Antimicrobial Stewardship: Principles and Practice (Table 1.14). ……………………………… 132
Overview of Antimicrobial Therapy
FACTORS IN ANTIBIOTIC SELECTION
Antibiotic spectrum refers to the range of microorganisms an antibiotic is usually effective against, and is the basis for empiric antibiotic therapy.
In vitro susceptibility does not always predict in vivo effectiveness.
Antibiotics unable to reach the site of infection will be ineffective. Antibiotic tissue penetration depends on antibiotic properties, e.g., lipid solubility, molecular size, adequacy of blood supply and presence of inflammation.
Antibiotics cannot be expected to eradicate organisms from difficult to penetrate areas. Abscesses usually require surgical drainage for cure.
Device associated infections usually need device removal for cure.
Bacterial resistance to antimicrobial therapy may be classified as natural/intrinsic or acquired relative or absolute.
Pathogens not covered by the usual spectrum of an antibiotic are termed naturally/intrinsically resistant, e.g., 25% of S. pneumoniae are naturally resistant to macrolides.
Acquired resistance refers to a previously susceptible pathogen that is no longer susceptible to an antibiotic, e.g., ampicillin resistant H. influenzae. Organisms with intermediate level (relative) resistance manifest as an increase in minimum inhibitory concentrations (MICs), but are susceptible if achievable serum/tissue concentrations > MIC, e.g., penicillin resistant S. pneumoniae and PCN.
Organisms with high level (absolute) resistance cannot be overcome by higher-than-usual antibiotic doses, e.g., gentamicin resistant P. aeruginosa.
Most acquired antibiotic resistance is agent specific, not class related resistance (usually limited to one or two species).
Resistance is not related, per se, to volume or duration of use, e.g., doxycycline, ceftriaxone.
Some antibiotics have little resistance potential, i.e., “low resistance” potential even when used in high volume. Other antibiotics can induce resistance, e.g., “high resistance” potential even with limited use.3
Resistance Potential of Selected Antibiotics
“High Resistance Potential” Antibiotics to Avoid
Usual Organism(s) Resistant to Each Antibiotic
Preferred “Low Resistance Potential” Antibiotic Alternatives in Same Class
Preferred “Low Resistance Potential” Antibiotic Alternatives in Different Classes
Gentamicin or Tobramycin
Levofloxacin, Colistin, Cefepime
Levofloxacin, Colistin, Polymyxin B
S. pneumoniae S. aureus
Amikacin, Colistin, Cefepime
Linezolid, Daptomycin, Minocycline, Tigecycline
Amikacin, Cefepime, Colistin, Polymyxin B
Doxycycline, Levofloxacin, Moxifloxacin
Dihydrofolate Reductase Inhibitors
Doxycycline, Levofloxacin, Moxifloxacin
Whenever possible, avoid antibiotics with serious/frequent side effects.
Mode of Antibiotic and Excretion/Excretory Organ Toxicity
The mode of elimination/excretion does not predispose to excretory organ toxicity per se, e.g., nafcillin (hepatically eliminated) is not hepatotoxic, and it's main side effect is nephrotoxicity (interstitial nephritis). In contrast, oxacillin (renally eliminated), is not nephrotoxic, it's main side effect is hepatotoxicity (hepatitis).
The primary route of antibiotic elimination is protective and does not predispose to excretory organ toxicity.
Switching early from IV to PO antibiotics is the single most important cost saving strategy in hospitalized patients.
Institutional cost of IV administration (~ $10/dose) may exceed the cost of the antibiotic.
Antibiotic costs can also be minimized by using antibiotics with long half-lives, by choosing monotherapy over combination therapy, and decreasing duration of therapy.
HEMODIALYSIS (HD) DOSING STRATEGIES
For renally eliminated antibiotics, begin therapy with the usual initial dose, then decrease the maintenance doses/intervals for a CrCl < 10 mg/min (for specific antibiotics, seeChapter 11, Drug Summaries).
