Bedside measurement of D-dimer in the identification of bacteremia in the Emergency Department

Abstract

The objective of this pilot study was to determine the clinical utility of the SimpliRed D-dimer bedside assay to identify patients with bacteremia in a university hospital Emergency Department. We tested 265 patients and compared blood culture results with a novel D-dimer semiquantitative whole blood assay. Bacteremia was confirmed in 25/262 patients. Sensitivity of D-dimer assay was 66.7% for Gram-positive bactermia and 61.5% for Gram-negative bacteremia with negative predictive value of 98% for Gram-positive and 96% for Gram-negative bacteremia patients. Measurement of D-dimer appears to be of value in identifying patients at low risk for bacteremia and can be accomplished rapidly using a whole blood semiquantitative bedside assay. Although increases in D-dimer are not detected in all patients subsequently documented to have bacteremia on a single sampling, the results of this and other earlier studies suggest assay of D-dimer is useful in rapid differentiation of patients with bacteremia from those who have no bacteremia using blood culture positivity as the standard for bacteremia.

Introduction

Sepsis occurs with an estimated incidence of 70,000 to 300,000 cases per year and is associated with a mortality of approximately 50% due to complications of hypotension and shock, adult respiratory distress syndrome and disseminated intravascular coagulation (1). Bone et al. define the sepsis syndrome as the presence of infection plus a systemic response inducing tachypnea, tachycardia, and hypothermia or hyperthermia (1). It is often preceded or triggered by bacteremia. Bacteremia, defined as the presence of viable bacteria in the blood, occupies an area of overlap between infection, sepsis, and the systemic inflammatory response syndrome as defined by Bone et al. (2). Though the sepsis syndrome is often preceded by Gram-negative bacteremia, the 1990s saw a rise in sepsis induced by Gram-positive bacteremia, particularly in the in-patient setting. This may be due to increased line sepsis from vascular access and other invasive procedures. Infection is defined as a microbial or viral invasion characterized by an inflammatory response to the presence of microorganisms in normally sterile host tissue. Sepsis is defined as the systemic response to infection and is manifested by two or more of the following conditions as a result of infection: 1) temperature > 38°C or < 36°C; 2) heart rate > 90 beats/min; 3) respiratory rate > 20 breaths/min or Pa CO₂ < 32; torr 4) WBC > 12,000 cells/mm³, < 4000 cells/mm³, or > 10% immature (band) forms. Septicemia refers to either organisms, toxins, or breakdown products in the blood. The systemic inflammatory response syndrome (SIRS) is defined as the response to a variety of severe clinical insults as manifested by the same set of conditions resulting from infection but caused by trauma, burns, pancreatitis, or other severe clinical conditions besides infection (2).

Though physicians deal with these pathologic classifications as if they are static and distinct categories, in reality they represent a dynamic clinical spectrum. Infection may lead to bacteremia, which may lead to sepsis; pancreatitis may lead to SIRS and sepsis may complicate the clinical picture. The accurate and timely diagnosis of sepsis at the bedside continues to elude physicians. The prospective identification of high-risk patients who will experience a complicated hospital course from sepsis or other infections is problematic in the Emergency Department (ED) setting. General measurement of clinical parameters is not sufficiently sensitive, and no specific test available to the Emergency Physician is helpful in making these decisions. In their work in defining the sepsis syndrome, Bone et al. provide evidence that it is possible to identify a group of patients with a high risk of developing life-threatening sepsis or septicemia, although the rate of blood culture positivity in their study population was only 45% (2). Others have disputed the validity of Bone’s approach, arguing that it is quite clear that a patient can fulfill the criteria of the sepsis syndrome whereas, in fact, the patient is suffering from a drug reaction, tissue necrosis related to surgery or trauma, pancreatitis, myeloproliferative disorders, or vasculitis (3).

Doyle and Rivers et al. employed the Acute Physiology and Chronic Health Evaluation II (APACHE II) and the Simplified Acute Physiology II (SAPS II) scores derived from a combination of age, admitting diagnosis, chronic health history, vital signs, Glasgow Coma Score, respiratory variables, and laboratory variables to objectively evaluate illness severity, evaluate efficacy of ED intervention, and predict mortality in the ED setting. They were able to demonstrate that the APACHE II and SAPS II scores objectively show that ED intervention significantly decreases morbidity and predicts in-house mortality for critically ill ED patients (4).

