Evaluation and comparison of multiple test methods, including real-time PCR, for Legionella detection in clinical specimens.
Legionella is a gram-negative bacterium that can cause Pontiac fever, a mild upper respiratory infection and Legionnaire’s disease, a more severe illness. We aimed to compare the performance of urine antigen, culture and PCR test methods and to determine if sputum is an alternative to the use of mor...
Main Authors: | , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Frontiers Media S.A.
2016-08-01
|
Series: | Frontiers in Public Health |
Subjects: | |
Online Access: | http://journal.frontiersin.org/Journal/10.3389/fpubh.2016.00175/full |
id |
doaj-1017ded906b744969375bc16dcac38cc |
---|---|
record_format |
Article |
spelling |
doaj-1017ded906b744969375bc16dcac38cc2020-11-25T01:01:07ZengFrontiers Media S.A.Frontiers in Public Health2296-25652016-08-01410.3389/fpubh.2016.00175214734Evaluation and comparison of multiple test methods, including real-time PCR, for Legionella detection in clinical specimens.Adriana Peci0Anne-Luise Winter1Jonathan B. Gubbay2PHOPHOPHOLegionella is a gram-negative bacterium that can cause Pontiac fever, a mild upper respiratory infection and Legionnaire’s disease, a more severe illness. We aimed to compare the performance of urine antigen, culture and PCR test methods and to determine if sputum is an alternative to the use of more invasive bronchoalveolar lavage (BAL). Data for this study included specimens tested for Legionella at PHOL from January 1, 2010 to April 30, 2014, as part of routine clinical testing. We found sensitivity of UAT compared to culture to be 87%, specificity 94.7%, positive predictive value (PPV) 63.8% and negative predictive value (NPV) 98.5%. Sensitivity of UAT compared to PCR was 74.7%, specificity 98.3%, PPV 77.7% and NPV 98.1%. Of 146 patients who had a Legionella positive result by PCR, only 66(45.2%) also had a positive result by culture. Sensitivity for culture was the same using either sputum or BAL (13.6%); sensitivity for PCR was 10.3% for sputum and 12.8% for BAL. Both sputum and BAL yield similar results despite testing methods (Fisher Exact p-values=1.0, for each test). In summary, all test methods have inherent weaknesses in identifying Legionella; thereforemore than one testing method should be used. Obtaining a single specimen type from patients with pneumonia limits the ability to diagnose Legionella, particularly when urine is the specimen type submitted. Given ease of collection, and similar sensitivity to BAL, clinicians are encouraged to submit sputum in addition to urine when BAL submission is not practical, from patients being tested for Legionella.http://journal.frontiersin.org/Journal/10.3389/fpubh.2016.00175/fullLegionellaSputumPCRcultureBALUrine antigen |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Adriana Peci Anne-Luise Winter Jonathan B. Gubbay |
spellingShingle |
Adriana Peci Anne-Luise Winter Jonathan B. Gubbay Evaluation and comparison of multiple test methods, including real-time PCR, for Legionella detection in clinical specimens. Frontiers in Public Health Legionella Sputum PCR culture BAL Urine antigen |
author_facet |
Adriana Peci Anne-Luise Winter Jonathan B. Gubbay |
author_sort |
Adriana Peci |
title |
Evaluation and comparison of multiple test methods, including real-time PCR, for Legionella detection in clinical specimens. |
title_short |
Evaluation and comparison of multiple test methods, including real-time PCR, for Legionella detection in clinical specimens. |
title_full |
Evaluation and comparison of multiple test methods, including real-time PCR, for Legionella detection in clinical specimens. |
title_fullStr |
Evaluation and comparison of multiple test methods, including real-time PCR, for Legionella detection in clinical specimens. |
title_full_unstemmed |
Evaluation and comparison of multiple test methods, including real-time PCR, for Legionella detection in clinical specimens. |
title_sort |
evaluation and comparison of multiple test methods, including real-time pcr, for legionella detection in clinical specimens. |
publisher |
Frontiers Media S.A. |
series |
Frontiers in Public Health |
issn |
2296-2565 |
publishDate |
2016-08-01 |
description |
Legionella is a gram-negative bacterium that can cause Pontiac fever, a mild upper respiratory infection and Legionnaire’s disease, a more severe illness. We aimed to compare the performance of urine antigen, culture and PCR test methods and to determine if sputum is an alternative to the use of more invasive bronchoalveolar lavage (BAL). Data for this study included specimens tested for Legionella at PHOL from January 1, 2010 to April 30, 2014, as part of routine clinical testing. We found sensitivity of UAT compared to culture to be 87%, specificity 94.7%, positive predictive value (PPV) 63.8% and negative predictive value (NPV) 98.5%. Sensitivity of UAT compared to PCR was 74.7%, specificity 98.3%, PPV 77.7% and NPV 98.1%. Of 146 patients who had a Legionella positive result by PCR, only 66(45.2%) also had a positive result by culture. Sensitivity for culture was the same using either sputum or BAL (13.6%); sensitivity for PCR was 10.3% for sputum and 12.8% for BAL. Both sputum and BAL yield similar results despite testing methods (Fisher Exact p-values=1.0, for each test). In summary, all test methods have inherent weaknesses in identifying Legionella; thereforemore than one testing method should be used. Obtaining a single specimen type from patients with pneumonia limits the ability to diagnose Legionella, particularly when urine is the specimen type submitted. Given ease of collection, and similar sensitivity to BAL, clinicians are encouraged to submit sputum in addition to urine when BAL submission is not practical, from patients being tested for Legionella. |
topic |
Legionella Sputum PCR culture BAL Urine antigen |
url |
http://journal.frontiersin.org/Journal/10.3389/fpubh.2016.00175/full |
work_keys_str_mv |
AT adrianapeci evaluationandcomparisonofmultipletestmethodsincludingrealtimepcrforlegionelladetectioninclinicalspecimens AT anneluisewinter evaluationandcomparisonofmultipletestmethodsincludingrealtimepcrforlegionelladetectioninclinicalspecimens AT jonathanbgubbay evaluationandcomparisonofmultipletestmethodsincludingrealtimepcrforlegionelladetectioninclinicalspecimens |
_version_ |
1725210671249883136 |