Do you order blood cultures for all your ED patients with a fever? Obviously no. What’s your decision making process on ordering this test? There are really no findings or tests with high specificity (rules-IN bacteremia), except interestingly “shaking chills”. Notice almost all the criteria listed below approach a likelihood ratio (LR) of 1.0. Two prediction rules do exist, however, to help you virtually rule-OUT bacteremia:
- Shapiro prediction rule
The list of LRs also will be helpful to show learners in the ED that an isolated serum WBC number is useless risk-stratifier.
A 55 y/o man with a PMH of hypertension presents with a community-acquired pneumonia on CXR, no fevers, no chills, no vomiting.
- Temperature 37.8 C, BP 160/90, HR 100, RR 16, Sat 100% RA
- Serum WBC 20K (no bands)
- Platelets 300K
- Creatinine 1.1 mg/dL
What is the patient’s pre-test and post-test probability for having bacteremia? Use these helpful stats from the Rational Clinical Examination series from JAMA.
PV Card: Blood Cultures for Suspected Bacteremia
Adapted from  Go to ALiEM (PV) Cards for more resources.
Answer to patient case
- Start with 7% pretest probability for bacteremia with a community acquired pneumonia.
- Using the clinical prediction rules, the WBC 20K and HR 100 bpm are criteria for SIRS but do not fulfill the Shapiro prediction criteria. LR = 1.8 * 0.08 = 0.144. Post-test probability for bacteremia = 0.06%.
- If the patient had instead a normal HR of 80 bpm, both the SIRS and Shapiro criteria would have been negative. LR = 0.09 * 0.08 = 0.0072. Post-test probability for bacteremia = << 0.1%.
This discussion doesn’t address WHETHER we should get blood cultures despite a risk for bacteremia in the setting of uncomplicated pneumonia receiving IV antibiotics or pyelonephritis with a pending urine culture.
- Coburn B, Morris A, Tomlinson G, Detsky A. Does this adult patient with suspected bacteremia require blood cultures? JAMA. 2012;308(5):502-511. [PubMed]
- Shapiro N, Wolfe R, Wright S, Moore R, Bates D. Who needs a blood culture? A prospectively derived and validated prediction rule. J Emerg Med. 2008;35(3):255-264. [PubMed]