Antimicrobial stewardship (AMS) programs optimize antimicrobial use to improve patient outcomes, reduce antimicrobial resistance, and minimize adverse effects including Clostridioides difficile infection. This skill covers the principles, strategies, and practical implementation of antimicrobial stewardship.
Core Principles: Start Smart, Then Focus
The UK "Start Smart Then Focus" framework (PHE/ARHAI) provides a practical approach:
Start Smart (Prescribing Initiation)
Do not start antibiotics unless there is clear evidence of infection — Distinguish infection from colonization, contamination, or inflammatory conditions without infection
Take cultures before starting antibiotics — Blood cultures, urine, sputum, wound swabs as clinically indicated. Do not delay treatment in sepsis.
Prescribe according to local guidelines — Local empiric guidelines reflect local resistance patterns and are more reliable than national guidelines for initial therapy
Skills relacionados
Document the indication, drug, dose, route, and planned duration in the medical record and on the drug chart
Include a review date — Typically 48-72 hours
Then Focus (Antimicrobial Review at 48-72 Hours)
At clinical review, one of five actions:
Stop — If no evidence of infection confirmed
Switch IV to oral — If clinically improving and oral criteria met
Change — To narrow-spectrum targeted therapy based on culture results
Continue — If indication remains appropriate and no culture data to guide change
Outpatient parenteral antibiotic therapy (OPAT) — If IV antibiotics needed but patient otherwise suitable for discharge
Empiric vs Targeted Therapy
Empiric Therapy
Initiated before the causative organism is identified. Guided by:
Likely pathogens based on site of infection, patient risk factors, and clinical syndrome
Local antibiogram (institutional resistance data)
Severity of illness — Broader coverage justified in severe sepsis/septic shock
Narrow the spectrum to the most targeted effective agent
Verify clinical response supports the culture result (contamination vs true pathogen)
Adjust dose for MIC data when available (especially for gram-negatives)
De-escalation
Narrowing antibiotic spectrum from broad empiric to targeted therapy:
When: Culture results available and patient is clinically improving (typically 48-72 hours)
How: Change from broad-spectrum (e.g., piperacillin-tazobactam) to narrow-spectrum (e.g., amoxicillin if susceptible Streptococcus)
Barriers: Clinician reluctance ("the patient is improving on current therapy — why change?"), lack of culture data, diagnostic uncertainty
Evidence: De-escalation is associated with equivalent or improved outcomes compared to maintaining broad-spectrum therapy
IV-to-Oral Switch Criteria
Standard Criteria (OPAT National Outcomes Registry, BSAC)
Switch from IV to oral when ALL of the following are met:
Clinical improvement — Temperature <38°C for ≥24 hours, improving inflammatory markers, resolution of sepsis signs
Functioning GI tract — Tolerating oral intake, no vomiting, no malabsorption, no ileus
Appropriate oral option available — Adequate bioavailability, effective against identified/likely pathogen
No indication for prolonged IV therapy — Some infections require IV: endocarditis, meningitis (initial phase), osteomyelitis (debated — OVIVA trial supports oral in many cases), deep-seated abscess without source control
High-Bioavailability Oral Antibiotics
Some oral antibiotics approach IV bioavailability and are excellent switch options:
A cumulative antibiotic susceptibility report based on all clinical isolates from a defined period (usually annual), stratified by organism and source.
How to Read It
Rows: Organisms (e.g., E. coli, S. aureus, P. aeruginosa)
Columns: Antibiotics tested
Values: Percentage of isolates susceptible
>90% susceptibility: Generally suitable for empiric use
80-90% susceptibility: Consider for empiric use with risk factor assessment
<80% susceptibility: Unreliable for empiric use — reserve for targeted therapy based on susceptibility results
Limitations
Reflects the institution's patient population (may include high-risk patients inflating resistance rates)
Does not distinguish community-acquired from hospital-acquired isolates (some antibiograms do)
Annual data may not reflect recent shifts in resistance
Does not account for PK/PD optimization (a drug might be effective with optimized dosing despite borderline MICs)
PK/PD Optimization
Antimicrobial Killing Patterns
Different antibiotics have different relationships between drug exposure and bacterial killing:
Altered clearance (augmented renal clearance in early sepsis, impaired clearance later)
Reduced protein binding (hypoalbuminemia)
Strategies: Loading doses, extended/continuous infusions of beta-lactams, therapeutic drug monitoring, consideration of augmented renal clearance (CrCl >130 mL/min → may need higher doses).
Duration of Therapy
Shorter Courses Are Often Sufficient
Growing evidence supports shorter antibiotic courses for many common infections:
Infection
Traditional Duration
Evidence-Based Duration
Community-acquired pneumonia
7-10 days
3-5 days (if clinically stable by day 3)
Hospital-acquired pneumonia
14-21 days
7 days (PneumA trial)
Urinary tract infection (uncomplicated)
7 days
3 days (women), single dose (some agents)
Intra-abdominal infection (source controlled)
7-14 days
4 days (STOP-IT trial)
Skin and soft tissue (uncomplicated)
10 days
5-6 days (if improved)
Acute bacterial sinusitis
10-14 days
5-7 days
Biomarker-Guided Duration
Procalcitonin (PCT) — Serial monitoring can guide antibiotic discontinuation in lower respiratory tract infections and sepsis. Multiple RCTs show reduced duration without adverse outcomes.
CRP — Slower to normalize. Rising CRP despite treatment suggests treatment failure or complication. Can support discontinuation when clearly trending down.
Treatment: Carbapenems are the standard for serious infections. Piperacillin-tazobactam is debated (MERINO trial showed inferiority to meropenem for ESBL E. coli bacteremia).
Carbapenem-sparing options for UTI: Nitrofurantoin, fosfomycin, trimethoprim (if susceptible). Avoid cephalosporins even if reported susceptible in vitro.
VRE (Vancomycin-Resistant Enterococcus)
Treatment: Linezolid, daptomycin (higher doses for E. faecium)
Key point: Distinguish colonization from infection. Many VRE-positive cultures represent colonization and do not require treatment.
Carbapenem-Resistant Organisms (CRE/CPE)
Treatment: Emerging options include ceftazidime-avibactam, meropenem-vaborbactam, imipenem-relebactam, cefiderocol. Choice depends on resistance mechanism (KPC, OXA-48, NDM, VIM).
Stewardship: Reserve novel agents for confirmed resistant infections to preserve their utility.
Applying This Skill
When advising on antimicrobial use:
Apply the Start Smart Then Focus framework
Recommend empiric therapy based on likely pathogens and local resistance patterns
Prompt de-escalation at 48-72 hours based on culture results
Identify opportunities for IV-to-oral switch
Recommend evidence-based durations (shorter courses where supported)
Optimize PK/PD for the specific drug class (extended infusion for beta-lactams, high-dose for aminoglycosides, AUC-guided for vancomycin)
Flag high-risk antibiotics for C. difficile and recommend alternatives when available