Fellow: Sarah Bram, MD/ Washington University School of Medicine
Article: Lipsett SC, Hall M, Ambroggio M, et al. Predictors of Bacteremia in Children Hospitalized With Community-Acquired Pneumonia. Hospital Pediatrics. Oct 2019; 9(10): 770-778.
Summary: IDSA guidelines recommend obtaining blood cultures only in children requiring hospitalization for presumed moderate, severe, or complicated bacterial pneumonia. This was a secondary analysis of a retrospective study using the PHIS+ database of children aged 3 months to 18 years hospitalized with community acquired pneumonia from 2007-2011. The aim of the study was to identify characteristics of children at higher risk for bacteremia and determine if blood cultures were warranted in a specific subset of children. Of the 2,568 children with blood cultures obtained (4,941 patients admitted with community acquired pneumonia total), only 2.5% had bacteremia and 0.4% had bacteremia with a penicillin nonsusceptible pathogen. Children with bacteremia had higher median admission WBC, CRP, and ESR, were more likely to have definite radiographic pneumonia on admission CXR, had higher prevalence of complicated pneumonia, and were more likely to be admitted to the ICU. WBC >20 and definite radiographic pneumonia were both independently associated with presence of bacteremia on a multivariable analysis. Of those patients not in the ICU, the prevalence of bacteremia was 2.2%, prevalence of bacteremia with a penicillin nonsusceptible pathogen was 0.25%, and prevalence of contaminated blood culture was 1.2%. Only 2 of the 4941 patients who did not have blood cultures initially had bacteremia detected later in their hospital course.
Key Strengths: Although the study was done at tertiary children’s hospitals, the authors excluded patients with complex chronic conditions making the results more applicable to our community sites as well. The multivariable analysis was an appropriate study methodology to identify individual predictors of bacteremia and added important data for the study.
Limitations: The main limitation was the time period for the study, which was performed between 2007-2011, before the introduction of the pneumococcal conjugate vaccine (PCV13). This may have impacted the rates of bacteremia in this study population as well as the epidemiology of the organisms causing such bacteremia. In addition, it is difficult to make clinical decisions using WBC count (and other lab markers including CRP and ESR that outperformed WBC on the ROC curve) given that labs are usually obtained at the same time as a blood culture in the pediatric population.
Major Takeaway: This article supports the recommendation against obtaining blood cultures and even a CBC in children hospitalized with community acquired pneumonia in the non-ICU setting. The number needed to treat to identify 1 child with penicillin non-susceptible bacteremia was 1 in 400. As ampicillin is the empiric antibiotic recommended for community acquired pneumonia, with such a low reported prevalence of penicillin resistant bacteremia (0.4% total, 0.25% in non-ICU setting), clinicians should feel comfortable holding off on obtaining blood cultures and routine labs in this population. There is still data to support obtaining blood cultures in those children that are admitted to the ICU, especially with complicated pneumonia.
Impact on Practice: This article argues the point that we should have clear indications for when a blood culture is warranted for any disease process, including community acquired pneumonia. As my quality improvement work is currently focused on contaminated blood cultures (see below), I think it is important to consider the impact contaminated blood cultures can have on patients, families, and the health care system. Even with a low prevalence rate of 1.2% in non-ICU patients in this study, contaminated blood cultures were 5 times more prevalent than penicillin non-susceptible bacteremia. This article is another example of the importance of value-based care.
Article: Lipsett SC, Hall M, Ambroggio M, et al. Predictors of Bacteremia in Children Hospitalized With Community-Acquired Pneumonia. Hospital Pediatrics. Oct 2019; 9(10): 770-778.
Summary: IDSA guidelines recommend obtaining blood cultures only in children requiring hospitalization for presumed moderate, severe, or complicated bacterial pneumonia. This was a secondary analysis of a retrospective study using the PHIS+ database of children aged 3 months to 18 years hospitalized with community acquired pneumonia from 2007-2011. The aim of the study was to identify characteristics of children at higher risk for bacteremia and determine if blood cultures were warranted in a specific subset of children. Of the 2,568 children with blood cultures obtained (4,941 patients admitted with community acquired pneumonia total), only 2.5% had bacteremia and 0.4% had bacteremia with a penicillin nonsusceptible pathogen. Children with bacteremia had higher median admission WBC, CRP, and ESR, were more likely to have definite radiographic pneumonia on admission CXR, had higher prevalence of complicated pneumonia, and were more likely to be admitted to the ICU. WBC >20 and definite radiographic pneumonia were both independently associated with presence of bacteremia on a multivariable analysis. Of those patients not in the ICU, the prevalence of bacteremia was 2.2%, prevalence of bacteremia with a penicillin nonsusceptible pathogen was 0.25%, and prevalence of contaminated blood culture was 1.2%. Only 2 of the 4941 patients who did not have blood cultures initially had bacteremia detected later in their hospital course.
Key Strengths: Although the study was done at tertiary children’s hospitals, the authors excluded patients with complex chronic conditions making the results more applicable to our community sites as well. The multivariable analysis was an appropriate study methodology to identify individual predictors of bacteremia and added important data for the study.
Limitations: The main limitation was the time period for the study, which was performed between 2007-2011, before the introduction of the pneumococcal conjugate vaccine (PCV13). This may have impacted the rates of bacteremia in this study population as well as the epidemiology of the organisms causing such bacteremia. In addition, it is difficult to make clinical decisions using WBC count (and other lab markers including CRP and ESR that outperformed WBC on the ROC curve) given that labs are usually obtained at the same time as a blood culture in the pediatric population.
Major Takeaway: This article supports the recommendation against obtaining blood cultures and even a CBC in children hospitalized with community acquired pneumonia in the non-ICU setting. The number needed to treat to identify 1 child with penicillin non-susceptible bacteremia was 1 in 400. As ampicillin is the empiric antibiotic recommended for community acquired pneumonia, with such a low reported prevalence of penicillin resistant bacteremia (0.4% total, 0.25% in non-ICU setting), clinicians should feel comfortable holding off on obtaining blood cultures and routine labs in this population. There is still data to support obtaining blood cultures in those children that are admitted to the ICU, especially with complicated pneumonia.
Impact on Practice: This article argues the point that we should have clear indications for when a blood culture is warranted for any disease process, including community acquired pneumonia. As my quality improvement work is currently focused on contaminated blood cultures (see below), I think it is important to consider the impact contaminated blood cultures can have on patients, families, and the health care system. Even with a low prevalence rate of 1.2% in non-ICU patients in this study, contaminated blood cultures were 5 times more prevalent than penicillin non-susceptible bacteremia. This article is another example of the importance of value-based care.