UMEM Educational Pearls

When we initiate the sepsis bundle in the ED for patients with suspected sepsis, what probability that those patients who received broad spectrum antibiotics in the ED would have bacterial infection.

This study (Shappell et al) provides us with a glimpse of those number.

 

Settings: Retrospective study of adults presenting to 4 EDs in Massachusetts.

Patients: patients with suspected serious bacterial infection in ED, defined as blood cultures and initiation of at least one broad spectrum antibiotics.  Random selection of 75 patients per hospital.

Patients were categorized in 4 groups:

• Definite bacterial infection: clinical syndrome, pathologic diagnosis of infection (positive cultures from blood, urine; pus; radiographic evidence of abscess, consolidations in lungs)
• Likely bacterial infection: not meeting criteria for definite infection, but having a compatible clinical syndrome responsive to antibiotics and no clear etiology or reason for clinical improvement.
• Unlikely bacterial infection: clinical syndrome consistent with infection, but an alternate diagnosis is more likely.
• Definitely no bacterial infection: there was clear non-infectious diagnosis and no evidence of concurrent bacterial process.

Outcome: Prevalence of each category.

Study Results: 300 patients who received broad spectrum antibiotics.

1. Prevalence of bacterial infection:
   1. 81 (27%) had definite bacterial infection
   2. 104 (34.7%) had likely bacterial infection
   3. 55 (18.3%) had unlikely bacterial infection
   4. 49 (16.3%) with definitely no bacterial infection
2. For 96 patients with suspicion of sepsis vs. the rest of the cohort (P = 0.36)
   1. Definite 42.7%
   2. Likely 29.2%
   3. Unlikely 16.7%
   4. Definitely no 11.5%

       3. For patients who were admitted to the ICU (P = 0.26)

  a.   Definite 16.5%

                b.   Likely 8.6%

  c.   Unlikely 16.4%

                d.   Definitely no 20.4%

4. Source of infection

5.  Definite/Likely bacterial infection
   1. GU = 69 (35%)
   2. Respiratory = 48 (24.4%)
   3. Skin or soft tissue = 45 (22.8%)
   4. Bacteremia or endovascular = 42 (21.3%)
   5. Abdominal = 24 (12.2%) 
6. Unlikely/definitely not bacterial infection

1. Viral = 27%
2. Volume overload/cardiac disease = 10%
3. Hypovolemia = 8%

 

Discussion:

1. Slightly more than half of the patient we covered with broad spectrum antibiotics would have definitely or likely bacterial infection.
2. This study agreed with previous studies (2), which suggested that for patients treated prophylactically for sepsis, 13% had a “none” likelihood, 30% of only "possible" likelihood for bacterial infection.
3. The study highlighted that it was not easy for Emergency clinicians to recognize bacterial infection when we operate on a limited source of information and a limited timeline (think about the bundle of sepsis).
4. However, overtreatment is also bad, so we just need to be cognizant.

Conclusion:

Approximately 30% of patients who had blood cultures drawn and received broad spectrum antibiotics in ED have low likelihood of bacterial infection.

Reference:

1. Shappell CN, Klompas M, Ochoa A, Rhee C; CDC Prevention Epicenters Program. Likelihood of Bacterial Infection in Patients Treated With Broad-Spectrum IV Antibiotics in the Emergency Department. Crit Care Med. 2021 Nov 1;49(11):e1144-e1150. doi: 10.1097/CCM.0000000000005090. PMID: 33967206; PMCID: PMC8516665.

2. Klein Klouwenberg PM, Cremer OL, van Vught LA, Ong DS, Frencken JF, Schultz MJ, Bonten MJ, van der Poll T. Likelihood of infection in patients with presumed sepsis at the time of intensive care unit admission: a cohort study. Crit Care. 2015 Sep 7;19(1):319. doi: 10.1186/s13054-015-1035-1. PMID: 26346055; PMCID: PMC4562354.