Emergency Medicine Reports February 3, 1997

The Trauma Panel: Laboratory Test Utilization in the Initial Evaluation of Trauma Patients

Author: Andrew W. Asimos, MD, Director of Resource Utilization, Department of Emergency Medicine, Carolinas Medical Center, Charlotte, NC.

Peer Reviewers: Allan B. Wolfson, MD, FACEP, FACP, Program Director, Affiliated Residency in Emergency Medicine; Professor of Emergency Medicine and Medicine, University of Pittsburgh,Pittsburgh, PA.

Thomas Rebbecchi, MD, Assistant Professor of Medicine, Director of Undergraduate Emergency Medicine Education, Temple University Hospital and School of Medicine, Philadelphia, PA.

A recent survey found that 95% of U.S. Level I trauma centers perform costly routine laboratory testing of trauma patients, despite the paucity of scientific data supporting this practice.1 Such standardized testing has been advocated as a method of fostering consistency in patient care, legitimizing the oral orders given during patient evaluation, and helping to achieve the best possible outcome for trauma patients.2 The "trauma panel" usually consists of some constellation of the following studies: a complete blood count (CBC), electrolytes, BUN, creatinine, amylase, protime (PT), prothrombin time (PTT), blood type and screen (T&S) or type and crossmatch (T&C), urinalysis (UA), ethanol level, urine toxicology screen, arterial blood gas (ABG), lactate, and an electrocardiogram (ECG). The patient charge for this entire battery of tests typically exceeds $500.

There have been several, mostly retrospective, studies in trauma patients that have evaluated the utility of routinely obtaining various components of the trauma panel.

By reviewing these studies and considering the individual components of trauma test batteries, some rational guidelines for test ordering in the setting of acute trauma can be recommended.

—The Editor

Point-of-Care Testing

Point-of-care testing (POCT) represents the latest phenomenon of laboratory testing. It refers to any testing performed outside the hospital’s main laboratory. Many POCT analyzers afford the rapid availability of a profile of blood gas and chemistry parameters from a whole blood sample. However, virtually all studies on POCT reveal a considerably higher cost-per-test for performing POCT vs. centralized lab testing.3-5 For example, one analysis concluded that it would cost $243,442 per year to perform a routine Chem7 on 10 patients/shift/day via a POCT analyzer vs. $91,678 in the hospital’s main laboratory.3,4 Despite this increased cost-per-test, proponents of POCT emphasize that this cost is justified because having results available sooner will ultimately decrease overall patient care costs. Theoretically, this occurs due to reduced morbidity from earlier and more accurate diagnoses, the performance of fewer total therapeutic or diagnostic tests, and shortened hospitalizations.

Frankel et al evaluated the utility of a battery of tests available through a POCT analyzer in the initial evaluation of 200 consecutive trauma patients.6 Tests available through the POCT instrument included the sodium (Na+), chloride (Cl-), potassium (K+), glucose, pH, PCO2, PO2, hematocrit (Hct), and calcium (Ca++). No patient outcomes other than the need for emergent surgery were evaluated. With regard to K+, no patients were hyperkalemic and 34 patients were hypokalemic (K+ < 3.5), of whom only 12 were treated. Hypokalemia was more prevalent in hypotensive patients, those with an Injury Severity Score of at least 9, and those requiring emergent surgery. Considering glucose levels, 21 patients had levels higher than 200 mg/dL, but only one patient, whose glucose level was higher than 400 mg/dL, received insulin. While 30 patients were acidemic (pH < 7.34) and 29 patients were alkalemic (pH > 7.46), it is unclear whether those results affected patient management in any way. Additionally, it is unclear whether abnormalities in PCO2, PO2, and Ca++ levels influenced patient management. Finally, regarding the Hct level, only 10 of 36 patients with Hct less than or equal to 0.33 received packed red blood cell (PRBC) transfusions in the resuscitation area, and seven others received blood products within 24 hours of admission. Of note, no comment was made regarding the contribution of the Hct level in deciding which patients were transfused.

The authors concluded that, with the exception of Na+ and Cl-, all POCT values were "abnormal with sufficient frequency in a variety of patients to warrant their acquisition"; however, they largely did not describe how abnormal values affected the patients’ management. Moreover, in cases where management was changed based on abnormal values (i.e., treating hypokalemia), no comparison was made between the outcomes of patients who were treated vs. those who were not. Finally, while the authors state that the POCT analytes measured are of special interest in injured patients who require anesthesia, surgical intervention, or sophisticated critical care, their impact on management of those situations is debatable.7

POCT analyzers offer considerable potential in the evaluation of trauma patients because they provide rapid test results concurrent with a patient’s initial evaluation and stabilization. As discussed in a subsequent section, if they provide the capability to perform blood gas and lactate measurements substantially faster than the central lab, their use may be warranted. However, before POCT analyzers can be recommended for routinely performing these tests on all major trauma patients, prospective studies are needed to demonstrate that POCT changes either diagnostic management or therapeutic management, outcome, or overall patient care costs.

Electrolyte Panels

Several groups have looked at the utility of a standard electrolyte panel in the management of trauma victims. Tortella et al retrospectively studied the utility of serum chemistry panels in 913 trauma patients.8 Only 6% (n = 54) had "clinically significant laboratory values," and only six of these 54 prompted a change in resuscitation or treatment (therapeutic K+ infusions for hypokalemia in all six cases). Moreover, there were an additional six cases of hypokalemia that, for unclear reasons, were not treated, but all values normalized with no incidence of significant cardiovascular side effects. Of note, the authors concluded that a history of hypertension, age older than 50, and a Glasgow Coma Scale (GCS) score less than or equal to 10 appeared to be useful criteria for selectively ordering chemistry profiles.