For antibiotics partially/totally removed by HD, a post-HD is needed (for specific antibiotics, seeChapter 11, Drug Summaries).
HD Dosing Only
For selected antibiotics with the requisite PK/PD properties, treatment may be given at each HD (Table 1.2).
Selected Antibiotics for Dialysis Dosing (Regimens that do not require Intra-HD Dosing)
2 gm (IV) q HD
2 gm (IV) q HD5
2 gm (IV) q HD
6-12 mg/kg (IV) q HD*
500 mg (IV/PO) q HD
P. aeruginosa, Aerobic GNBs, MSSA, VSE
2 gm (IV) q HD
* For bacteremia due to MSSA/MRSA use 6 mg/kg, for VSE/VRE use 12 mg/kg.
MICROBIOLOGY AND SUSCEPTIBILITY TESTING
Limitations of Microbiology Susceptibility Testing
In vitrodata do not differentiate between colonizers and pathogens.
In vitrodata do not necessarily translate intoin vivoefficacy.
Antibiotic activity-effectiveness depends on body site concentrations local pH, degree of inflammation, cellular debris, local oxygen levels, blood supply and penetrability.
In vitrosusceptibility testing is dependent on the microbe, methodology, pH, and antibiotic concentration.
In vitro susceptibility testing assumes the isolate was recovered from blood, using serum concentrations of an antibiotic given in the usual dose.
Since some body site, e.g., bladder urine contains higher antibiotic concentrations than found in serum, and other body sites, e.g., CSF levels may be lower than in serum, i.e., in vitrosusceptibility may be misleading for non-bloodstream infections.
Antibiotics should be prescribed at the usual recommended doses; attempts to lower cost by reducing dosage may decrease antibiotic efficacy, e.g., cefoxitin 2 gm IV inhibits ~ 85% of B. fragilis, whereas 1 gm IV inhibits only ~ 20%.6
Antibiotic-Organism Combinations for Which In Vitro Susceptibility Testing Does Not Predict In Vivo Effectiveness
H. influenzae, Yersinia pestis, VSE*
Klebsiella, VSE, Bartonella
Streptococci, Salmonella, Shigella
Fusobacteria, Clostridia, Listeria
1st, 2nd generation cephalosporins
Salmonella, Shigella, Bartonella
3rd, 4th generation cephalosporins
Listeria, Bartonella, MRSA†
† In spite of apparent in vitro susceptibility of antibiotics against MRSA, only vancomycin, minocycline, quinupristin/dalfopristin, linezolid, tedizolid, daptomycin, ceftaroline fosamil, telavancin, dalbavancin, oritavancin, and tigecycline are effective in vivo.
* Effective penicillin therapy for systemic enterococcal infections due to VSE requires an aminoglycoside, e.g., gentamicin.
PK/PD AND OTHER CONSIDERATIONS IN ANTIMICROBIAL THERAPY
Antibiotic Dosing: Concentration vs. Time Dependent Kinetics
PCN: maintain concentrations > MIC for ≥ 60% of the dosing interval
β-lactams: maintain concentrations > MIC for ≥ 75% of the dosing interval
Carbapenems: maintain concentrations > MIC for ≥ 40% of the dosing interval
Vancomycin (if MIC ≤ 1 mcg/mL) use 1 gm (IV) q12h
Use high doses (which increase serum concentrations which also increase T > MIC for more of the dosing interval)
Other Antibiotics (Cmax: MIC/T>MIC and or AUC 0-24/MIC)
> 125 (effective)
> 250 (more effective)
Use highest effective dose (without toxicity)
Bactericidal vs. Bacteriostatic Therapy
For most infections, bacteriostatic and bactericidal antibiotics inhibit/kill organisms at the same rate, and should not be a factor in antibiotic selection.
Bactericidal antibiotics have an advantage in certain infections, such endocarditis, meningitis, and febrile leukopenia, but there are exceptions even in these cases.
