Skip Navigation LinksHome > May/June 2012 - Volume 18 - Issue 3 > Variations in Positive Predictive Values for Rapid Influenza...
Journal of Public Health Management & Practice:
doi: 10.1097/PHH.0b013e3182214730
Research Articles

Variations in Positive Predictive Values for Rapid Influenza Tests for 2009 Pandemic Influenza A (pH1N1)—Arizona, April-October 2009

Baty, Steven A. DVM, MPH; D'Souza, Aarikha MPH; Sunenshine, Rebecca MD; Bisgard, Kris DVM, MPH; Erhart, Laura M. MPH

Collapse Box


Context: Rapid influenza diagnostic tests (RIDTs) are used for influenza screening, clinical decision making, and influenza surveillance. In August 2009, a hospital reported increased false-positive RIDT results to the Arizona Department of Health Services. Because of reported RIDT low sensitivities (40%-62%) for 2009 pandemic influenza A (pH1N1), the hospital's report raised further concerns about the specificity and clinical utility of RIDTs.

Objective: To determine the positive predictive value (PPV) of RIDTs compared with real-time reverse transcription–polymerase chain reaction assay (rRT-PCR) using Centers for Disease Control and Prevention (CDC) protocols and primers as a standard.

Design: A standardized survey collected information including RIDT brand/lot number, training of personnel performing test, type of laboratory, swab and specimen type, time from collection to testing, sample storage, and viral transport medium.

Setting: Arizona.

Participants: Seven Arizona laboratories submitted positive RIDT clinical samples to Arizona State Public Health Laboratory (ASPHL) for confirmatory rRT-PCR testing.

Main Outcome Measure: The PPV was calculated on the basis of rRT-PCR-positive results for April through October.

Results: Results from 600 specimens using 1 of 4 RIDTs were available. Median pH1N1 PPV was 80% (range: 62%–91%) when calculated by RIDT brand. A significant difference in PPV was identified between the 2 largest facilities, which used the same RIDT brand, BinaxNOW Influenza A&B, (Laboratories A, 33% and B, 92%, [P < .01]). The facilities reported similar testing practices except lot numbers used and timing of testing. Laboratory A used lot 003684 and performed testing within 1 hour of collection; Laboratory B used multiple lots, excluding lot 003684, and performed testing within 24 hours. Laboratory A switched RIDT brands and noted a significant PPV increase from 33% to 91% (P < .01).

Conclusions: Wide PPV variability combined with documented low sensitivity among RIDTs for pH1N1 diagnosis increases concerns about their specificity and clinical and epidemiologic utility for influenza.

© 2012 Lippincott Williams & Wilkins, Inc.


Article Level Metrics

Search for Similar Articles
You may search for similar articles that contain these same keywords or you may modify the keyword list to augment your search.