Recently, in 456 blunt and penetrating trauma patients, Namias et al examined the impact of chemistry profiles on clinical interventions.9 While 93% of the profiles obtained had at least one abnormal component, the authors concluded that an intervention was made in response to only five of the abnormal chemistry results. These included administration of K+ to a patient with a serum K+ of 3.1 and treatment of four hyperglycemic patients: two known diabetics received insulin, one patient with severe head injuries received insulin 13 hours after the hyperglycemia was reported by the laboratory, and one patient had a medical workup of diabetes initiated, but received no insulin. It is important to note, however, that the authors used the number of interventions made based on abnormal results of that test as the sole measure of the utility of a laboratory test. Furthermore, there was no attempt to distinguish the reason for any intervention. Indeed, it is impossible to know whether any intervention identified in their study resulted from an abnormal laboratory result vs. a clinical finding.

Lowe et al have prospectively validated 10 clinical criteria predictive of electrolyte abnormalities affecting diagnosis or treatment (sensitivity, 94%; specificity, 28%).10 They include: a history of poor oral intake; vomiting; chronic hypertension; taking diuretic medication; recent seizure; muscle weakness; age at least 65 years; alcoholism; abnormal mental status; or recent history of electrolyte abnormality. While these criteria were primarily applied to medical patients in their prospective validation, three out of eight patients with clinically significant electrolyte abnormalities missed by Lowe’s clinical criteria (out of a total of 982 cases) were trauma victims. One was a victim of a motor vehicle accident (MVA) who had transient hypotension and an HCO3- of 16; one was an MVA victim with hip and nasal fractures and a K+ of 3.0; and one was a victim of a superficial stab wound with an HCO3- of 19. None of these three abnormalities was concluded to have influenced patient management. The initial patient’s HCO3- was normal two days later, the hypokalemic patient was discharged on the next day with no repeat potassium level checked, and no other comment or treatment was offered on the stab wound victim’s low HCO3-. Additionally, Nelson et al found that 78% (n = 48) of electrolyte panels were normal in trauma patients, but they did not correlate treatment, outcome, injuries, or exam with either normal or abnormal values in their study.11

Finally, regarding indices of renal function, Frankel et al evaluated the utility of BUN and creatinine in trauma victims.6 They identified 10 trauma patients with elevated BUN or creatinine levels, including six patients with prior chronic renal failure, hypertension, or diabetes mellitus. All four patients with elevated BUN or creatinine levels without premorbid renal dysfunction (hypertension or diabetes) were hypotensive on arrival but returned to normal renal function subsequently, including two who received IV contrast for "one-shot" pyelography. In Namias et al’s study, they found that 13% of patients had elevated BUN levels and 7.5% of patients had elevated creatinine levels. No interventions based on abnormal BUN or creatinine were identified.

Based on these studies, some criteria can be suggested in which obtaining electrolytes and indices of renal function may yield abnormal values. (See Table 1.) What remains unclear, however, is to what extent any abnormalities identified by performing these tests are of any clinical consequence in the setting of acute trauma.

Complete Blood Count

While the complete blood count (CBC) is commonly ordered on trauma patients, there are data suggesting that the entire CBC is frequently not needed in many patients for either the ED or anesthetic management.7 Nelson et al found that 62% (n = 37/60) of CBCs were normal in trauma patients but in their study did not correlate treatment, outcome, injuries, or exam with normal or abnormal components of the CBC.11 The hemoglobin or hematocrit level is usually of most interest from the CBC profile. While a low hemoglobin level observed after injury is usually an indicator of serious ongoing hemorrhage, its impact on patient management, and even emergent blood transfusion, is questionable.6,12 Furthermore, the sensitivity of an initial hemoglobin or hematocrit for detecting ongoing hemorrhage is unclear.

The white blood cell (WBC) level has been shown to have no relation to intra-abdominal injury in trauma patients and it does not reliably predict severity or cause of disease in acutely ill adults of various causes.6,13 Frankel et al found that the presence of intra-abdominal injury was not associated with an increased relative risk for an elevated WBC count.6 However, they did conclude that the platelet count can be of some use in patients with severe head injuries, but it rarely contributes to the initial management in other trauma patients.

Based upon these observations, most would agree that a baseline hemoglobin or hematocrit should be routinely obtained in all major trauma patients because it provides a useful starting point for patients thought or known to have ongoing hemorrhage. However, a WBC count by convention offers no help in the initial evaluation of trauma patients, and, beyond the setting of severely head-injured patients, a platelet count is not routinely needed.

Coagulation Profiles

Indices of coagulation status are often of interest in the setting of acute trauma, particularly in patients with severe head trauma who are at risk for the rapid development of disseminated intravascular coagulopathy (DIC). Frankel et al found that of 12 patients with a GCS score less than or equal to 8, seven had a low platelet count or prolonged PT/PTT with abnormalities that appeared to occur in proportion to the severity of the head injury.6 Namias et al found that either PT or PTT, or both, were abnormal in 59% of the 375 trauma patients studied.9 Increased incidences of abnormal PT and PTT, independent of one another, were associated to each of a low systolic blood pressure (SBP) (£ 90 mmHg), RR (£ 10/minute), and GCS less than or equal to 12. Importantly, 36% of PT abnormalities were more than 15 seconds, and all but one were less than 31 seconds. While PTT was elevated in only 3% of all patients, 33% of these patients were given fresh frozen plasma (FFP). FFP was administered to another 3% of the total patients with some abnormality in their clotting profile.

