Journal of Clinical Virology 30 (2004) 165–174 Evaluation of a multiplex reverse transcriptase PCR ELISA for the detection of nine respiratory tract pathogens W. Puppe a,∗ , J.A.I. Weigl a,b , G. Aron c , B. Gröndahl b , H.-J. Schmitt b , H.G.M. Niesters c , J. Groen c a Department of Pediatrics, Pediatric Infectious Diseases, University Hospital Schleswig–Holstein Campus Kiel, Schwanenweg 20, 24105 Kiel, Germany b Department of Pediatrics, Pediatric Infectious Diseases, Johannes Gutenberg University, Laagenbeckstraße 1, 55101 Mainz, Germany c Institute of Virology, Erasmus MC, 3015 GD Rotterdam, The Netherlands Received 26 March 2003; received in revised form 26 July 2003; accepted 6 October 2003 Abstract Background: A multiplex reverse transcription (RT) polymerase chain reaction combined with a microwell hybridization assay (m-RT-PCR- ELISA) was previously developed to detect nine different microorganisms: enterovirus (EV), influenza virus type A (IVA) and type B (IVB), respiratory syncytial virus (RSV), parainfluenzavirus type 1 (PIV1) and type 3 (PIV3), adenovirus (AV), Mycoplasma pneumoniae (Mpn), Chlamydia pneumoniae (Cpn) in a single test. These organisms do not usually colonize the respiratory tract of humans, but, if present, it may be assumed they are involved in respiratory disease. Objectives and study design: The m-RT-PCR-ELISA was tested on (i) culture supernatants of unknown contents, (ii) by determining the analytical sensitivity of 10-fold serial dilutions of culture supernatants and (iii) by determining clinical sensitivity in a retrospective study on 411 clinical specimens. The specimens were re-tested in parallel by m-RT-PCR-ELISA versus the gold standard culture and immunfluorescence, and versus individual RT-PCR. Results: (i) The 9-valent m-RT-PCR-ELISA shows 83% to 100% concordant results on 103 culture supernatants containing different organisms. (ii) The analytical sensitivity was as follows: higher sensitivity of the 9-valent m-RT-PCR-ELISA in comparison to culture in the cases of PIV3, IVA and IVB (factor 10) and AV and EV (factor 100), and lower sensitivity in case of RSV and PIV1 (factor 10). (iii) The agreement with the gold standard in the kappa statistic was excellent for RSV (κ = 0.937), IVA (κ = 0.940), very good for PIV1 (κ = 0.914), IVB (κ = 0.907) and satisfactory for PIV3 (κ = 0.410). For AV, EV and Mpn the m-RT-PCR-ELISA preliminary could be qualified as very good, based on the data derived on culture supernatants. Information about the validity for Cpn is limited. Conclusion: The m-RT-PCR-ELISA is a feasible, sensitive and specific method for detection of a broad spectrum of organisms. It is suitable for individual as well as epidemiological diagnosis. © 2003 Elsevier B.V. All rights reserved. Keywords: m-RT-PCR-ELISA; Respiratory tract infections; Validation; Kappa-statistics; Sensitivity 1. Introduction Acute respiratory tract infections (ARI) inflict a high burden of disease in children worldwide (Leowski, 1986). Streptococcus pneumoniae, Haemophilus influenzae and Moraxella catarrhalis (Bartlett and Mundy, 1995) are the most common bacteria causing ARI. However, as commen- sals of the upper respiratory tract, they usually contaminate sputum samples, nasopharyngeal aspirates or swabs, and ∗ Corresponding author. Present address: Universitätsklinikum Schleswig-Holstein Campus Kiel, Allgemeine Pädiatrie, Pädiatrische In- fektiologie, Schwanenweg 20, 24105 Kiel, Germany. Tel.: +49-431-597-1679; fax: +49-431-597-1680. E-mail address: puppe@pediatrics.uni-kiel.de (W. Puppe). thus their etiological role in ARI is impossible to prove by upper respiratory tract sampling. Evidence of acute infec- tion is given by the detection of Mpn, Cpn and also by the detection of viruses in a child with respiratory symptoms in most instances. There are numerous diseases as well as numerous pathogens involved in ARI, and precise data about causative pathogens are important for individual care as well as for intervention strategies in public health. Molecular am- plification techniques are advantageous in getting more epidemiological data about these pathogens, because they are faster and less expensive than other methods, and do not need viable organisms. The logistic for specimen ship- ment is therefore simple, and the detection of particularly fastidious or “difficult to culture” organisms, such as Mpn 1386-6532/$ – see front matter © 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.jcv.2003.10.003