Hospital Medicine

Antibiotic selection

I. Problem/Challenge.

Infections are a common cause of hospital admissions and appropriately identifying the infection and pathogen are desirable for optimal resolution. At times, empiric antimicrobial therapy must be instituted, and knowledge of common pathogens typical for a suspected site of infection, and corresponding effective antimicrobials is necessary while awaiting definitive diagnoses.

II. Identify the Goal Behavior

The goal of antimicrobial therapy is to correctly identify infection and then select the most appropriate anti-infective that can sufficiently treat the infection. This, while being as narrow spectrum as possible, providing the best penetration into the infected organ, while minimizing adverse reactions, including further organ damage or risks due to antimicrobial therapy itself, such as need for intravenous access, renal toxicity, or acquisition of superinfection such as Clostridium difficile colitis.

III. Describe a Step-by-Step approach/method to this problem.

Correctly identify this is an infectious disease process. Fever and laboratory signs such as leukocytosis can be caused by non-infectious processes. Even positive cultures can reflect contamination rather than true infection. In this patient’s clinical context, with the constellation of symptoms they are presenting with, including their given history and co-morbidities, is infection likely?

Choose the spectrum of antibiotic coverage. Initially and particularly in a very sick patient, empiric broad-spectrum antibiotic treatment should be initiated. When selecting empiric antibiotics consider the following:

  • The site of infection and the organisms most commonly colonizing that site. When determining the likely source of infection, patient characteristics must be defined. Is the patient immunocompromised or diabetic? Does the patient have any foreign bodies or have they undergone a recent invasive procedure? Does he or she have advanced age? Have they recently been in a healthcare facility? Does the patient have any concerning exposures or travel?

  • Is the patient known to have been infected with particular organisms in the past and were they treated with a particular antibiotic previously?

  • What are the local resistance patterns within the community or hospital?

Once the specific pathogen has been identified, narrow the antibiotic agent according to organism and susceptibility testing. Always communicate the source of the culture to the microbiology lab, as work-up of the culture can vary depending on the source.

  • Minimal inhibitory concentration (MIC) is the lowest concentration of an antibiotic that inhibits visible growth of an organism.

  • MICs are tested in vitro, assume bloodstream infection and are tested at recommended dosages

  • MICs are particular to individual antibiotics and cannot be directly compared between antibiotics

Choose an antibiotic with adequate tissue penetration for the site of infection. Tissue penetration is based on properties of both the antibiotic (i.e. lipid solubility, molecular size, protein-binding, and volume of distribution) and target organ (adequate blood supply, presence of inflammation, pH at site of infection, and anatomic barriers). Be mindful that some sites such as ocular fluid, cerebrospinal fluid (CSF), abscess cavity, prostate, and bone achieve much lower concentrations of antimicrobials than plasma concentrations.

In general, tissue penetration is easier in acute infections, because of increased vascular permeability, than in chronic infections (i.e. chronic osteomyelitis) where blood flow is often compromised or necrotic tissue is present.

Decide optimal dosing of antibiotics. This is based on both the pharmacokinetics and pharmacodynamics of the medications themselves, along with the hepatic and renal function of the patient since most medications are metabolized and/or cleared in these organs.

Antibiotic pharmacokinetics and pharmacodynamics are categorized in the following ways:

  • Concentration-dependent, meaning that as the concentration is increased, the rate and extent of killing bacteria is also increased because they have a prolonged post-antibiotic effect. Thus, it is more important to give a higher dose less frequently than the same dose divided over intervals. Examples of concentration-dependent antimicrobials are aminoglycosides, fluoroquinolones, metronidazole and daptomycin.

  • Time-dependent, meaning that these medications have little post-antibiotic effect, and so more frequent dosing with increased exposure to the antibiotic will produce more effective results. Examples of time-dependent antimicrobials are beta-lactams, including carbapenems and cephalosporins;, erythromycin, sulfonamides, and linezolid.

  • Mixed concentration and time-dependent. These medications generally follow time-dependent killing properties, but also exhibit some moderately prolonged post-antibiotic effect. The ideal dosing for these medications is to maximize the amount of medication given. Examples in this category include vancomycin, tetracyclines and azithromycin.

Dose adjustments for renal insufficiency are based on creatinine clearance and vary depending on the antibiotic, and are particularly important in those medications with a low therapeutic index. Usually, the initial loading dose is the same as for the general population and the maintenance dose is reduced. Unfortunately, there is no comparative measure like creatinine clearance for hepatic failure. However, for patients with severe liver disease, if a hepatically-cleared antimicrobial must be used, the dose should be reduced by 50%.

Selecting between bacteriostatic and bactericidal antibiotics is generally less significant in vivo, though in general, for sick patients a bactericidal agent should be utilized. Exceptions to this are serious infections like meningitis, endocarditis, and febrile neutropenia. In these instances bactericidal agents are preferred to achieve a rapid cure.

Consider host factors and safety profile of medications.

  • Adjust dosages based on renal and hepatic function.

  • When possible, avoid medications with serious and/or frequent side effects, for example, immediate immunoglobulin E (IgE) mediated hypersensitivity reactions.

