JCM Figure table search 04
Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
This Article
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Koksalan, O. K.
Right arrow Articles by Behr, M. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Koksalan, O. K.
Right arrow Articles by Behr, M. A.

 Previous Article

Journal of Clinical Microbiology, June 2005, p. 3031-3032, Vol. 43, No. 6
0095-1137/05/$08.00+0     doi:10.1128/JCM.43.6.3031-3032.2005

LETTER TO THE EDITOR

Low Positive Predictive Values and Specificities of Spoligotyping and Mycobacterial Interspersed Repetitive Unit-Variable-Number Tandem Repeat Typing Methods for Performing Population-Based Molecular Epidemiology Studies of Tuberculosis


   LETTER
Top
 Letter
References
 Letter 
 References 
 
According to a recent paper by Scott and colleagues (1), the specificities and the positive predictive values of spoligotyping and mycobacterial interspersed repetitive unit-variable-number tandem repeat typing (MIRU-VNTR), compared with those of IS6110 restriction fragment length polymorphism (RFLP) analysis, are too low for these methods to be useful for studying population-based molecular epidemiology of tuberculosis.

However, the low calculated specificities of the MIRU-VNTR and spoligotyping methods sound controversial, for the authors took IS6110 RFLP results to be entirely true. Unfortunately, however, no molecular epidemiological method for tuberculosis, including IS6110 RFLP analysis, is 100% sensitive and specific. This is why each new molecular epidemiological method is evaluated most correctly by comparison of its results with the data obtained by conventional epidemiological surveillances. Otherwise, as was the case in the cited study (1), it is not easy to decide whether and to what extent the low calculated specificities of the newer methods really depict their lesser discriminatory power or—just in contrast—whether and to what extent these low calculated specificities are caused by false discrimination of isolates by the IS6110 RFLP method which are in fact epidemiologically related.

On the other hand, there seem to be two reasons for the finding of the study that MIRU-VNTR and spoligotyping have positive predictive values as low as 10%. Firstly, the positive predictive value of a test is primarily affected by the specificity; i.e., the lower the specificity the lower the positive predictive value of a test will be. In other words, the low calculated specificities in this study dramatically decreased the positive predictive values of both tests. Secondly, these low positive predictive values were further decreased due to the nature of the study population. The very scarce number of clustered isolates and relatively higher number of isolates with orphan types obtained in this study, which is typical for developed countries, is a contributing factor for such unusually low positive predictive values. Generally, in the developed world, where tuberculosis is rare and under control, as in Montreal, recent transmissions and clustering are expected to be low.

Keeping the points in mind that more than 95% of the new cases of tuberculosis in the world occur in developing countries (2) and that the low rates of the MIRU-VNTR and spoligotyping obtained in this study were based solely on IS6110 RFLP comparison and not on data based on conventional epidemiological surveillances, the conclusions of the study should be interpreted cautiously.


   REFERENCES
Top
 Letter
 References
Letter 
 References 
 

  1. Scott, A. N., D. Menzies, T. Tannenbaum, L. Thiebert, R. Kozak, L. Joseph, K. Schwartzman, and M. A. Behr. 2005. Sensitivities and specificities of spoligotyping and mycobacterial interspersed repetitive unit-variable-number tandem repeat typing methods for studying molecular epidemiology of tuberculosis. J. Clin. Microbiol. 43:89-94.[Abstract/Free Full Text]
  2. World Health Organization. 2003. Global tuberculosis control: surveillance, planning, financing. Report W.H.O./CDS/TB2003.316. World Health Organization, Geneva, Switzerland.
O. Kaya Koksalan*
Istanbul University
Institute for Medical Experimental Research
Asagi Vakif Gureba cd
34270 Istanbul, Turkey

* Phone: 00 90 212 631 1351, Fax: 0090 212 631 1351-113, E-mail: okkoksalan{at}hotmail.com


Authors' Reply


   LETTER 
Top
 Letter
 References
 Letter 
References 
 
We appreciate Dr. Koksalan's interest in our report and agree that the analysis of diagnostic tests in the absence of a gold standard can be problematic. Very few diagnostic tests can be considered true gold standards, so unless complex modeling is employed, the results of a study such as this one will always be subject to greater or lesser amounts of bias. However, we feel using IS6110 RFLP as the comparator is justified based on over a decade of demonstrated utility since it was first proposed as the standard method for tuberculosis molecular epidemiology (6). Moreover, studies comparing IS6110 RFLP to the previous standard, conventional contact tracing, have shown that named contacts provide relatively large numbers of false-negative (5) and false-positive (1) results. This suggests, but does not prove, that IS6110 RFLP is a better marker for tuberculosis (TB) transmission. In an attempt to address this issue, we are currently studying the use of a Bayesian analysis approach to estimate test characteristics without the assumption of a gold standard (4).