Monotherapy vs. Combination Therapy
Monotherapy is preferred to combination therapy for nearly all infections.
In addition to cost savings, monotherapy results in less chance of medication error and fewer missed doses/drug interactions.
Combination therapy may be useful for drug synergy or for extending spectrum beyond what can be obtained with a single drug.
Combination therapy is not effective in preventing antibiotic resistance, except in very few situations.
IV to PO Switch Therapy
Patients admitted to the hospital are usually started on IV antibiotic therapy, then switched to equivalent oral therapy after clinical improvement/defervescence (usually within 72 hours).
Advantages of early IV-to-PO switch programs include reduced cost, early hospital discharge, less need for home IV therapy, and virtual elimination of IV line infections.
Drugs well-suited for IV-to-PO switch or for treatment entirely by the oral route have high bioavailibility, e.g., doxycycline, minocycline, clindamycin, metronidazole, chloramphenicol, amoxicillin, trimethoprim-sulfamethoxazole, quinolones, and linezolid.8
Bioavailability of Oral Antimicrobials
Excellent (> 90%)
Nitazoxanide (with food)
Poor (< 60%)
Nitazoxanide (without food)
Oral Antibiotic Therapy for Serious Systemic Infections
Most infectious diseases should be treated orally unless the patient is critically ill, cannot take antibiotics by mouth, or there is no equivalent oral antibiotic.
If the patient is able to take/absorb oral antibiotics, there is no difference in clinical outcome using equivalent IV or PO antibiotics.
It is more important to think in terms of antibiotic spectrum, bioavailability and tissue penetration, rather than route of administration.
Nearly all non-critically ill patients may be treated in part or entirely with oral antibiotics.
When switching from IV to PO therapy, oral antibiotic chosen should have the same spectrum/degree of activity against the presumed/known pathogen and achieve the same blood and tissue levels as the equivalent IV antibiotic.
OPAT (outpatient parenteral antibiotic therapy)
OPAT has been used to treat infections IV on an outpatient basis or to complete IV therapy begun during hospitalization. Preferred OPAT antibiotics are those with few adverse effects and those with a long serum half life. The most frequently used OPAT antibiotics are ceftriaxone and vancomycin.9
Other agents with long t1/2 ideal for OPAT of Gram positive cSSSIs due to MRSA are telavancin 10 mg (IV) q 24 h, dalbavancin 1 gm (IV) × 1 doses then 500 mg (IV) × 1 dose 7 days later; tedizolid 200 mg (IV)q 24 h × 6 days, then 200 mg (PO) q 24 h × 6 days; and oritavancin 1200 mg (IV) × 1 dose.
The preferred alternative to OPAT is oral antibiotic therapy, e.g., for MRSA, minocycline or linezolid are equally efficacious as OPAT regimens.
Duration of Therapy
Most bacterial infections in normal hosts are treated with antibiotics for 1–2 weeks.
The duration of therapy may need to be extended in patients with impaired immunity, e.g., diabetes, SLE, alcoholic liver disease, neutropenia, diminished splenic function, etc., chronic bacterial infections e.g., endocarditis, osteomyelitis, chronic viral and fungal infections, or certain bacterial intracellular pathogens.
COLONIZATION VS. INFECTION
Colonization vs. Infection†: Acute Uncomplicated Cystitis (AUC) and Catheter Associated Bacteriuria (CAB)*
Pyuria (> 30 WBCs/hpf) with no or low grade bacteriuria (< 50 cfu/mL).
AUC: bacteriuria with minimal pyuria (< 10 WBCs/hpf) and low grade bacteriuria (< 50 cfu/mL).
CAB: bacteriuria with pyuria (> 10 WBCs/hpf)§ and bacteriuria (> 50 cfu/mL)
AUC with moderate – high grade pyuria (>10 WBCs/hpf) and high grade bacteriuria (>100 cfu/mL)*.
