Central to all good antibiotic (or any other) therapy is diagnosis. Whenever an antibiotic is prescribed it should be for a specific diagnosis whether it is a tentative clinical diagnosis or not. An antibiotic should not be prescribed for a “urinary tract infection” it should be prescribed for cystitis, pyelonephritis or prostatitis as an example. Prescribing and antibiotic for a “cough” is ludicrous. Cough could be a symptom of many conditions with pneumonia or severe COPD exacerbation being the only specific conditions for which antibiotic therapy is indicated.
Each time you contemplate prescription of an antibiotic decide upon a specific diagnosis. Your antibiotic choices can then be informed by trusted therapy guidelines such as these created in the Vancouver Island Health Authority.
Principles of Antibiotic Therapy
While the issue of antibiotic misuse and resistance is a large one, antibiotics are life-saving therapy. Several general principles can help you make decisions about initiating, selecting, and discontinuing antibiotic therapy.
- Error on the side of not prescribing when diagnosis is unclear for those who are not seriously unwell and are immuno-competent. These patients are in the majority in primary care.
- Error on the side of prescribing when diagnosis is unclear for those seriously ill or immunocompromised (many of these patients should considered for hospital care)
- Use the narrowest spectrum antibiotic possible
- Use the least expensive effective antibiotic
- Use the shortest effective duration
- Utilize local, trusted guidelines for empiric therapy
- In serious, life threatening infections – “Broad” empiric therapy is appropriate. Hopefully microbiology will help narrow the spectrum. “Shoot first and ask questions later”. Rapid administration is also very important. Fortunately, these circumstances are quite infrequent.
- In non – life threatening circumstances “covering” for all possible pathogens is neither appropriate nor desirable.
General introduction to antibiotics can be viewed here and the following are short discussions of the major classes and individual antibacterial antibiotics that are in common usage in Canada.
Groupings of Bacteria for Rational Antibiotic Selection
This section is not “textbook” information but is an attempt to explain my way of thinking about possible infecting organisms in clinical circumstances and how creating a list of potential groups of organisms influences the selection of empiric therapy.
Gram positive Bacteria
The term “Gram-positive coverage” is often used when describing the spectrum required for a particular clinical situation. It is a meaningless term and should be avoided. Consideration must be given to the potential for specific Gram positive organisms to be involved.
If S. aureus is high on the list of potential pathogens, as it is for all skin and soft tissue infections, empiric coverage must include an agent that is reliably active. Cloxacillin, cefazolin, and clindamycin are examples of such therapy in Canada.
However, methicillin resistance Staphylococcus aureus (MRSA) is increasingly pushing us toward the limited empiric choices of intravenous-only vancomycin or the new oxyzolidinone linezolid at $140.00 per day. This is a very good example of why prudent antibiotic use and good routine infection control practices are so valuable. See Community-associated MRSA from the treatment guidelines section.
Coagulase-negative Staphylococci (CNS)
S. epidermidis and other CNS colonize all people and rarely cause severe disease. As such, they are not considered potential pathogens in most clinical circumstances. However, when patients develop infections in the setting of indwelling intravascular catheters or other prosthetic devices (heart valves, hips, knees etc.) CNS are common pathogens.
Methicillin resistance in S. epidermidis is very common worldwide and empiric choices are limited to the glycopeptides (vancomycin and teichoplanin) and oxyzolidinones (linezolid) as with MRSA. The primary mode of therapy is often removal of the prosthetic material if possible.
Enterococcus spp. are increasing in importance as nosocomial pathogens and are inherently resistant to antibiotics. Some isolates are ampicillin susceptible but the only predictably effective therapies are vancomycin and oxyzolidinones and resistance is already a problem in some areas in the case of glycopeptides (so-called vancomycin resistant enterococci or VRE).
This organism deserves special attention because it is the fourth most common cause of bacterial meningitis and is resistant to cephalosporins – the most commonly advocated therapy for meningitis. For this reason a second agent (ampicillin) is necessary.
No single organism is driving change in outpatient empiric therapy recommendations more than S. pneumoniae. Resistance inexorably increases and knowledge of local susceptibility is essential in determining appropriate empiric treatment for conditions potentially caused by this organism.
Gram negative Bacteria
Similar to the situation with Gram positive bacteria, the oft used term “Gram negative coverage” is meaningless because specific Gram negatives and groups of Gram negatives have varying susceptibility profiles and recommended empiric therapies.
While resistance increases, E. coli tends to be the least resistant of the Gram negative enteric organisms. Many antibiotics that are used for the therapy of hospital acquired infections are active against this organism. Local susceptibility profiles guide empiric recommendations for therapy for outpatient urinary tract infections.
“Nasty” Gram negatives
I use this term to group the hospital-associated, antibiotic-resistant members of the family Enterobacteriaciae namely Klebsiella, Enterobacter, Serratia, Citrobacter and Proteus, and Morganella. Hospital patients become colonized with these organisms and they are common causes of nosocomial infections, in particular pneumonia and urinary tract infections.
In many scenarios carbapenems are the only reliable empiric therapy and resistance to this class is increasing.
This is the prototypic “non-fermenter”, an environmental organism that only causes disease in opportunistic circumstances. However, the disease it causes can be very severe and P. aeruginosa is inherently resistant to antibiotics. If this organism is on the list of potential pathogens, such as in the case of febrile neutropenia, options are limited. Antipseudomonal penicillins (piperacillin and ticarcillin), antipseudomonal cephalosporins (ceftazidime and cefepime) and carbapenems (imipenem and meropenem) are the only reliable therapies. Therapy often includes one of these agents plus an aminoglycoside (gentamicin or tobramycin). When you are “covering” for P. aeruginosa you are giving very broad spectrum therapy.
The empiric selection of antibiotics for the treatment of anaerobic infections is relatively easy. Metronidazole is the “workhorse” agent with most important anaerobes susceptible and very few patients allergic or intolerant. Antibiotic therapy is often secondary in importance to surgery for serious anaerobic infections. Clindamycin is another agent often employed for its activity against many anaerobic bacteria.
Mycoplasma spp., Chlamydia spp. and Legionella spp. Only particular classes of antibiotics (esp. macrolides, tetracyclines and fluoroquinolones) are reliably active.