When should rapid viral testing occur for the diagnosis of CAP?

-We strongly recommended based on high quality evidence obtaining rapid viral testing (RSV/influenza and/or Respiratory Viral PCR) during the appropriate seasons.
Differentiation of bacterial pneumonia from viral lower respiratory tract illness is difficult on clinical grounds. Multiple studies have identified that rapid viral studies are useful during the appropriate seasons; during influenza season, testing is indicated in febrile patients of all ages (≥ 39°) who present with or without respiratory symptoms (Woo, Chiu et al. 1997; Abanses, Dowd et al. 2006; Mandell, Wunderink et al. 2007). Testing is typically performed December to March for influenza and October to March for RSV. Start and end dates of testing are determined by the department of microbiology (Local Expert Consensus). The British Thoracic Society guideline recommends viral PCR and/or immunofluorescence to aid in an alternative diagnosis to bacterial pneumonia. The PIDS/IDSA guideline recommends both rapid viral testing (influenza and RSV) as well as testing for other viral pathogens (respiratory viral PCR at CMH). Identification of a viral pathogen has a strong potential to alter treatment options and limit unnecessary antibiotic use (Bradley, Byington et al. 2011; Harris, Clark et al. 2011).

Should a CBC or inflammatory marker laboratory tests be obtained to assist in the diagnosis of CAP?

-We strongly recommended based on moderate quality evidence not routinely obtaining labs, including CBC and inflammatory markers, for patients with suspected bacterial CAP.

Patients evaluated for PNA in the outpatient setting will unlikely benefit from the addition of laboratory analysis including CBC and inflammatory markers for diagnosis or treatment. The British Thoracic Society guideline recommendations discuss that acute phase reactants and CRP do not distinguish viral from bacterial lower respiratory tract infection and therefore should not routinely be performed (Harris, Clark et al. 2011). The PIDS/IDSA guideline recommendations agree, specifically that acute phase reactants or inflammatory markers cannot distinguish bacterial from viral causes. However, for patients that require hospitalization and/or have serious disease, acute phase reactants may be beneficial for clinical management or to assess response to therapy (Bradley, Byington et al. 2011). Flood et al, published a meta-analysis including 8 studies and 1230 children with the goal to determine the utility of CRP as a predictor for bacterial pneumonia; concluding that a serum CRP >40-60 mg/L (4-6mg/dL as reported by CMH laboratory) is weakly predictive of a bacterial etiology. Included studies in this analysis were only required to have a criteria applied to differentiate bacterial from nonbacterial viral pneumonia, allowing for heterogeneity of diagnosis between studies (Flood, Badik et al. 2008).

Complete blood count is also not recommended by the PIDS/IDSA guideline for patients in the outpatient setting as routine measurements are unlikely to aid diagnosis or change management. For patients considered to have severe disease, it is suggested that a CBC with differential may be beneficial when interpreted in conjunction with other labs, cultures and imaging (Bradley, Byington et al. 2011). Three recent studies have looked at the association of total WBC count and the diagnosis of PNA on CXR. Brauner et al found an increased rate of serious bacterial infection (SBI), specifically pneumonia in a retrospective review of 146 Israeli patients. Patients with extreme leukocytosis described as WBC >25,000 had segmental or lobar pneumonia diagnosed in 41 (28%) patients compared to 27 (9.2%) controls with moderate leukocytosis of 15,000-24,999 (p<0.001, OR 3.83, 95% CI 2.25 to 6.52) (Brauner, Goldman et al. 2010). Shah and colleagues published a similar retrospective review and found that for patients less than 24 months of age being evaluated in the ED for fever, the rates of SBI for patients with extreme leukocytosis and modest leukocytosis were similar. Specifically PNA rates were 13% and 15% between the two groups (Shah, Shofer et al. 2005). Finally, Don et al published a prospective analysis of 68 children who presented in Italy for suspected CAP. Lab studies including WBC, ESR, CRP and procalcitonin were obtained as well as methods for determination of pneumonia etiology including immunoglobulins and enzyme immunoassay. When a combination of lab values was found to be elevated, the likelihood ratio for bacterial etiology was increased. However and likely more importantly, low values did not rule out bacterial etiology (Don, Valent et al. 2009). 