CAB (UA/UC after Foley removed/replaced) with >10 WBCs/hpf and high grade bacteriuria (> 100 cfu/mL)*
* If indwelling urinary catheter, change/replace catheter and repeat AU/UC before considering treatment† UTI = high grade pyuria with high grade bacteriuria. Pyuria ≠ UTI; Dysuria ≠ UTI; Bacteriuria ≠ UTI
§ clumps = high grade pyuria.10
COLONIZATION VS. INFECTION
Colonization vs. Infection: Respiratory Secretions
Colonization of respiratory secretions in ventilated patients is the rule and a function of days intubated.
GNB from the ICU milieu colonize respiratory secretions.
Common GNB colonizers include Klebsiella, Enterobacter, Serratia, S. maltophilia, B. cepacia and P. aeruginosa.
Most GNB colonizing organisms rarely, if ever, cause NP/VAP, e.g., Enterobacter, S. maltophilia, B. cepacia.
In ventilated patients with fever, leukocytosis, and infiltrates on CXR organisms cultured from respiratory secretions should be considered as airway colonizers and not the cause of NP/VAP unless accompanied by the characteristic clinical features of the pathogen.
Klebsiella pneumonia presents with fevers and rapid cavitation (3-5 days after infiltrates) P. aeruginosa pneumonia presents with high spiking fevers, and rapid cavitation (<3 days after infiltrates).
Ventilated patients treated with antibiotics with minimal S. aureus activity, e.g., ciprofloxacin, ceftazidime are often colonized with MSSA or MRSA, but do not develop S. aureus NP/VAP.
S. aureus (MSSA or MRSA) may cause CAP with influenza pneumonia but rarely, if ever, causes NP/VAP.
Data suggesting that S. aureus in ventilated patients with fever, leukocytosis and infiltrates on CXR is a common cause of NP/VAP is based on respiratory secretion cultures and represents colonization and not the cause of NP.
Such data represents colonization of respiratory secretions in ventilated patients. Without the distinctive clinical features of MSSA or MRSA pneumonia, e.g., high fevers, clinical deterioration and rapid cavitation (<3 days) on CXR the patients should be considered as colonized or may have tracheobronchitis.
COLONIZATION VS. INFECTION
Colonization vs. Infection: Draining Wounds
Wound Gram stain
Few or some WBCs
Few or some WBCs
Few or some WBCs
Colonization vs. Infection: Sacral Decubitus (Stage III/IV) and Diabetic Foot Ulcers (± osteomyelitis)
Always treat the usual pathogens (related to body site flora) rather than just “covering the cultured organism” (particularly if the specimen is not representative of the infected tissue).
Diabetic foot ulcers with infections at the body site, chronic osteomyelitis
Cover the usual pathogens: GAS, GBS, common coliforms, S. aureus, and B. fragilis (not P. aeruginosa)
Do not cover surface ulcer colonizers cultured: P. aeruginosa, Acinetobactor, VSE/VRE, Enterobacter, Burkholderia, Stenotrophomonas
Do not rely on deep ulcer/fistula cultures (which represent skin flora) and are not reflective of bone pathogens, i.e., osteomyelitis.
If P. aeruginosa is cultured from deep ulcer/fistula, do not cover only for P. aeruginosa. Over 95% of diabetic foot ulcers/fistulas will be culture positive from P. aeruginosa (due to P. aeruginosa colonization from wet socks, wet dressings, whirlpool baths). In aseptically collected bone specimens in the OR, P. aeruginosa is NOT a bone pathogen in diabetics with chronic osteomyelitis.
