EMPIRIC ANTIMICROBIAL THERAPY
Antimicrobial agents are frequently used before the pathogen responsible for a particular illness or the susceptibility to a particular antimicrobial agent is known. This use of antimicrobial agents is called empiric (or presumptive) therapy and is based on experience with a particular clinical entity. The usual justification for empiric therapy is the hope that early intervention will improve the outcome; in the best cases, this has been established by placebo-controlled, double-blind prospective clinical trials. For example, treatment of febrile episodes in neutropenic cancer patients with empiric antimicrobial therapy has been demon-strated to have impressive morbidity and mortality benefits even though the specific bacterial agent responsible for fever is deter-mined for only a minority of such episodes.
Finally, there are many clinical entities, such as certain epi-sodes of community-acquired pneumonia, in which it is diffi-cult to identify a specific pathogen. In such cases, a clinical response to empiric therapy may be an important clue to the likely pathogen.
Frequently, the signs and symptoms of infection diminish as a result of empiric therapy, and microbiologic test results become available that establish a specific microbiologic diagnosis. At the time that the pathogenic organism responsible for the illness is identified, empiric therapy is optimally modified to definitivetherapy, which is typically narrower in coverage and is given foran appropriate duration based on the results of clinical trials or experience when clinical trial data are not available.
Initiation of empiric therapy should follow a specific and system-atic approach.
Using all available data, the clinician should determine that there is anatomic evidence of infection (eg, pneumonia, cellulitis, sinusitis).
Examination of stained specimens by microscopy or simple examination of an uncentrifuged sample of urine for white blood cells and bacteria may provide important etiologic clues in a very short time. Cultures of selected anatomic sites (blood, sputum, urine, cerebrospinal fluid, and stool) and nonculture methods (antigen testing, polymerase chain reaction, and serology) may also confirm specific etiologic agents.
The history, physical examination, and immediately available laboratory results (eg, Gram stain of urine or sputum) may pro-vide highly specific information. For example, in a young man with urethritis and a Gram-stained smear from the urethral meatus demonstrating intracellular gram-negative diplococci, the most likely pathogen is Neisseria gonorrhoeae. In the latter instance, however, the clinician should be aware that a significant number of patients with gonococcal urethritis have uninformative Gram stains for the organism and that a significant number of patients with gonococcal urethritis harbor concurrent chlamydial infection that is not demonstrated on the Gram-stained smear.
Whether or not to initiate empiric therapy is an important clinical decision based partly on experience and partly on data from clini-cal trials. Empiric therapy is indicated when there is a significant risk of serious morbidity if therapy is withheld until a specific pathogen is detected by the clinical laboratory.
In other settings, empiric therapy may be indicated for public health reasons rather than for demonstrated superior outcome of therapy in a specific patient. For example, urethritis in a young sexually active man usually requires treatment for N gonorrhoeae and Chlamydia trachomatis despite the absence of microbiologic confirmation at the time of diagnosis. Because the risk of noncom-pliance with follow-up visits in this patient population may lead to further transmission of these sexually transmitted pathogens, empiric therapy is warranted.
Selection of empiric therapy may be based on the microbiologic diagnosis or a clinical diagnosis without available microbiologic clues. If no microbiologic information is available, the antimicro-bial spectrum of the agent or agents chosen must necessarily be broader, taking into account the most likely pathogens responsible for the patient’s illness.
Selection from among several drugs depends on host factors that include the following: (1) concomitant disease states (eg, AIDS, neutropenia due to the use of cytotoxic chemotherapy, organ transplantation, severe chronic liver or kidney disease) or the use of immunosuppressive medications; (2) prior adverse drug effects; (3) impaired elimination or detoxification of the drug (may be genetically predetermined but more frequently is associated with impaired renal or hepatic function due to underlying disease); (4) age of the patient; (5) pregnancy status; and (6) epidemiologic exposure (eg, exposure to a sick family member or pet, recent hospitalization, recent travel, occupational exposure, or new sexual partner).
Pharmacologic factors include (1) the kinetics of absorption,distribution, and elimination; (2) the ability of the drug to be delivered to the site of infection; (3) the potential toxicity of an agent; and (4) pharmacokinetic or pharmacodynamic interactions with other drugs.
Knowledge of the susceptibility of an organism to a specific agent in a hospital or community setting is important in the selec-tion of empiric therapy. Pharmacokinetic differences among agents with similar antimicrobial spectrums may be exploited to reduce the frequency of dosing (eg, ceftriaxone may be conveniently given once every 24 hours). Finally, increasing consideration is being given to the cost of antimicrobial therapy, especially when multiple agents with comparable efficacy and toxicity are available for a specific infection. Changing from intravenous to oral antibiotics for prolonged administration can be particularly cost-effective.
Brief guides to empiric therapy based on presumptive microbial diagnosis and site of infection are given in Tables 51–1 and 51–2.