In light of these challenging dilemmas, the identification of an appropriate screening test that would have predictive value is of significant clinical relevance. Objective markers, which may determine the position of the patient within the clinical spectrum of sepsis and the sepsis syndrome, may assist the physician in defining the patient’s illness along that spectrum and aid in both diagnostic testing and selection of treatment recognized as efficacious for the given condition. One would expect that, over time, objective collection of standardized data points would begin to allow prediction of clinical outcomes.

Although the diagnosis of sepsis depends on blood culture (BC) growth of the specific pathogenic species of bacteria, the bacterial sepsis syndrome may also occur in patients in whom blood cultures do not yield bacterial growth. The sepsis syndrome in such instances may reflect infection at a distant site that isolates bacteria but not endotoxin from blood, or bacteremia that is not present at the time blood is sampled for culture.

Because there is a delay before blood culture results are available, empiric antimicrobial therapy is indicated whenever bacterial sepsis is suspected. Although prompt treatment of potential bacterial sepsis is a reasonable strategy, prophylactic treatment with broad-spectrum antibiotics increases the risk of antibiotic resistance and exposes the patient to potential toxic side effects of antibiotics.

Because the diagnosis of bacterial sepsis is sometimes difficult to ascertain definitively at the bedside, an objective tool to aid physicians in the bedside diagnosis of sepsis could be beneficial in directing therapy. In a prospective pilot study using a novel bedside D-dimer assay, SimpliRed (AGEN Biomedical, Ltd., Australia), Deitcher and Eisenberg demonstrate a high negative predictive value for Gram-negative bacteremia (5). Results of that study suggest that patients in whom D-dimer is not elevated have a low likelihood (95% negative predictive value) of having subsequently documented Gram-negative bacteremia. In their study, the SimpliRed assay, which is based on a semi-quantitative assessment of red cell agglutination, was performed through a centralized laboratory by trained technicians. The sensitivity of this assay depends on the accurate assessment of even minimal increases in D-dimer, a cross-linked fibrin degradation product that is induced by activation of the coagulant and fibrinolytic systems by endotoxin 6, 7. The SimpliRed assay is designed so that clearly visible hemagglutination occurs within 3 min, when cross-linked fibrin degradation product concentration exceeds 200 ng/mL. (3) The results of earlier studies indicate SimpliRed D-dimer assay is a sensitive measurement of cross-linked fibrin degradation products, which increase early in patients with Gram-negative bacteremia and in many patients with other forms of bacteremia (5). Thus, in the ED setting, a measurement of cross-linked fibrin degradation products with the SimpliRed assay, which permits rapid bedside measurement of ordinary whole blood, should be useful in identifying patients with a low likelihood of having community-acquired bacteremia in the ED. Although the SimpliRed semiquantitative assay is slightly less sensitive than the ELISA, the SimpliRed assay can be performed at the bedside in less than 3 min, and results in a high negative predictive value (96%) for Gram-negative bacteremia (8). Deitcher and Eisenberg also find that both the ELISA and the SimpliRed assay have a greater predictive value for Gram-negative bacteremia than previously studied criteria based on clinical and laboratory variables such as temperature, presence of shaking chills, intravenous (i.v.) drug use, and major co-morbidity (5).

Deitcher and Eisenberg also observe that standard coagulation assays and platelet counts are of less value than measurement of cross-linked fibrin degradation products in the initial evaluation of patients with suspected bacteremia.

Often underlying illness or anticoagulant therapy contribute to the abnormal coagulation parameters. The main limitation of measurement of cross-linked fibrin degradation products in an in-patient population is the lack of specificity (5). Hospitalized patients may exhibit modest increases in fibrinolytic activity in response to non-bacteremic conditions such as intravascular thrombosis, inflammatory conditions, and the post-operative state 9, 10.

Identification of patients at low-risk for bacteremia may permit stratification of the intensity and type of antimicrobial therapy used. The objective of this study was to determine the clinical utility of the SimpliRed assay for identifying patients with bacteremia in a university hospital ED population characterized by a relatively low prevalence of bacteremic patients. A second purpose was to determine whether the assay could be performed by attending Emergency Physicians in the ED to define the utility of the assay as a point of care test.