Low level/low grade BC positivity (1/4 − 2/4 + BCs)
TTPC = > 2 days
No clinical source of CoNS + BCs (CVC, implanted orthopedic/cardiac devices, prosthetic materials, severe/prolonged neutropenia)
CoNS Bacteremia (infection likely) with:
Persistently positive BCs
High level/high grade bacteremia (3/4 – 4/4 BCs +)
TTPC = < 2 days
Clinical source of + BCs for CoNS apparent (CVC, implanted orthopedic/cardiac devices, pros- thetic materials, severe/prolonged neutro- penia)
Implanted/prosthetic device associated: Dx = gallium or indium scan. ABE: Dx = cardiac vegetation. Abscess Dx = Gallium scan or CT scan. CVC associated: Dx = SQ removed CVC tip culture with > 15 col of same organism as in BCs not drawn from the CVC. TTPC = time to positive culture
Clinical Features of Drug Fever
Many but not all individuals are atopic
Patients have been on a sensitizing medication for days or years “without a problem”
Fevers may be low- or high-grade, but usually range between 102°–104°F and may exceed 106°F
Patient appears “inappropriately well” for degree of fever
Elevated WBC count (usually with left shift)
Eosinophils almost always present, (but eosinophilia is uncommon)
Elevated erythrocyte sedimentation rate in most
Early, transient, mild elevations of serum transaminases (common)
Provide empiric coverage primarily directed against the most probable pathogens causing the infection at the body site.
Avoid “covering” or “chasing” multiple organisms cultured that are (pathogens and non-pathogens) at the body site cultured.
Selectively treat CAB in immunocompromised hosts. Avoid treating CAB in normal hosts.
Colonization of respiratory secretions, wounds, or urine with “water” (S. maltophilia, B. cepacia, P. aeruginosa) or skin organisms (MSSA, MRSA, CoNS, VSE, VRE) is the rule.
Narrow vs. Broad Spectrum Therapy
Narrow vs broad spectrum doesn't prevent resistance, e.g., in treating E. coli urosepsis switching from a carbapenem (broad spectrum) to ampicillin (narrow spectrum) may actually increase resistance potential.14
Narrow spectrum vs broad spectrum is not clinically superior to well chosen broad spectrum therapy, e.g., switching from ceftriaxone (broad spectrum) to penicillin (narrow spectrum in treating S. pneumoniae has no clinical rationale or clinical advantage and has no effect on controlling resistance.
Antibiotic resistance potential is related to individual antibiotics and not antibiotics class, e.g., meropenem (low resistance potential) vs. imipenem (high resistance potential).
Antibiotic resistance is not related to spectrum narrowness or broadness, e.g., levofloxacin (broad spectrum but “low resistance potential”) vs ampicillin (narrow spectrum but “high resistance potential”).
The best way to control resistance is a selectively restricted formulary; restricted only to “high resistance potential” antibiotics, e.g., ciprofloxacin (not levofloxacin or moxifloxacin), imipenem (not meropenem or ertapenem), cefazidime (not other 3rd of 4th GC), gentamicin/tobramycin (not amikacin).
Some antibiotics may be restricted for other reasons, e.g., excessive vancomycin (IV not PO) use predisposes to VRE emergence and vancomycin may cause cell wall thickening in S. aureus resulting in permeability related resistance (to vancomycin and other antibiotics, e.g., daptomycin).
Over-restriction of antibiotics may impair timely effective therapy and does not, per se, decrease resistance.
Preferentially select antibiotics (all other things being equal) with a “low resistance potential”. Avoid, if possible, “high resistance potential” antibiotics, e.g., macrolides (for respiratory infections), TMP-SMX (for UTIs).
Except for TB therapy, combination therapy doesn't prevent resistance. The only examples of combination therapies that prevent resistance are carbenicillin + gentamicin and FC + amphotericin B.
Since resistance is, in part, concentrations dependant, subtherapeutic or low antibiotic tissue concentrations, (all other things being equal) predisposes to resistance.
Suboptimal dosing or usual dosing with inadequate tissue penetration, e.g., into the body fluids or undrained abscesses (source control is key) also predisposes to resistance.
Treat for the shortest duration of therapy that is effective in eliminating the infection.
Monotherapy vs. Combination Therapy
Preferably use monotherapy whenever possible to cover the most likely pathogen or cultured pathogen clinically relevant to the site of infection.