These studies suggest that a PT/PTT may provide clinically useful information in patients with severe head injuries, and performing these studies in this patient population may be prudent. Additionally, certain vital-sign thresholds appear to be associated with PT and PTT values outside the normal range. (See Table 1.) However, the clinical significance of these abnormal values is unclear.

Amylase

Several groups of investigators have looked at the value of obtaining a serum amylase in victims of blunt trauma as an indicator of intra-abdominal injury. Mure et al, along with Frankel et al, found the sensitivity, specificity, and predictive value of amylase to be poor for determining intra-abdominal injury whether accompanied by craniofacial injury or not.6,14 Positive predictive value (PPV) of amylase for intra-abdominal injury was only 44% in patients with associated head/maxillofacial injury and 28% in patients without head or maxillofacial trauma. Other groups agree that elevated amylase is neither sufficiently sensitive nor specific to be of use in either blunt abdominal trauma patients or those in traumatic shock.15-17 More recently, while hyperamylasemia was found in 7% of 429 trauma patients studied by Namias et al, they found no interventions made on the basis of these elevated levels.9

The overwhelming scientific data indicate that serum amylase adds nothing to either the diagnostic or therapeutic management of trauma patients. It cannot be advocated as a routine appropriate test, even in the setting of abdominal trauma.

Blood Bank Testing

The blood bank represents perhaps one of the most inappropriately used and costly aspects of the trauma panel. This is largely a result of a misunderstanding of what blood bank compatibility testing accomplishes, along with concern over the likelihood of serious transfusion reactions when using abbreviated compatibility testing. Furthermore, before recommending guidelines for routinely performing a T&C vs. a T&S, the time required to convert a T&S to a T&C needs to be considered in addition to the differences in patient charges between these two services. In most instances, it takes approximately 10 minutes for blood bank personnel to crossmatch units of blood previously typed, barring any complicated serologic incompatibility. Moreover, crossmatching even only one unit of blood is a costly practice. Typically, the patient charge for crossmatching each unit of blood is about $75. Therefore, for example, a $300 patient charge is incurred for a T&C for four units vs. only about $50 charge for a T&S. In addition, when a T&C is performed vs. a T&S, the rate of blood outdating and blood wastage can be higher.18 When a crossmatch is performed and compatible donor units are identified, they can be reserved for a patient for up to two days and can be taken out of the available blood pool.

A few groups of investigators have retrospectively reviewed ED transfusions in trauma patients to identify positive predictors for blood transfusion and establish criteria for obtaining a T&C vs. a T&S in trauma patients. In 1987, West et al reviewed clinical variables in trauma patients to develop guidelines for ED requests for blood.19 They concluded the best predictor of blood use was the trauma score (TS). Of 250 trauma victims, 71% had a TS of greater than 14, and 91% of these patients did not require transfusion. By contrast, of patients with a TS of less than or equal to 14, 70% required transfusion (P < 0.001). Of note, this study was performed prior to the development of the Revised Trauma Score (RTS).20 There has been no published retrospective or prospective investigation of the ability of the RTS to predict the need for prbc usage in trauma victims. Clarke et al conducted a retrospective review of all types of patients (not just trauma victims) for whom crossmatching of blood was considered.21 Via multiple regression analysis of several variables, criteria were constructed from parameters that had a significant relationship with the likelihood of transfusion:

• SBP < 90 mmHg

• SBP < 100 mmHg and pulse > 120 BPM, unless a tilt test was negative

• SBP < 110 mmHg with a pulse > 140 BPM, unless the tilt test was negative

• A positive tilt test

• Hematocrit < 30%

• Observed blood loss of at least 500 cc or grossly visible GI bleeding

• Emergency operation with anticipated blood loss

There has been no publication of the prospective validation of Clarke et al’s criteria. Hooker et al, via a retrospective study, found that significantly more trauma victims with a prehospital SBP less than 100 mmHg required transfusion.22 A subsequent prospective validation study evaluated prehospital hypotension as a guide to crossmatching in 136 trauma patients. Only eight of 109 patients without prehospital hypotension received a transfusion. Six of these received blood because of operative procedures, and none of these patients received uncrossmatched blood. Of the two remaining patients, one had a massive head injury with DIC and was typed and crossmatched by the emergency physician upon arrival. The second patient had pelvic fractures and became hypotensive in the ED after initially being hemodynamically stable but did not require the use of uncrossmatched blood. The authors concluded that prehospital blood pressure is a useful adjunct to clinical judgment in identifying major trauma patients who can be initially managed safely without crossmatching. Recently, Davis et al conducted a retrospective study of the association between base deficit (BD) and blood transfusion in trauma patients.23 They concluded that patients with a BD less than or equal to -6 should undergo a T&C vs. a T&S, based on the finding that transfusions were required within 24 hours of admission in 72% of patients with a BD less than or equal to -6 vs. only 16% of patients with a BD greater than -6. Finally, in Frankel et al’s study, only 8.5% (n = 17) of the trauma patients evaluated required a blood transfusion within 48 hours of admission.6 Moreover, nine out of 10 patients who were transfused while in the resuscitation area received uncrossmatched blood. They concluded that the need for performing even a T&S routinely for all trauma patients may need to be re-evaluated.

