Sensitivity, specificity, and positivity predictors of the pneumococcal urinary antigen test in community-acquired pneumonia

Abstract

Rationale: Detection of the C-polysaccharide of Streptococcus pneumoniae in urine by an immune-chromatographic test is increasingly used to evaluate patients with community-acquired pneumonia.

Objectives: We assessed the sensitivity and specificity of this test in the largest series of cases to date and used logistic regression models to determine predictors of positivity in patients hospitalized with community-acquired pneumonia.

Methods: We performed a multicenter, prospective, observational study of 4,374 patients hospitalized with community-acquired pneumonia.

Measurements and Main Results: The urinary antigen test was done in 3,874 cases. Pneumococcal infection was diagnosed in 916 cases (21%); 653 (71%) of these cases were diagnosed exclusively by the urinary antigen test. Sensitivity and specificity were 60 and 99.7%, respectively. Predictors of urinary antigen positivity were female sex; heart rate ≥125 bpm, systolic blood pressure <90 mm Hg, and SaO2 <90%; absence of antibiotic treatment; pleuritic chest pain; chills; pleural effusion; and blood urea nitrogen ≥30 mg/dl. With at least six of all these predictors present, the probability of positivity was 52%. With only one factor present, the probability was only 12%.

Conclusions: The urinary antigen test is a method with good sensitivity and excellent specificity in diagnosing pneumococcal pneumonia, and its use greatly increased the recognition of community-acquired pneumonia due to S. pneumoniae. With a specificity of 99.7%, this test could be used to direct simplified antibiotic therapy, thereby avoiding excess costs and risk for bacterial resistance that result from broad-spectrum antibiotics. We also identified predictors of positivity that could increase suspicion for pneumococcal infection or avoid the unnecessary use of this test.