Incidence and Impact of False Positive Urine Pneumococcal Antigen Screening in Hospitalized Patients

Background: Prior studies, conducted largely in patients with community acquired pneumonia, have ascribed high specificity (94-100%) to the Streptococcus pneumonia (pneumococcal) urine immunochromatographic membrane test (ICT). Potential reasons for ICT false positivity include nasopharyngeal carriage, prior invasive pneumococcal infection, and recent 23-valent pneumococcal polysaccharide vaccine (PPV) administration. The performance of this assay has not been studied in a real world clinical setting.

Objective: To evaluate the incidence of and patient characteristics associated with false positive (FP) ICT results, and to determine if clinical management is impacted by these FP results in hospitalized patients.

Methods: We conducted a single-center, retrospective cohort study to assess the incidence of FP ICT results among hospitalized adults from November 21, 2007 to June 30, 2009. All patients with a positive ICT result, identified through an electronic hospital database, had medical records abstracted for demographic and clinical data, documented receipt of PPV, and signs and symptoms of invasive pneumococcal infection. We defined a FP ICT result as one obtained from a patient who met all of the following criteria: 1. No clinical syndrome consistent with invasive pneumococcal infection; 2. No radiographic evidence of lower respiratory tract disease; 3. No microbiological evidence of invasive streptococcal infection; 4. Presence of an alternate diagnosis.  All patients determined to have FP ICT results were evaluated to determine impact to clinical management including initiation, discontinuation, or change in antimicrobial therapy and/or initiation or deferral of further diagnostic work up.

Results: During the study period, ICT results were obtained from 850 patients. Of these patients, 52 (6.1%) had positive ICT results. Of the 52 patients with a positive pneumococcal antigen result, 5 (9.6%) were determined to be FP. All five patients with FP results were immunosuppressed including two who had received PPV in the 5 day period prior to testing (to comply with the Centers for Medicare and Medicaid Services (CMS) core measure for PPV administration). In no case did clinicians suspect FP results and in all 5 instances, clinical management was impacted in some way including overuse of antimicrobials, delay in correct diagnosis, and prolonged hospital stay.

Conclusions: In contrast with prior studies, we demonstrated a high rate of FP ICT results among hospitalized adults. As PPV administration is now occurring earlier in hospitalization at our institution, the potential for increasing FP ICT results exists. Clinicians should be educated about conditions that can lead to FP results and should interpret positive ICT results within a careful clinical framework.