Before offering any suggestions on blood bank utilization based on the studies described above, it is important to recognize the incidence of clinically significant transfusion reactions from administering uncrossmatched blood. The risk of transfusing uncrossmatched blood into a patient entails the possibility of transfusing incompatible blood due to a clinically significant, non-ABO antibody. Fortunately, this occurrence is rare, with the likelihood of encountering non-ABO antibodies ranging from 0.04% in an individual who has neither been transfused nor pregnant to about 0.3% in a previously transfused multiparous woman.24 More importantly, most of these antibodies do not cause clinically significant hemolysis in patients who are transfused. In 1992, Heddle et al conducted a prospective study in 9128 patients to evaluate this issue.25 Of the 27 transfusion episodes where the antibody crossmatch was positive, there were no clinical or serological episodes of hemolysis.

Most experts agree that the main priority in trauma patients should be to ensure the availability of type-specific blood as soon as possible.26,27 Beyond ABO/Rh typing, one needs to consider the likelihood of blood transfusion, along with its urgency, and the time required by the blood bank to perform both a T&S and a T&C. While the studies above describe criteria for obtaining a T&C, given the remote likelihood of a serious transfusion reaction with blood that is compatible based on the T&S and given the fact that a T&S can be converted to a T&C within a matter of time that rarely result in increased morbidity or mortality to the patient, a T&S may suffice in many of these scenarios.

Electrocardiogram

Most published reports on the topic of obtaining an initial ECG in trauma patients involve their use in detecting myocardial contusion. However, any consideration of the scientific literature on myocardial contusion is significantly complicated by the lack of an agreed-upon "gold standard" for diagnosing this entity. Nonetheless, most of the literature on this topic suggests that there is some utility in obtaining an initial ECG on patients who are suspected to have sustained a myocardial contusion.28-33 It is recognized, however, that an abnormal admission ECG does not necessarily positively correlate with subsequent cardiac complications.34 Moreover, some studies have shown that abnormalities were not present in a significant number of initial ECGs of patients who subsequently had significant arrhythmias during their evaluation for myocardial contusion.30,31

For trauma patients in whom myocardial contusion is not suspected, the benefit of performing routine ECGs is unknown. Nelson et al found that 86% (n = 31) of ECGs were normal in all trauma patients they evaluated, but their study did not correlate treatment, outcome, injuries, or exam with normal or abnormal ECGs.11 Garland and Wolfson have retrospectively evaluated predetermined criteria for obtaining an admission ECG in medical patients.35 Their criteria consist of:

• history of coronary artery disease, arrhythmia, congestive heart failure (CHF), conduction disturbance, cor pulmonale, pericarditis, or cardiomyopathy

• palpitations

• syncope or coma

• symptom complex suggestive of angina or CHF

• suspected cardiotoxic overdose, irregular pulse

• pulse greater than 120 or less than 60 BPM,

• SBP greater than 200 or less than 90 mmHg, or DBP greater than 120

• serum K+ greater than 5.7 or less than 3.0.

In their study, 32% (n = 202/631) of patients had an admission ECG performed, but none met the criteria above. In only three of these patients did the ECG change patient management; however, in none of these instances was patient outcome ultimately affected. While these criteria were not developed for or applied to trauma patients, the development and application of a similar set a criteria for trauma patients would be worthwhile. Until such criteria are established and prospectively validated, the decision to obtain an ECG needs to be made on a case-by-case basis and should be particularly considered when a myocardial contusion is suspected.

Urinalysis

There have been several studies that have investigated the utility of urinalysis after blunt trauma. Nicolaisen et al prospectively studied 359 consecutive patients with blunt (n = 306) or penetrating (n = 53) renal trauma to refine the indications for radiographic evaluation for renal trauma.36 They concluded that radiographic evaluation is warranted in patients with gross hematuria or microscopic hematuria and shock (< 90 mmHg) after blunt trauma but not in blunt trauma victims with microscopic hematuria without shock. Furthermore, no combination of parameters was able to predict a severe injury in patients with penetrating renal trauma. These criteria were further studied in 506 consecutive patients who presented with a history of blunt trauma and hematuria.37 Twelve percent (n = 3/25) of patients with urinary tract injuries were not detected by these criteria, all of whom had microhematuria and no shock. One patient had a minor renal laceration and two patients had renal contusions. All three patients were managed conservatively and were hospitalized chiefly for associated nonurological injuries. Four other studies have achieved similar results.38-41 The conclusion from these studies is that, in blunt trauma victims with neither gross hematuria nor shock (SBP < 90 mmHg), a urinalysis, particularly one performed in the lab vs. a bedside dipstick, contributes no worthwhile information affecting patient management.