Combination therapy should be avoided if possible. Always tty to preferentially use monotherapy.
Monotherapy is usually less expensive than combination therapy and has less potential for adverse effects and drug-drug interactions.15
Combination therapy is often used for potential synergy (rarely occurs and if used must be based on microbiology laboratory synergy studies), to increase spectrum (preferable to use monotherapy with same spectrum), or to prevent resistance (except for TB, ineffective in nearly all contributors).
PO and IV-to-PO Switch Antibiotic Therapy (see also Table 1.5)
Wherever possible, treat with entirely oral antibiotic therapy instead of IV therapy.
Switch from IV-to-PO antibiotic therapy after clinical defervescence (usually < 72 hours).
Early IV-to-PO switch therapy eliminates phlebitis and IV line associated infections.
De-escalation is problematic if based on microbiology data alone without site-pathogen correlation.
De-escalation is appropriate in the setting of broad spectrum coverage of “presumed urosepsis” which can be narrowed after the uropathogen is identified in blood/urine.
In intubated/ventilate patients, microbiology data from respiratory secretion cultures are usually misleading and not representative of NP or VAP lung pathogens.
In patients with NP or VAP, it is more prudent to treat the most likely pathogen, e.g., P. aeruginosa (even if not cultured from respiratory secretions) than to be misguided into treating multiple colonizing organisms in respiratory secretions.
De-escalation can be harmful if microbiology data is misleading, e.g., represents colonization rather than being reflective of the pathogen (underlying bone pathogen, not ulcer organisms), e.g., diabetic foot ulcers/chronic osteomyelitis or sacral ulcers/chronic osteomyelitis.
C. difficile Diarrhea/Colitis
Preferentially select antibiotics (all other things being equal) with low C. difficile potential.
Predisposing factors to C. difficile include relatively few antibiotics, e.g., clindamycin, b-lactams, ciprofloxacin.
Most antibiotics have little/no C. difficile potential, e.g., aminoglycosides, aztreonam, maerolides, TMP-SMX, colistin, polymyxin B, daptomycin, Q/D, doxycycline, minocycline, tigecycline, vancomycin, linezolid.
Some antibiotics are protective against C. difficile, e.g., doxycycline, tigecycline.
Always consider non-antibiotic factors that may predispose to C. difficile, e.g., cancer chemotherapy, anti-depressants, statins, PPIs.
Also consider person to person spread or acquisition for the environment.
Empiric Antibiotics for Fever and Leukocytosis
Avoid treating unexplained fever/leukocytosis. If due to infection, let the infection declare itself and then initiate early therapy.
Try to diagnose the many non-infectious causes of fever/leukocytosis rather than treating empirically with antibiotics.16
Avoid prolonged antibiotic therapy of fever/leukocytosis in the presence of a device associated infection or undrained/inadequately drained abscesses.
Avoid antibiotic therapy of non-infectious fevers or non-antibiotic responsive infections, e.g., drug fevers, malignancies, hematomas, rheumatic/inflammatory disorders, viral infections.
The least expensive therapy is usually not the best therapy.
The least expensive antibiotic (acquisition cost) may, in fact, be expensive (re: total cost) when considering the cost implications to the institution of dosing frequency, C. difficile potential, resistance potential, and degree of activity against the known or likely pathogen, and the cost of potential therapeutic failure vis-à-vis ↑ LOS and medicolegally.
Stewardship savings are best achieved by decreasing duration of antibiotic therapy, and by treating entirely with oral antibiotic therapy or early IV-to-PO switch therapy.
Bennett JE, Dolin R, Blaser MJ (eds). Mandell, Douglas, and Bennett's Principles and Practice of Infectious Diseases (8th Ed). Philadelphia Elsevier Churchill Livingstone,
Bryskier A (ed). Antimicrobial Agents. ASM Press,
Washington, D.C., 2005.