Author: John Genova, MD

 

Background:

Blood cultures remain a cornerstone test for detecting bacteremia, fungemia, and endovascular infection. In the ED, they influence decisions on admission, antibiotics, and follow-up. Misuse leads to unnecessary antibiotics, false positives, phone-call callbacks, and missed high-risk infections.

When to Obtain Blood Cultures

High likelihood of bacteremia — Obtain blood cultures before antibiotics in adults with:

    • Sepsis or septic shock
    • Suspected infective endocarditis
    • Infected prosthetic material (ICD/pacemaker, vascular grafts, prosthetic valves)
    • IV catheter infection / central line associated bloodstream infection (CLABSI)
    • Epidural abscess, vertebral osteomyelitis/discitis
    • Septic thrombophlebitis
    • Meningitis, septic arthritis, CNS shunt infection
    • ≥2 SIRS criteria + suspected bacterial infection

Moderate likelihood — consider blood cultures

    • Pyelonephritis
    • Cholangitis
    • Moderate–severe CAP
    • Cellulitis in the immunocompromised
    • When source cultures are unavailable prior to antibiotics
    • Patients at risk for endovascular infection

Low likelihood — generally do not obtain blood cultures

    • Isolated fever or leukocytosis with stable vitals
    • Non-severe community acquired pneumonia
    • Simple cellulitis
    • Cystitis / prostatitis
    • Post-op fever <48 hours

Avoiding unnecessary cultures reduces false positives, unnecessary antibiotics, and follow-up burdens.

Pediatric guidance varies by age, vaccination status, and immunocompetence.

Collecting Blood Cultures 

    • 2 sets from 2 separate venipuncture sites
      (3 sets from 3 sites if suspecting endocarditis)
    • Never obtain a single set
    • Avoid drawing from existing lines, if possible
    • Fill aerobic bottle first
    • Maximize volume (biggest driver of yield)
    • Maintain strict aseptic technique

Poor collection technique leads to contamination and unnecessary antibiotics / return to care visits.

Interpreting Gram Stains

Gram-Positive Cocci in Clusters → Staphylococcus spp.

  • Always significant: S. aureus
  • Usually contaminant: Coagulase-negative staph
  • If prosthetic material/endocarditis risk: treat as real until proven otherwise.

Gram-Positive Cocci in Chains or Pairs → Streptococci or Enterococci

  • Usually clinically significant
  • Viridans streptococci + valve disease = endocarditis until proven otherwise.

Gram-Positive Bacilli

  • Significant: Listeria, Clostridium
  • Usually contaminants: Corynebacterium, Cutibacterium, Bacillus spp. (except anthrax)

Gram-Negative Rods

  • Nearly always significant
  • Includes Enterobacterales, Pseudomonas, Bacteroides, etc.

Coccobacilli

  • Significant pathogens: H. influenzae, Brucella, Coxiella, Francisella, Yersinia, Kingella, Acinetobacter

Organisms that are always clinically significant

    • Staphylococcus aureus
    • Streptococcus pneumoniae
    • Group A streptococci
    • Enterobacteriaceae (Escherichia, Klebsiella, Proteus, Salmonella, Shigella, etc.)
    • Haemophilus influenzae
    • Pseudomonas spp.
    • Bacteroidaceae
    • Campylobacter spp.
    • Neisseria meningitidis
    • Listeria monocytogenes
    • Enterococcus spp.
    • Brucella spp.
    • Pasteurella spp.
    • HACEK organisms
    • Fungal growth: Aspergillus, Candida spp., Fusarium

Organisms requiring clinical consideration:

  • Enterococci
    • Frequent cause (after staphylococci) of:
      • Nosocomial infections
      • UTIs / CAUTIs
      • CLABSI bacteremias
      • Hospital-associated endocarditis
  • Viridans streptococci
      • Significance depends on clinical context

Likely contaminants (skin flora)

    • Coagulase-negative staphylococci
    • Corynebacterium
    • Propionibacterium / Cutibacterium acnes
    • Micrococcus spp.
    • Bacillus spp. (except B. anthracis)
    • Lactobacillus spp.
    • Others consistent with normal skin flora
    • Exception: Coagulase-negative Staphylococcus saprophyticus
      • Frequent cause (after E. coli) of community-acquired UTIs

Common pediatric blood culture pathogens

  • Neonates
    • Group B streptococci
    • E. coli
    • Klebsiella
    • Enterococci
    • Coagulase-negative staphylococci
    • Staphylococcus aureus
  • Infants & Children
    • Streptococcus pneumoniae
    • Staphylococcus aureus
    • Streptococcus pyogenes
    • Neisseria meningitidis
    • Escherichia coli

Fungal Blood Cultures

  • Blood cultures are poor at detecting most invasive fungi.
    • Candida → may grow from routine blood cultures, requires prompt therapy
  • Consider β-D-glucan (BDG) for suspected invasive fungal infection
    • Useful for Candida, Aspergillus, Fusarium, Pneumocystis
    • Does not detect Cryptococcus or Mucorales (e.g., Mucor, Rhizopus)
    • Sensitivity ~80%, specificity ~82%
  • Fungal isolator cultures in select scenarios endemic molds (Blastomyces, Coccidioides, Histoplasma), Malessezi species, and other filamentous fungi.
  • ID consultation strongly recommended in all cases of suspected fungal infection

Effect of Prior Antibiotics

  • Prior oral or IV antibiotics significantly lowers blood culture positivity rates (26% → 13.5%).
  • This increases:
    • False negatives
    • Diagnostic delay
    • Need for repeat cultures
  • Always obtain blood cultures before antibiotics unless unsafe to delay.