Blood Alcohol Level and Toxicology Screen

While the American College of Surgeons’ Committee on Trauma recommends obtaining blood alcohol levels and drug screens in trauma patients, a survey indicated that, despite available resources, only 64% of Level I and 40% of Level II trauma centers routinely perform these tests on trauma victims.42 The primary reason cited in the survey for not obtaining these tests was that they are "clinically not important." Despite the fact that toxicology screens frequently do not affect patient management, arguable justification for continuing this costly practice are the epidemiological reasons and the potential for substance abuse intervention. However, there are no data to suggest an increased rate of successful substance abuse intervention at trauma centers that maintain the routine practice of drug and alcohol screening. Additionally, Sloan et al found that toxicology screen data predicted neither overall ISS data, nor the need for an emergent airway, laparotomy, or a neurosurgical procedure.43 However, the authors concluded that the directed use of such testing is warranted because, for example, the finding of an unexpectedly low or negative level in a lethargic or comatose trauma patient will prompt the more rapid investigation of neurosurgical etiologies of altered mental status. While this seems intuitive, the authors of this study present no data to support this claim. New POCT drug screening methods, including breath analyzers for detecting ethanol, represent alternatives to traditional blood alcohol and toxicology screens and afford the benefit of almost immediate results. While their impact on management remains to be proven, such studies may be of some help in patients with altered mental status.

Arterial Blood Gas and Lactate

The arterial blood gas (ABG) obtained during the initial management of trauma patients has been investigated in studies attempting to correlate its results with patient outcome. Additionally, some of these studies have suggested that ABG results are important in a real-time fashion for guiding resuscitation or indicating ongoing blood loss. Falcone et al studied preresuscitation arterial pH as a predictor of outcome of injury in 191 trauma patients.44 While they found that metabolic acidosis (i.e., pH, HCO3-, and BD) did not improve on the combination of trauma score (TS) and age for predicting outcome, it was predictive of total blood products used during the initial resuscitation. The authors conclude that their results suggested that acid-base status is important in a real-time fashion as a guide to resuscitation or as an indicator of occult ongoing blood loss. Similarly, in a retrospective study of 209 trauma patients, Davis et al found that the volume of blood transfusion increased with increasing BD (P < 0.001).45 They conclude that BD is a useful guide to volume replacement in the resuscitation of trauma patients. Rutherford et al retrospectively studied 3791 trauma patients and concluded that base deficit was a marker for significant risk of mortality.46 Their data suggest that the following two BD thresholds were markers for significant risk of mortality: greater than or equal to 15 mmol/L in a patient younger than 55 years of age without a head injury; or greater than or equal to 8 mmol/L in elderly patients (55 years or older) without a head injury or a young patient with a head injury. They conclude that the BD should be used to help guide the decision for aggressive resuscitation and early invasive hemodynamic monitoring in trauma patients. Other groups have reached similar conclusions.47,48 Finally, Roux et al prospectively studied 50 adult trauma patients and the contribution that obtaining routine laboratory, radiographic, and electrocardiographic data made to the anesthetic management of trauma patients.7 The only investigation that gave information additional to that obtained by the clinical examination was the ABG. The Hgb and serum glucose also were rarely contributory, but the electrolytes, clotting profile, full CBC, BUN, creatinine, ECG, and chest x-ray (CXR) did not make a contribution to the assessment and anesthetic management of this patient group.

While the studies described thus far suggest that many components commonly obtained in trauma panels do not contribute to the management of patients, there are data to suggest that a serum lactate level, which is infrequently obtained in most trauma panels, is useful in the management of trauma patients.49 While the association between serum lactate levels and hypo-volemic shock has been well-documented for several years, a group has only recently investigated the correlation between lactate levels and blood product usage in trauma patients.50-53 Milzman et al conducted a retrospective study of 4367 trauma patients to examine the relationship between admission blood lactate and the composition and amount of fluids given during trauma patient treatment for the first day.54 Patients were divided into five groups based on serum lactate level and compared based on fluid use. Their results are listed in Table 2.

Lactate correlated with the need for each fluid type used in resuscitation (P < 0.0001). These data suggest that lactate may be an important predictor of fluid use in trauma patients. The same author has correlated admission lactate level in adult trauma patients to mortality rates, and several other groups have correlated serum lactate concentrations to clinical outcome in trauma victims.47,48,55-59 Additionally, investigators have recently shown that elevated lactate on admission is a better triage tool than standard triage criteria over a wide range of ISS scores.60 Finally, while future investigations may show that lactate measurements obtained from peripheral venous blood specimens can supplant arterial lactates, currently there are conflicting data regarding this issue.61,62

Considering all of the data that has been published correlating certain ABG parameters and lactate levels with the amount of fluid used during resuscitation and severity of injury, it appears worthwhile to obtain at least one of these measurements in major trauma patients, particularly if results can be processed and made available rapidly. However, it needs to be emphasized that prospective studies evaluating the contribution of these tests to the management of trauma patients are lacking.