When should blood cultures be obtained for the diagnosis of CAP?

-We strongly recommended based on moderate quality evidence obtaining a blood culture in patients requiring hospitalization.

Results of a blood culture in a hospitalized child suspected of bacterial PNA has the potential to significantly alter management by identification of a pathogen and altering both antibiotic choice and length of therapy. However, the yield of a positive culture in a patient that does not require hospitalization is low and potentially reaches the rate of false positive cultures; leading to unnecessary repeat cultures, antibiotics and possible hospitalizations. Neither the British Thoracic Society or PIDS/IDSA guidelines recommend obtaining a blood culture in patients treated as an outpatient (Bradley, Byington et al. 2011; Harris, Clark et al. 2011). 

Patients determined to be ill enough to require hospitalization, specifically patients admitted to the ICU or patients suspected of pneumonia complications are recommended by both guidelines to have a blood culture obtained (Bradley, Byington et al. 2011; Harris, Clark et al. 2011). Bacteremia rates have been published as high as 15-25% in patients with complicated pneumonia (Byington, Spencer et al. 2002) when strict definitions of pneumonia were applied. Shah et al evaluated the results of blood cultures obtained in the ED in Philadelphia in 2011. Of 877 patients evaluated, 291 (33.2%) had blood cultures obtained. The overall prevalence of bacteremia was 2.1% (CI 0.8-4.4%), with 2.6% (CI 1.0-5.6%), of patients with an infiltrate on CXR and 13% (CI 2.8-33.6%), of patients with a complicated PNA. The study found an overall contamination rate of 1.0% (CI 0.2-3.0%). Blood cultures were more likely to be obtained in patients suggestive of severe disease including hypoxic patients. No bacteremia was identified in patients determined well enough to be discharged home or in patients without infiltrate identified on CXR (Shah, Dugan et al. 2011). 

Since patients with complicated pneumonia are not exclusively identified when lab evaluation is obtained, we recommend obtaining a blood culture on all patients considered for hospitalization. Data from CMH in the 12 months after the original CAP CPG was published identified that only 54% of admitted patients with ICD9 codes for CAP had a blood culture obtained despite it being recommended for all inpatients being treated for CAP (Newman, Hedican et al. 2012).

When should TB testing occur for the diagnosis of CAP?

-We strongly recommended based on low quality evidence testing for TB in high risk patients.

Tuberculosis infection caused by Mycobacterium tuberculosis complex needs to be considered in the differential diagnosis of a patient with a lower respiratory tract infection (LRTI). High risk patients for latent tuberculosis infection include: immigrants, international adoptees, refugees from high prevalence areas (including Asia, Africa, Latin America and former Soviet Union countries), homeless persons and incarcerated persons. Infants and post-pubertal adolescents are at increased risk for progression of latent TB to tuberculosis LRTI with other predictive factors including immunodeficiency, use of immunodeficiency drugs including high dose corticosteroids, chemotherapy, diabetes mellitus, chronic renal failure and malnutrition (Pickering 2009). Testing for TB in indicated by tuberculin skin test (TST) in children less than 5 years of age. For children greater than 5 years of age, Quanti-FERON-TB Gold assay can be used and is recommended by this guideline(Mazurek, Jereb et al. 2010).


These guidelines do not establish a standard of care to be followed in every case. It is recognized that each case is different and those individuals involved in providing health care are expected to use their judgment in determining what is in the best interests of the patient based on the circumstances existing at the time. It is impossible to anticipate all possible situations that may exist and to prepare guidelines for each. Accordingly these guidelines should guide care with the understanding that departures from them may be required at times.

Copyright © 1996-2014 The Children's Mercy Hospital