Sepsis / Septic Shock — Source Not Yet Known
A severity-first empiric strategy that covers likely pathogens without reflexive maximal-spectrum therapy in every patient, then narrows aggressively as source and microbiology become clearer.
Immediate priorities
- If septic shock or a high likelihood of sepsis is present, administer effective antimicrobials immediately—ideally within 1 hour of recognition
- Obtain appropriate cultures before antibiotics only when this does not create a meaningful treatment delay
- Resuscitate shock and pursue source control in parallel rather than sequentially
- Review prior cultures, recent antibiotics, healthcare exposure and local antibiogram before choosing empiric coverage
Diagnostics before and during therapy
- At least two blood-culture sets from appropriate sites
- Source-directed specimens: urine, respiratory sample, CSF, drainage material, joint fluid or wound/tissue when clinically indicated
- Imaging guided by the suspected source
- Do not use procalcitonin alone to decide whether to start antibiotics
Empiric antimicrobial strategy
These are concrete empiric starting examples for the syndrome. Choose the option that fits acquisition setting, prior cultures and the local ICU antibiogram. Exact loading doses, renal/CRRT adjustments and infusion details stay in the medication monographs/local protocol.
Use when: Sepsis/septic shock with no source yet and no strong history suggesting ESBL/CRE; choose between them using the likely source and local antibiogram.
- Add vancomycin (or another local anti-MRSA agent) only when MRSA risk/source supports it.
- If cefepime is selected and an intra-abdominal/deep anaerobic source is suspected, add metronidazole or choose a single agent with anaerobic activity instead.
Use when: Septic shock or severe infection with prior ESBL Enterobacterales, recent broad beta-lactam exposure, or local epidemiology making a standard backbone unreliable.
- Add MRSA coverage only if indicated by source/risk.
- Use prolonged beta-lactam maintenance infusion after the loading dose when feasible in sepsis/septic shock.
Use when: The patient is unstable and has recent colonization/infection with CRE, DTR Pseudomonas, CRAB or another specific MDR organism.
- Use the prior susceptibility profile and current IDSA AMR guidance to choose the actual drug.
- Obtain ID/microbiology input early.
Choose one reliable broad gram-negative backbone matched to local susceptibility and source
Add MRSA coverage only when patient/source/risk profile justifies it
Add anaerobic coverage when the suspected source requires it; do not duplicate anaerobic therapy unnecessarily
Use a second gram-negative agent only in selected high-risk MDR/severe shock situations where the probability of inadequate initial therapy is substantial
Account for meningitis, endocarditis, neutropenia, severe immunosuppression and recent colonization because these change the empiric target set
Risk modifiers
Source control
- Drain infected collections
- Relieve urinary/biliary obstruction
- Remove infected intravascular devices when indicated
- Debride necrotic tissue
- Obtain urgent surgery/interventional radiology consultation when definitive source control is required
Reassessment
- Reassess diagnosis and antibiotic spectrum at 24–48 hours or sooner when microbiology returns
- If an alternative noninfectious diagnosis becomes likely, discontinue empiric antibiotics rather than completing an arbitrary course
- Use clinical trajectory, source control and microbiology—not fever alone—to judge response
De-escalation
- Narrow to the most targeted active agent once organism/susceptibilities are known
- Stop redundant combination therapy
- Convert IV to an active oral agent only when absorption, source, susceptibility and clinical stability make this appropriate
Duration principles
- Use the shortest evidence-supported duration after adequate source control and clinical response
- Duration starts to make sense only after defining the syndrome and controlling the source
- Persistent bacteremia, endovascular infection, uncontrolled source, osteomyelitis and profound immunosuppression require syndrome-specific duration rather than a generic sepsis duration