Pediatric Trauma Panel Studies

Several studies in pediatric trauma patients have paralleled those in adults. Isaacman et al conducted a two-phase study assessing the prevalence of laboratory abnormalities (CBC, electrolytes, BUN, creatinine, glucose, ALT/AST, amylase, lipase, urinalysis) and the sensitivity and specificity of the physical exam and screening laboratory tests for identifying intra-abdominal injury in moderately injured pediatric patients.63 Patients classified as moderately injured had an intact airway, stable vital signs, and a GCS score of 8-12. In phase one of their study, the authors retrospectively found a low prevalence of laboratory abnormalities in the 285 patients studied but found that a physical exam in addition to a urinalysis (UA) showing more than 5 rbc/hpf had a sensitivity of 100%, specificity of 64%, PPV of 13% and negative predictive value (NPV) of 100% for detecting intra-abdominal injury. In phase two of their study, when applying this decision rule to 91 patients known to have intra-abdominal injuries, the physical exam and UA identified an abnormality in 98% of cases. Bryant et al advocate a routine protocol of only hemoglobin/hematocrit, T&S, and UA on the basis of a retrospective review they conducted of pediatric trauma patients.64 Their rationale for obtaining the UA and T&S was not included in their report; however, after their study, amylase, platelet count, and PT/PTT were dropped from their pediatric trauma panel, since the incidence of abnormalities of these lab tests was extremely low (5.2% for amylase and 0% for platelet and PT/PTT). Finally, Ford et al reviewed the records of 100 pediatric blunt injury patients to determine the utility of leukocyte count, serum amylase, electrolytes, and UA in the management of this patient population.65 Serum amylase was found to be normal in all patients admitted, including all patients with documented intra-abdominal and retroperitoneal injuries. Leukocyte determinations were elevated in eight patients with CT-proven intra-abdominal injuries and in another 12 patients with normal CT findings. Patients uniformly had normal serum sodium, chloride, and potassium levels, and all patients had low bicarbonate determinations. Urine dipstick had an equal sensitivity to laboratory urinalysis for detecting hematuria. The authors conclude that evaluation of pediatric blunt trauma patients should begin with a thorough physical examination, and that a hematocrit determination is their only "routine" blood test in this patient population.

Summary

As resource utilization becomes increasingly scrutinized, the question remaining is whether any laboratory test battery should be recommended as a "routine protocol" across the entire spectrum of trauma patients. Considering the heterogeneity of this patient population, both in premorbid disease and acute traumatic injuries, few studies can be advocated as routine. Furthermore, it cannot be overemphasized that many of the studies cited in this review were retrospective, and prospective studies need to be conducted. However, based upon current literature, the following conclusions seem reasonable. While a normal hematocrit or hemoglobin level by no means excludes significant ongoing hemorrhage, if present, it does alert the physician to carefully evaluate the patient for a source of acute hemorrhage. Additionally, a baseline measurement provides a useful starting point for patients considered to have ongoing hemorrhage who may be followed with serial hemoglobin measurements, though it must be kept in mind that, even in the absence of blood loss, IV fluids may cause some fall in hematocrit. For this reason, a reliable measurement of hemoglobin or hematocrit should be obtained as soon as possible in virtually all victims of major trauma. In addition, several studies suggest that BD and/or a lactate level are both reliable early physiologic predictors of injury severity and may be helpful in guiding the early evaluation and management of trauma patients. If these studies can be processed promptly, the current body of evidence supports obtaining one or both of these measurements. Beyond these studies, factors such as age, acuity level, mechanism of injury, premorbid disease, areas of anatomic injury, and need for emergent surgery need to be specifically considered in each case and may help to decide what laboratory testing is likely to be of importance. Table 1 summarizes criteria for which obtaining each of the corresponding studies may be indicated.

As increasing attention and emphasis are placed on the cost-effective evaluation of trauma patients, it will be important to justify all ancillary testing that is performed. Many may argue that in the grand scheme of the costs and resources spent on treating any given trauma patient, the cost of initial ED testing comprises a small fraction of the total money spent. However, in aggregate, these cost are great, with $172 million being one estimation for total annual charges for routine initial testing at Level I trauma centers for major trauma victims.1 As studies are conducted that not only evaluate the yield of abnormal results but also the diagnostic and therapeutic management changes that result from normal or abnormal laboratory values, current recommendations may change. Additionally, studies that evaluate the impact of intervening or not intervening based on laboratory abnormalities on acceptable outcomes would add further worthwhile information to this area.

References

1. Burton J, Wolfson A, Rockoff S. Routine laboratory testing in emergency department trauma patients at Level 1 trauma centers: A national survey. Acad Emerg Med 1995;2:408.

2. Myers M, Norwood S. Standing orders for trauma care. J Emerg Nurs 1994;20:111-117.

3. Nosanchuk JS, Keefner R. Cost analysis of point-of-care laboratory testing in a community hospital. Am J Clin Pathol 1995;103:240-243.

4. Lindsley J, Eble JN. Cost analysis of point-of-care laboratory testing in a community hospital. Am J Clin Pathol 1995;104:107-108.

5. Tsai WW, Nash DB, Seamonds B, et al. Point-of-care versus central laboratory testing: An economic analysis in an academic medical center. Clin Ther 1994;16:898-910.

6. Frankel H, Rozycki G, Ochsner G, et al. Minimizing admission laboratory testing in trauma patients: Use of a microanalyzer. J Trauma 1994;37:728-736.

7. Roux A, Lourens L, Richards E. Contribution of preoperative investigations to the anaesthetic management of adult trauma patients. Injury 1993;24:17-20.

8. Tortella B, Lavery R, Rekant M. Utility of routine admission serum chemistry panels in adult trauma patients. Acad Emerg Med 1995;2:190-194.

9. Namias N, McKenney MG, Martin LC. Utility of admission chemistry and coagulation profiles in trauma patients: A reappraisal of traditional practice. J Trauma 1996;41:21-25.

10. Lowe R, Arst H, Ellis B. Rational ordering of electrolytes in the emergency department. Ann Emerg Med 1991;20:35-40.

11. Nelson E, Gilmartin D, Codd M, et al. The use of baseline investigations in patients admitted from an accident and emergency department. BMJ 1992;85:100-102.

