraints of weight and cube, even at level 3 combat support hospitals;
• personnel shortcomings and shortages, such as the absence of a burn surgeon at all MTFs and shortages of ICU physicians;
• high injury severity scores of combat casualties, without equal in CONUS centers;
• frequent mass casualty scenarios; and
• multiple echelons of care, with loss of continuity as a patient moves through the system.
Thus, we must maintain a balance between meeting the unique needs of the military environment and developing products that can have wider application in the civilian world as well. The focus of C3E is on delivery of products within two- and five-year time frames—meaning an emphasis on products that are approved or nearly approved by the Food and Drug Administration, and that have viable commercial pathways to market.
To fill the critical care technology gap, C3E is committed to three specific tasks: new vital signs such as improved patient monitors, designed to predict the need for lifesaving interventions more accurately; automated critical care, including decision support, open-loop and ultimately closedloop automation of difficult critical care tasks; and better effectors such as improved devices for support of organ function in critically injured combat casualties, with the major emphasis on lung support.
NEW VITAL SIGNS
New vital signs will be developed to meet the battlefield need for earlier and more accurate diagnosis. There are four basic approaches to getting new vital signs: repackaging, in which existing technology is reconfigured to make it smaller, lighter and wireless; data fusion, in which traditional vital signs are combined using multivariate analysis; extraction, through which existing sensors such as the electrocardiogram (EKG) waveform produce more information; and new sensors that can measure physiologic endorgan effects of trauma, such as tissue oxygen saturation and tissue carbon dioxide content.
There are advantages to all of these approaches. Repackaging is the focus of an ongoing effort in conjunction with Athena GTX to develop the Wireless Vital Signs Monitor (WVSM), which is intended for use as far forward in battlefield medical response as the casualty evacuation (CASEVAC) point. This effort continues, with collection of prehospital trauma patient data in 2009–2010.
Data fusion has successfully been performed in the FDA-approved Visensia system, made by Carmel, Ind.-based OBS Medical. This device uses the traditional vital signs, subjects them to a machine-learning process and generates a single risk index for cardiopulmonary instability. This device has not, though, been validated in acutely injured patients.
Similarly, the C3E program has developed an artificial neural network (ANN) that predicts the need for a lifesaving intervention using the trauma vitals database from Houston and San Antonio. This algorithm is now being tested prospectively in the same trauma systems.
Extraction of more information from the EKG of the heart by means of complexity or variability analysis is particularly appealing, because it uses a sensor (the EKG) that is uniformly used in patient care already. Our laboratory has pursued this line of research for several years, demonstrating that heart-rate complexity served as a superior method of predicting the need for lifesaving interventions in trauma patients. New vital signs derived from heart-rate complexity analysis have been imbedded in the ANN mentioned above, and will likely be embodied in decision support monitors (see below).
We have identified a heart-rate complexity metric, PD2i, which is now FDAapproved and marketed by Boca Raton, Fla.-based Vicor Technologies. This device will be tested in trauma patients under C3E.
Finally, there are several promising approaches to new sensor development. These include pulse oximetric estimation of hemoglobin and of stroke volume; continuous sublingual or buccal capnometry; and next-generation tissue oxygen saturation measurement by near-infrared spectroscopy. Further product development and trauma patient studies are needed for all of these devices.
AUTOMATED CRITICAL CARE
Automated critical care takes the C3E program a step beyond new vital signs and represents a new paradigm in critical care medicine intended to avoid the mistake of overwhelming providers at any level with more data.