Pitfalls & High-Risk Scenarios

1. Over-ordering in low-yield conditions

    • Up to 40% of positive blood cultures are contaminants, leading to unnecessary antibiotics, imaging, callbacks, and increased length of stay.
    • Avoid in isolated fever/leukocytosis, mild CAP/cellulitis, simple cystitis/prostatitis.
    • ED contamination rates (3–7%) often rise when blood cultures are used reflexively.

2. Contamination from poor collection technique

    • Most ED contaminants come from:
      • Drawing off lines
      • Suboptimal antisepsis
      • Inadequate venipuncture technique
    • Solution: strict chlorhexidine–alcohol prep + peripheral draws + two separate sites.

3. Under-recognizing high-risk patients

    • Elderly, immunosuppressed, and patients with prosthetic valves, vascular grafts, pacemakers, or recent cardiac surgery often present with normal vitals/labs despite true bacteremia.
    • Normal WBC, lactate, or post-antipyretic defervescence ≠ reassurance.

4. Misinterpreting “1 of 4” positives

    • One positive bottle of skin flora is often contamination.
    • One positive bottle of viridans streptococci, enterococci, or S. aureus in high-risk hosts is never dismissible—repeat cultures and assess for endovascular infection.

5. Missing infective endocarditis

    • Obtaining <3 sets prior to antibiotics
    • Dismissing viridans strep or enterococci in patients with valve disease/TAVR
    • Relying on a negative TTE when pretest probability is moderate–high
    • Any streptococcal bacteremia in valve disease or TAVR = endocarditis until proven otherwise.

6. Over-relying on culture results for disposition

    • Drawing blood cultures should not mandate admission or antibiotics in low-risk presentations.
    • But discharging high-risk patients with pending blood cultures requires robust follow-up and explicit return precautions.
      • Ensure patient contact information is correct and patient’s are counseled that they will receive a call in the setting of a positive test.

7. Breakdowns in follow-up

    • Many patients with true bacteremia are diagnosed after ED discharge.
    • Failure to contact patients or arrange timely return leads to delays in therapy and preventable morbidity. Emergency Departments should have a robust process for time sensitive patient notification in the setting of positive blood cultures.

References:

The CORE EM Pearl’s, Perils and Pitfall Series are adapted from safety bytes from the NYU Department of Emergency Medicine Division of Quality, Safety, and Practice Innovation

  1. Arnaout S et al. 2024; PMID: 39834750.
  2. Caldeira D et al. 2011; PMID: 21194791.
  3. Cheng MP et al. 2019; PMID: 31525774.
  4. Cruz AT et al. 2020; PMID: 33145549.
  5. Cummings LA et al. 2020; PMID: 32214207.
  6. Doern GV et al. 2019; PMID: 31666280.
  7. Doi M et al. 2024; PMID: 39739683.
  8. Duguid RC et al. 2025; PMID: 39848845.
  9. Ekwall-Larson A et al. 2022; PMID: 34851687.
  10. El Feghaly RE et al. 2018; PMID: 30217808.
  11. Fabre V et al. 2020; PMID: 31942949.
  12. Hirosawa T et al. 2023; PMID: 37408849.
  13. Huber S et al. 2020; PMID: 31654793.
  14. Garcia RA et al. 2015; PMID: 26298636.
  15. Kristóf K et al. 2016; PMID: 27683527.
  16. Lamy B et al. 2016; PMID: 27242721.
  17. Lee A et al. 2007; PMID: 17881544.
  18. MacVane SH et al. 2024; PMID: 39298359.
  19. Magadia RR et al. 2001; PMID: 11780265.
  20. Odabasi IO et al. 2020; PMID: 32617051.
  21. Onishi A et al. 2012; PMID: 22075593.
  22. Park WB et al. 2015; PMID: 26129718.
  23. Pien BC et al. 2010; PMID: 20800151.
  24. Prout AJ et al. 2020; PMID: 31789702.
  25. Snyder SR et al. 2012; PMID: 22709932.
  26. Song WS et al. 2020; PMID: 34886592.
  27. Varettas K et al. 2002; PMID: 12009100.
  28. Weinstein MP. AHRQ, Patient Safety Network, WebM&M: Case Studies, 2008.
  29. Weinstein MP et al. 1983; PMID: 6828811.
  30. White SK et al. 2020; PMID: 32693433.
  31. Wilson ML et al. 2020; PMID: 31377231.
  32. Wilson ML, 2025; UpToDate.
  33. Yu D et al. 2020.