12. Knottenbelt JD. Low initial hemoglobin levels in trauma patients: An important indicator of ongoing hemorrhage. J Trauma 1991;31:1396-1399.

13. Callaham M. Inaccuracy and expense of the leukocyte count in making urgent clinical decisions. Ann Emerg Med 1986;15:774-781.

14. Mure A, Josloff R, Rothberg J, et al. Serum amylase determination and blunt abdominal trauma. Am Surg 1991;57:210-213.

15. Olsen WR. The serum amylase in blunt abdominal trauma. J Trauma 1973;13:200-204.

16. Takahashi M, Maemura K, Sawada Y, et al. Hyperamylasemia in critically injured patients. J Trauma 1980;20:951-955.

17. Boulanger BR, Milzman DP, Rosati C. The clinical significance of acute hyperamylasemia after blunt trauma. Can J Surg 1993;36:63-69.

18. Feng C, Ng A. An analysis of donor blood wastage due to outdating in a large teaching hospital. Pathology 1991;23:195-197.

19. West H, Jurkovich G, Donnell C, et al. Immediate prediction of blood requirements in trauma victims. South Med J 1989;82:186-189.

20. Champion HR, Sacco WJ, Copes WS, et al. A revision of the trauma score. J Trauma 1995;29:623-629.

21. Clarke J, Davidson S, Bergman G, et al. Optimal blood ordering for emergency department patients. Ann Emerg Med 1980;9:9-13.

22. Hooker E, Miller F, Hollander J, et al. Do all trauma patients need early crossmatching for blood?. J Emerg Med 1994;12:447-451.

23. Davis JW, Parks SN, Kaups KL, et al. Admission base deficit predicts transfusion requirements and risk complications. J Trauma 1996;41:769-774.

24. Giblett ER. Blood group antibodies: An assessment of some laboratory practices. Transfusion 1977;17:299-308.

25. Heddle NM, O’Hoski P, Singer J, et al. A prospective study to determine the safety of omitting the antiglobulin crossmatch from pretransfusion testing. Br J Haematol 1992;81:579-584.

26. Gervin A, Fischer R. Resuscitation of trauma patients with type-specific uncrossmatched blood. J Trauma 1984;24:327-331.

27. Petz LD, Swisher, SN, eds. Clinical Practice of Transfusion Medicine. New York: Churchill Livingstone; 1989.

28. Norton M, Stanford G, Weigelt J. Early detection of myocardial contusion and its complications in patients with blunt trauma. Am J Surg 1990;160:577-582.

29. Baxter B, Moore E, Moore F, et al. A plea for sensible management of myocardial contusion. Am J Surg 1989;158:557-562.

30. Fabian T, Cicala R, Croce M, et al. A prospective evaluation of myocardial contusion: Correlation of significant arrhythmias and cardiac output with CPK-MB measurements. J Trauma 1991;31:653-660.

31. Wisner D, Reed W, Riddick R. Suspected myocardial contusion. Ann Surg 1990;212:82-86.

32. Cachecho R, Grindlinger G, Lee V. The clinical significance of myocardial contusion. J Trauma 1992;33:68-73.

33. Foil M, Mackersie R, Furst S, et al. The asymptomatic patient with suspected myocardial contusion. Am J Surg 1990;160:638-643.

34. Healey M, Brown R, Fleiszer D. Blunt cardiac injury: Is this diagnosis necessary?. J Trauma 1990;30:137-146.

35. Garland J, Wolfson A. Routine admission electrocardiography in emergency department patients. Ann Emerg Med 1994;23:275-280.

36. Nicholaisen G, McAninch J, Marshall G, et al. Renal trauma: Re-evaluation of the indications for radiographic assessment. J Urol 1985;133:183-187.

37. Hardeman S, Husmann D, Chinn H, et al. Blunt urinary tract trauma: Identifying those patients who require radiological diagnostic studies. J Urol 1987;138:99-101.

38. Guice K, Oldham K, Eide B, et al. Hematuria after blunt trauma: When is pyelography useful? J Trauma 1983;23:305-311.

39. Mee SL, McAninch JW, Robinson AL, et al. Radiographic assessment of renal trauma: A 10-year prospective study of patient selection. J Trauma 1989;141:1095-1098.

40. Klein S, Johs S, Fujitani R, et al. Hematuria following blunt abdominal trauma. Arch Surg 1988;123:1173-1177.

41. Eastham JA, Wilson TG, Ahlering TE. Radiographic assessment of blunt renal trauma. J Trauma 1991;31:1527-1528.

42. Soderstrom C, Dailey J, Kerns T. Alcohol and other drugs: An assessment of testing and clinical practices in U.S. trauma centers. J Trauma 1994;36:68-73.

43. Sloan EP, Zalenski RJ, Smith RF, et al. Toxicology screening in urban trauma patients: Drug prevalence and its relationship to trauma severity and management. J Trauma 1995;29:1647-1653.

44. Falcone RE, Santanello SA, Schulz MA, et al. Correlation of metabolic acidosis with outcome following injury and its value as a scoring tool. World J Surg 1993;17:575-579.

45. Davis JW, Shackford SR, Mackersie RC, et al. Base deficit as a guide to volume resuscitation. J Trauma 1988;28:1464-1467.

46. Rutherford EJ, Morris JA, Reed GW, et al. Base deficit stratifies mortality and determines therapy. J Trauma 1992;33:417-423.

47. Sauaia A, Moore FA, Moore EE, et al. Early predictors of postinjury multiple organ failure. Arch Surg 1994;129:39-45.

