Fellow: Christiane Lenzen, University of California San Diego, Rady Children's Hospital
Article: Stockman CS, Ampofo K, Killpack J, et al. Procalcitonin Accurately Identifies Hospitalized Children with Low Risk of Bacterial Community-Acquired Pneumonia. J Pediatric Infect Dis Soc. 2017; 1-8. https://doi.org/10.1093/jpids/piw091
Summary:
The authors examined the association between serum procalcitonin (PCT) levels with the type of pathogen and the severity of disease in a large cohort of hospitalized children (EPIC study/'etiology of PNA in the Community') with radiographically confirmed community-acquired pneumonia (CAP). Findings indicate children with typical bacterial infection had significantly higher PCT concentrations (p<0.001) than patients with atypical, viral, or unknown etiology. Low PCT levels yielded a high negative predictive value for typical bacterial infection (96-100% for cut-off PCT levels <0.25 vs 0.1 ng/ml) but high PCT levels had poor positive predictive value (15-17%). Overall lower PCT levels were associated with less severe disease, while higher PCT was associated with parapneumonic empyema.
What are the key strengths of the article?
The study utilized a highly sensitive PCT assay in a large cohort (>500) prospectively enrolled to evaluate the etiology of PNA. This cohort underwent more comprehensive diagnostic testing with one or more identified pathogen in 91% of patients.
Are there any limitations or flaws in the article?
Limitations include that the PCT testing was performed on residual sera, which were not available from all patients enrolled in the PIC study. Those with sera available were more frequently in the intensive care unit and skew to sicker patients, given more frequent laboratory testing in this population. The classifications allowed for co-infections with viruses in the typical bacterial pathogens and the atypical bacterial pathogens. Another limitation is that no serial testing was performed and all values are only representative of the first day of hospitalization.
What is the major takeaway message?
The study indicates very good negative predictive value for typical bacterial infection with low PCT values (100% for < 0.1ng/ml; 96% for < 0.25ng/ml), which could assist in identifying patients at low risk for typical bacterial infection. There is a potential for decreased antibiotic utilization. Higher PCT values demonstrated poor positive predictive value of typical bacterial infection given overlap with viral or atypical bacterial infections (65% of patients with PCT level> 0.5 had only atypical or viral etiology of CAP).
Describe how this article should impact our practice.
The vast majority of children hospitalized for CAP receive antibiotics. The EPIC study has shown that only 15% of these children have bacterial infection (8% with typical bacterial), but it remains difficult to distinguish the ones at low risk for typical bacterial infection. Given the excellent negative predictive value of low PCT levels, incorporating procalcitonin levels in treatment algorithms could help identifying these low risk patients and reduce unnecessary antibiotic treatments, thus improving antibiotic stewardship and save cost.
Article: Stockman CS, Ampofo K, Killpack J, et al. Procalcitonin Accurately Identifies Hospitalized Children with Low Risk of Bacterial Community-Acquired Pneumonia. J Pediatric Infect Dis Soc. 2017; 1-8. https://doi.org/10.1093/jpids/piw091
Summary:
The authors examined the association between serum procalcitonin (PCT) levels with the type of pathogen and the severity of disease in a large cohort of hospitalized children (EPIC study/'etiology of PNA in the Community') with radiographically confirmed community-acquired pneumonia (CAP). Findings indicate children with typical bacterial infection had significantly higher PCT concentrations (p<0.001) than patients with atypical, viral, or unknown etiology. Low PCT levels yielded a high negative predictive value for typical bacterial infection (96-100% for cut-off PCT levels <0.25 vs 0.1 ng/ml) but high PCT levels had poor positive predictive value (15-17%). Overall lower PCT levels were associated with less severe disease, while higher PCT was associated with parapneumonic empyema.
What are the key strengths of the article?
The study utilized a highly sensitive PCT assay in a large cohort (>500) prospectively enrolled to evaluate the etiology of PNA. This cohort underwent more comprehensive diagnostic testing with one or more identified pathogen in 91% of patients.
Are there any limitations or flaws in the article?
Limitations include that the PCT testing was performed on residual sera, which were not available from all patients enrolled in the PIC study. Those with sera available were more frequently in the intensive care unit and skew to sicker patients, given more frequent laboratory testing in this population. The classifications allowed for co-infections with viruses in the typical bacterial pathogens and the atypical bacterial pathogens. Another limitation is that no serial testing was performed and all values are only representative of the first day of hospitalization.
What is the major takeaway message?
The study indicates very good negative predictive value for typical bacterial infection with low PCT values (100% for < 0.1ng/ml; 96% for < 0.25ng/ml), which could assist in identifying patients at low risk for typical bacterial infection. There is a potential for decreased antibiotic utilization. Higher PCT values demonstrated poor positive predictive value of typical bacterial infection given overlap with viral or atypical bacterial infections (65% of patients with PCT level> 0.5 had only atypical or viral etiology of CAP).
Describe how this article should impact our practice.
The vast majority of children hospitalized for CAP receive antibiotics. The EPIC study has shown that only 15% of these children have bacterial infection (8% with typical bacterial), but it remains difficult to distinguish the ones at low risk for typical bacterial infection. Given the excellent negative predictive value of low PCT levels, incorporating procalcitonin levels in treatment algorithms could help identifying these low risk patients and reduce unnecessary antibiotic treatments, thus improving antibiotic stewardship and save cost.