We agree that positive and negative predictive values depend on the prevalence of a disease in a community, which for this example is the prevalence of recently transmitted TB in the defined study setting. Therefore the predictive values we reported here are specific to Montreal and other areas with similar prevalence rates and population densities. Studies with a higher incidence of TB in a tight geographic framework are likely to have a higher pretest probability of TB transmission between two individuals (7). Conversely, national and international databases providing the opportunity to compare genotypes between persons separated by thousands of kilometers are likely to present a lower pretest probability of TB transmission (2, 3), thereby producing more false-positive results and a lower positive predictive value. We urge other researchers to replicate our study in other settings and determine the predictive values of these typing methods for their own communities.

Finally, we emphatically agree that there is an important disconnect between the sites where much TB research is conducted and the global burden of TB. Efforts to perform research in high-burden countries should be encouraged. While we acknowledge the limitations of our study setting, we hope that our results contribute information applicable beyond our community.


   REFERENCES 
Top
 Letter
 References
 Letter 
 References 
 

  1. Behr, M. A., P. C. Hopewell, E. A. Paz, L. M. Kawamura, G. F. Schecter, and P. M. Small. 1998. Predictive value of contact investigation for identifying recent transmission of Mycobacterium tuberculosis. Am. J. Respir. Crit. Care Med. 158:465-469.[Abstract/Free Full Text]
  2. Cowan, L. S., and J. T. Crawford. 2002. Genotype analysis of Mycobacterium tuberculosis isolates from a sentinel surveillance population. Emerg. Infect. Dis. 8:1294-1302.[Medline]
  3. Filliol, I., J. R. Driscoll, D. Van Soolingen, B. N. Kreiswirth, K. Kremer, G. Valetudie, D. D. Anh, R. Barlow, D. Banerjee, P. J. Bifani, K. Brudey, A. Cataldi, R. C. Cooksey, D. V. Cousins, J. W. Dale, O. A. Dellagostin, F. Drobniewski, G. Engelmann, S. Ferdinand, D. Gascoyne-Binzi, M. Gordon, M. C. Gutierrez, W. H. Haas, H. Heersma, G. Kallenius, E. Kassa-Kelembho, T. Koivula, H. M. Ly, A. Makristathis, C. Mammina, G. Martin, P. Mostrom, I. Mokrousov, V. Narbonne, O. Narvskaya, A. Nastasi, S. N. Niobe-Eyangoh, J. W. Pape, V. Rasolofo-Razanamparany, M. Ridell, M. L. Rossetti, F. Stauffer, P. N. Suffys, H. Takiff, J. Texier-Maugein, V. Vincent, J. H. de Waard, C. Sola, and N. Rastogi. 2002. Global distribution of Mycobacterium tuberculosis spoligotypes. Emerg. Infect. Dis. 8:1347-1349.[Medline]
  4. Joseph, L., T. W. Gyorkos, and L. Coupal. 1995. Bayesian estimation of disease prevalence and the parameters of diagnostic tests in the absence of a gold standard. Am. J. Epidemiol. 141:263-272.[Abstract/Free Full Text]
  5. Small, P. M., P. C. Hopewell, S. P. Singh, A. Paz, J. Parsonnet, D. C. Ruston, G. F. Schecter, C. L. Daley, and G. K. Schoolnik. 1994. The epidemiology of tuberculosis in San Francisco. A population-based study using conventional and molecular methods. N. Engl. J. Med. 330:1703-1709.[Abstract/Free Full Text]
  6. Van Embden, J. D., M. D. Cave, J. T. Crawford, J. W. Dale, K. D. Eisenach, B. Gicquel, P. Hermans, C. Martin, R. McAdam, T. M. Shinnick, and. 1993. Strain identification of Mycobacterium tuberculosis by DNA fingerprinting: recommendations for a standardized methodology. J. Clin. Microbiol. 31:406-409.[Abstract/Free Full Text]
  7. Verver, S., R. M. Warren, Z. Munch, E. Vynnycky, P. D. van Helden, M. Richardson, G. D. van der Spuy, D. A. Enarson, M. W. Borgdorff, M. A. Behr, and N. Beyers. 2004. Transmission of tuberculosis in a high incidence urban community in South Africa. Int. J. Epidemiol. 33:351-357.[Abstract/Free Full Text]
Allison N. Scott
Department of Epidemiology and Biostatistics
McGill University
Montreal, Quebec, Canada

Lawrence Joseph
Department of Epidemiology and Biostatistics
McGill University
Montreal, Quebec, Canada

Marcel A. Behr*
Division of Infectious Diseases and Microbiology, A5.156, Montreal General Hospital, 1650 Cedar Ave., Montreal, Quebec, Canada H3G 1A4

* Phone: (514) 934-1934, ext. 42815, Fax: (514) 934-8423, E-mail: marcel.behr{at}mcgill.ca

Journal of Clinical Microbiology, June 2005, p. 3031-3032, Vol. 43, No. 6
0095-1137/05/$08.00+0     doi:10.1128/JCM.43.6.3031-3032.2005





This Article
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Koksalan, O. K.
Right arrow Articles by Behr, M. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Koksalan, O. K.
Right arrow Articles by Behr, M. A.

Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
Antimicrob. Agents Chemother. Clin. Microbiol. Rev.
Clin. Vaccine Immunol. ALL ASM JOURNALS