48. Siegel JH, Rivkind AI, Dalal S, et al. Early physiologic predictors of injury severity and death in blunt multiple trauma. Arch Surg 1990;125:498-508.

49. Burton JH, Wolfson AB, Rockoff S. Routine Laboratory Testing of Emergency Department Trauma Patients at Level I Trauma Centers: A National Survey. (Unpublished)

50. Mizock BA, Falk JL. Lactic acidosis in critical illness. Crit Care Med 1992;20:80-93.

51. Cady LD, Weil MH, Afifi AA, et al. Quantitation of severity of critical illness with special reference to blood lactate. Crit Care Med 1973;1:75-80.

52. Vitek V, Cowley RA. Blood lactate in the prognosis of various forms of shock. Ann Surg 1971;173:308-313.

53. Dunham CM, Siegel JH, Weireter L, et al. Oxygen debt and metabolic acidemia as quantitative predictors of mortality and the severity of the ischemic insult in hemorrhagic shock. Crit Care Med 1991;19:231-243.

54. Milzman D, Boulanger B, Wiles CE, et al. Admission lactate predicts fluid requirements for trauma victims during the initial 24 hours. Crit Care Med 1994;22:A73.

55. Milzman D, Menlove S, Boulanger B, et al. Admission lactate: A rapid predictor of survival following traumatic injury. Ann Emerg Med 1995;21:596.

56. Aduen J, Bernstein WK, Khastgir T, et al. The use and clinical importance of a substrate-specific electrode for rapid determination of blood lactate concentrations. JAMA 1994;272:1678-1685.

57. Abou-Khalil B, Scalea TM, Trooskin S, et al. Hemodynamic responses to shock in young trauma patients: Need for invasive monitoring. Crit Care Med 1994;22:633-639.

58. Abramson D, Scalea TM, Hitchcock R, et al. Lactate clearance and survival following injury. J Trauma 1993;35:584-589.

59. Grzybowski M, Rivers EP, Tuttle AP, et al. Prevalence of lactic acidosis on ED presentation and association with hospitalization rates. Acad Emerg Med 1996;3:479.

60. Livingston DH, Lavery RF, Tortella BJ, et al. Lactate identifies major trauma better than standard triage criteria. Acad Emerg Med 1996;3:532.

61. Bernstein WK, Aduen J, Bhatiani A, et al. Simultaneous arterial and venous lactate determinations in critically ill patients. Crit Care Med 1995;22:A227.

62. Rodriguez K, Gallagher EJ, Touger M. Prediction of arterial lactate levels from peripheral venous blood. Acad Emerg Med 1995;2:468-469.

63. Isaacman D, Scarfone R, Kost S, et al. Utility of routine laboratory testing for detecting intra-abdominal injury in the pediatric trauma patient. Pediatrics 1993;92:691-694.

64. Bryant M, Tepas J, Talbert J, et al. Impact of emergency room Laboratory studies on the ultimate triage and disposition of the injured child. Am Surg 1988;54:209-211.

65. Ford E, Karamanoukian H, McGrath N, et al. Emergency center laboratory evaluation of pediatric trauma victims. Am Surg 1990;56:752-757.

Physician CME Questions

17. Which of the following is recommended as routine testing on all major trauma patients?

A. Hemoglobin or hematocrit

B. WBC

C. Electrolyte panel

D. Potassium

E. BUN/Creatinine

18. Severely head injured patients should have at least which of the following performed?

A. WBC

B. Potassium

C. BUN/Creatinine

D. PT/PTT

E. Type and crossmatch

19. In an adult sustaining blunt renal trauma without gross hematuria, urinalysis is indicated when which of the following historical or clinical features are present?

A. Fall from a height of greater than 10 feet

B. Shock (SBP < 90 mm Hg)

C. Flank tenderness

D. Flank pain

E. Abdominal tenderness

20. Regarding blood bank utilization in major trauma, which of the following is correct?

A. Prehospital hypotension may be associated with the likelihood of transfusion.

B. Base deficit less than or equal to -6 may be an appropriate threshold for obtaining at least a type and screen.

C. The risk of a major hemolytic reaction from transfusing blood that has been typed but not crossmatched is very small.

D. The main priority in trauma patients should be to ensure the availability of type-specific blood as soon as possible.

E. All of the above.

21. What is the most frequently cited reason trauma centers do not perform toxicology screens routinely?

A. Results are not available soon enough

B. Legal restrictions

C. Fears of litigation

D. Inability to bill for the test

E. Results are clinically not important

22. Which of the following tests is least helpful in the initial evaluation of trauma patients?

A. Hemoglobin

B. Potassium

C. Glucose

D. Amylase

E. Lactate

23. Regarding measuring electrolytes routinely in trauma patients, which of the following is true?

A. Sodium and chloride are often abnormal.

B. In published studies, hyperkalemia appears more prevalent than hypokalemia.

C. Pediatric studies have found high rates of abnormal results.

D. Age older than 50 appears to be an appropriate age threshold for routinely measuring electrolytes.

24. Regarding BD and lactate, which is true?

A. Volume of blood transfusion has been found to increase with increasing BD.

B. Volume of blood transfusion has been found to increase with increasing lactate.

C. BD has been found to be a marker for significant risk of mortality.

D. Lactate level has been found to correlate with mortality rates in adult trauma patients.

E. All of the above