Base Deficit

1

I am 5 weeks into my rotation on the ICU team in a surgical/trauma ICU. The experience is exciting and challenging. The ICU team in this particular unit manages ventilation, sedation, pain management, and vascular access. The primary team (eg. trauma, neurosurgery) manages all other aspects of care. When a patient requires admission the ICU team is notified, and the patient is admitted concurrently by the primary team and the ICU team. My role on the ICU team is to admit the patient, gather a history, perform a physical, and write pertinent orders (ventilator settings, pain meds, etc). I will then present any patients I have admitted at ICU rounds and continue to manage their ICU needs. Of course, I don't do this independently; a resident and the fellow are close at hand. Every note I write and all my orders are co-signed by one of them. So far, this has been an invaluable learning experience.

The patient population in this ICU is varied. Most of the patients are traumatically injured, but we also admit neurology and neurosurgery patients. Additionally, we admit burn patients, as well as those who require ICU care after surgery. Ages range from very young to very old. Also, many of our patients are malnourished and homeless.

The acuity of illness and injury in these patients is usually quite high. They require aggressive resuscitation, immediate intervention, and intensive monitoring.

We monitor many of the usual parameters: blood pressure, heart rate, urine output, skin signs, mental status, etc. An additional parameter we monitor is base deficit. Practically every patient in the unit has an arterial catheter, and blood gases are drawn frequently.

While it would be interesting to assess the relationship between presence of arterial

catheters and frequency of blood gases, I am curious about the use of base deficit (BD) in

2

assessment of the trauma patient. BD is widely accepted as an indicator of resuscitation, but are there other reasons to assess BD? Could BD give us additional information relevant to a patient's status?

Recently, a 76-year-old woman was admitted after a motor vehicle crash. She had a prolonged extrication and arrived in the emergency department ninety minutes after the crash. She had multiple facial fractures, a fractured right femur, a right temporal subdural hematoma, and a ruptured spleen requiring emergent surgery. Her BD upon arrival in the emergency department was -16, which improved slightly to -10 upon admission to the ICU after surgery. She continued to be resuscitated, now with blood products. In the next bed was a 72-year-old man who had also been injured in a motor vehicle crash. He was admitted to the ICU with a right pneumothorax, a fractured right humerus, and a large liver laceration. He also required operative intervention. His admission BD was -10, improving to near normal with resuscitation. Here were two elderly people, both with multiple injuries, both requiring surgery, but with large differences in their BD. Could that difference have significance?

My curiosity lies not in the use of BD for determining resuscitation status, but rather in the prognostic value of BD values. What does a BD of -16 imply? Will someone with an initial BD of -16 have a worse outcome than someone with an initial BD of -10? Will their chance of dying be greater? Could the initial BD value have meaning in the determination of treatment?

Trauma services strive to provide aggressive, timely, and effective interventions to all patients after major traumatic injury. At the same time, efficient resource

3

management presents an ongoing challenge for most trauma centers. It has been noted that a readily available marker of both increased risk for poor outcome and increased resource consumption could facilitate the provision of effective care [Davis, 1996 #263].

Base deficit (BD), a nonspecific indicator of anaerobic metabolism and acidosis, has emerged as an important variable in the resuscitation of trauma patients. BD is the amount of base, in millimoles, required to titrate one liter of whole arterial blood to a pH of 7. 40. It is a calculated value, derived from the arterial blood gas based on the Astrup and Siggard-Anderson nomogram [Porter, 1998 #371]. BD has demonstrated a good cor-relation to lactate, a sensitive marker of anaerobic metabolism that has been well studied [Porter, 1998 #371]. Previous studies have demonstrated the utility of BD in assessing shock and resuscitation. The severity of admission BD correlates with the volume of fluid required for resuscitation [Davis, 1991 #534]. BD has been shown to normalize rapidly with adequate resuscitation and control of hemorrhage [Chang, 1993 #527].

Currently, there is interest in the ability of BD to predict outcome, complications, and resource consumption. The early identification of patients at high risk for death and shock-related complications would be valuable at the time of triage in order to allocate resources properly, as well as to determine the needed intensity of resuscitative efforts.

The purpose of this paper is to critique several key studies evaluating the ability of BD to predict mortality, morbidity, and resource consumption. Conclusions derived from these studies will be presented. Current findings will be summarized and ideas for future research provided.

4

Research Critiques

Prediction of Mortality

[Rutherford, 1992 #611] studied the association of BD with mortality. They retrospectively reviewed all trauma admissions between August 1984 and March 1990. Patients with an arterial blood gas sample drawn within the first 24 hours who had a base deficit were included in the study. They studied 3,038 patients. Using a logistic regression model, they found that BD, age, injury mechanism, and head injury were associated with mortality. Age < 55 years, no head injury, and a BD of -15 were associated with a 25% mortality. Age > 55 years with no head injury or age < 55 years with a head injury and a BD of -8 were associated with a 25% mortality. They concluded that the BD was an expedient and sensitive measure of inadequate perfusion.

While the magnitude of the sample size lends power to these findings, the retrospective nature of this study creates a problem. Confounding variables such as time from injury, use of resuscitative drugs, and the use of bicarbonate were uncontrolled in this study. Also, by using the worst BD during the first 24 hours, BD values obtained after significant resuscitation had occurred were included in data analysis.