This paradigm recognizes three levels of knowledge: data, including the raw vital signs and lab values, with minimal interpretation other than individual outof- range alarms for each data point (e.g., “Blood Pressure is 80/40.”); information that interprets the data and assigns it meaning (e.g., “Patient is hypotensive due to blood loss and has an 85 percent likelihood of requiring a lifesaving intervention.”); and automation that recommendations for a course of action (e.g., “Give the patient two units of plasma.”). Automated critical care includes the following levels of increasing sophistication (and regulatory challenge):
• decision support, in which the system makes a recommendation to the provider and the provider implements the recommendation if desired—for example, by manually changing the fluid rate on a pump;
• open-loop control, in which the system makes a recommendation, enables the provider to accept or reject the recommendation, such as with a push-button interface, and includes a method for performing the intervention—for example, by connecting with a fluid infusion pump); and
• closed-loop control, in which the system fully controls the care, with delivery of periodic updates to the provider—for example, the system measures the vital signs and automatically adjusts the fluid infusion rate to achieve a target blood pressure.
Our regulatory experience argues in most cases in favor of an incremental approach to automated critical care. That is, our efforts are currently focused on decision support rather than on closedloop control because of the more stringent regulatory requirements associated with closed-loop systems. Currently, in conjunction with the clinical trials program, we have developed a burn decision support algorithm that runs on either a laptop PC or on a PDA, takes the urine output of a patient, the burn size and the time postburn; subjects these data to an algorithm; and outputs a recommended resuscitation fluid rate. This product is ready for commercialization at this time. It is the prototype for a planned family of decision support algorithms; other products will include resuscitation—including massive transfusion—of patients with hemorrhagic shock, management of hypotension due to post-injury sepsis, management of cerebral perfusion pressure for headinjured patients, post-injury nutrition support, weaning of pressors, weaning of mechanical ventilation and documentation of wound healing.
The primary test beds for these products will be the U.S. Army Institute of Surgical Research Burn Center and participating U.S. trauma centers—with the combat support hospital, forward surgical team and en route care systems serving as the primary customers. In addition, we support efforts to develop closed-loop control for management of the fraction of inspired oxygen and of minute ventilation in mechanically ventilated patients.
BETTER EFFECTORS
C3E goals to develop better effectors are focused on lung support and include development of simplified ventilators for use in trauma patients, especially those with severe head injury; ventilators and other therapeutics for improved survival in patients with acute respiratory distress syndrome (ARDS) due to combat injuries, to include smoke inhalation, pulmonary contusion, and massive transfusion; and extracorporeal devices to augment or entirely replace the mechanical ventilator in patients with severe ARDS.
Simplified ventilators are important because mis-ventilation is common in head-injured casualties and can cause cerebral ischemia (over-ventilation) or cerebral hypoxia (under-ventilation), leading to more deaths. The Defense Advanced Research Projects Agency has fielded a simplified automated ventilator that provides push-button ventilation with no means of adjustment to accommodate differing patient sizes or metabolic needs. This concept needs to be tested in trauma patients.
The need for ventilators with ARDS, as well as the need for extracorporeal devices to augment or replace mechanical ventilators, reflects the continued high mortality of ARDS in trauma patients—about 30 percent. Although the 2000 ARDSnet trial established the efficacy of low-tidal volume ventilation in selected categories of ARDS patients, it did not provide a means for easily achieving these goals, and it failed to establish the best mode of ventilation in patients such as those with severe smoke inhalation injury. Our efforts in this area are directed at evaluation of airway pressure release ventilation and volumetric diffusive respiration modes in realistic, multi-day models of smoke inhalation injury, and at development of extracorporeal carbon dioxide removal devices to facilitate gentle mechanical ventilation.
CORE CAPABILITIES
The core capabilities that the C3E program brings include an integrated team of intramural and extramural collaborators, consisting of combat-experienced trauma surgeons and intensivists; computer scientists; biomedical engineers; and physiologists. Because of the heavy emphasis on trauma patient validation and rapid product delivery to the battlefield, the focus of C3E is on clinical trials in trauma and burn patients, and on product testing in clinically relevant models of severe injury where appropriate.
Colonel Leopoldo Cancio, M.D., a fellow of the American College of Surgeons, is with the U.S. Army Medical Corps at the U.S. Army Institute of Surgical Research at Fort Sam Houston, Texas. This article represents the private views of the author and are not to be construed as official or as reflecting the views of the Department of the Army or Department of Defense, nor does the mention of specific products constitute an endorsement. ♦