  • Special considerations must be given for the use of antimicrobials in pregnant and lactating women.

Determine monotherapy versus combination therapy. In general, monotherapy is preferred, however combination should be considered in a few scenarios:

  • For synergy, meaning that when tested in vitro, the combined effects of agents are greater than the sum of the independent activities when measured separately. While synergy is frequently demonstrated in vitro, there are only a few clinical scenarios where benefit has been described, for example, the use of an aminoglycoside along with penicillin for enterococcal endocarditis or the use of trimethoprim with a sulfonamide.

  • When empirically treating critically ill patients, and especially when there is potential resistance; in this case, the hope is that at least one of the antimicrobials will be effective against the organism.

  • When the spectrum must be expanded to cover likely polymicrobial infections (i.e. intra-abdominal infections).

  • Rarely, to prevent emergence of resistance. There are very few instances in which combination therapy prevents resistance such as combination therapy for tuberculosis.

Choose mode of therapy. Hospitalized patients with serious bacterial infections should be started on intravenous (IV) therapy. However, a switch from IV to oral therapy should be instituted as soon as the patient is stable and able to tolerate orals to avoid increased cost, decrease length of stay and avoid complications of IV lines.

When switching to oral therapy, the oral antibiotic should cover the same spectrum of organisms of the suspected/ known pathogen and should have excellent bioavailability. Exceptions to this include serious infections, which require high concentrations within targeted organs (i.e. meningitis or endocarditis); in this case, IV therapy is preferred for the duration of treatment.

The following is a partial list of medications based on bioavailability:

  • Excellent (greater than 90%): Quinolones, trimethoprim/sulfamethoxazole (TMP-SMX), metronidazole, clindamycin, doxycycline, minocycline, , fluconazole, rifampin, linezolid, voriconazole, and chloramphenicol.

  • Good (60-90%): Most beta-lactams, cefixime, cefpodoxime, cefuroxime, macrolides, cefaclor, valacyclovir, famciclovir, valganciclovir, itraconazole solution, 5-flucytosine, and nitazoxanide (with food).

  • Poor (less than 60%): Vancomycin, cefdinir, cefditoren, and nitazoxanide (without food).

Note: excellent bioavailability means that oral administration equals blood/tissue levels as that in IV form; good means that oral administration results in lower, but effective blood/tissue levels than IV dose; poor means suboptimal blood/tissue levels.

Duration of therapy. Little data exists on optimal duration of treatment and most recommendations are based on expert opinion rather than controlled studies. There is evidence for treating uncomplicated urinary tract infections (UTIs) in women for 3 days and community-acquired pneumonia for 5 days.

Given potential adverse reactions, a shorter duration of treatment is preferred, and in general most infections can be treated for 1-2 weeks. Longer duration of treatment (4-6 weeks) is warranted for certain infections, including endocarditis, osteomyelitis, brain abscesses, and mycobacterial infections.

Monitoring serum levels. Since most antimicrobials have a wide therapeutic index (ratio of toxic to therapeutic dose), there is no indication for routine monitoring of serum levels. However, a few medications with narrow therapeutic index require monitoring. Most common among these are aminoglycosides to monitor for toxicity at high levels and vancomycin to evaluate for treatment failure at subtherapeutic levels.

IV. Common Pitfalls.

  • Treating contaminants or colonizers rather than true pathogens

  • Underdosing medications

  • Choosing drugs with poor diffusion into the target site

  • Missing a pocket of localized pus/abscess

  • Inappropriate drug or route of administration

V. National Standards, Core Indicators and Quality Measures.

Effective October 2008, Centers for Medicare and Medicaid Services (CMS) guidelines state that certain hospital-acquired infections will no longer be eligible for reimbursement. Included in these are catheter-associated urinary tract infections and catheter-associated bloodstream infections.

What's the Evidence?

"Centers for Disease Control and Prevention. “Antibiotic Resistance Threats in the United States, 2013.”".

Eliopoulos, GM, Moellering, RC, Bennett, JE, Dolin, R, Blaser, MJ. "“Principles of Anti-infective Therapy.”". Mandell, Douglas, and Bennett's Principles and Practice of Infectious Diseases. Saunders. 2014.

Morgan, DJ, Croft, LD, Deloney, V, Popovich, KJ. "“Choosing Wisely in Healthcare Epidemiology and Antimicrobial Stewardship.”". Infect Control Hosp Epidemiol. vol. 28. 2016 Mar. pp. 1-6.

You must be a registered member of ONA to post a comment.

Sign Up for Free e-newsletters

Regimen and Drug Listings


Bone Cancer Regimens Drugs
Brain Cancer Regimens Drugs
Breast Cancer Regimens Drugs
Endocrine Cancer Regimens Drugs
Gastrointestinal Cancer Regimens Drugs
Genitourinary Cancer Regimens Drugs
Gynecologic Cancer Regimens Drugs
Head and Neck Cancer Regimens Drugs
Hematologic Cancer Regimens Drugs
Lung Cancer Regimens Drugs
Other Cancers Regimens
Rare Cancers Regimens
Skin Cancer Regimens Drugs