INT J TUBERC LUNG DIS 15(12):1566 © 2011 The Union http://dx.doi.org/10.5588/ijtld.11.0654
EDITORIAL
Contact tracing in low-incidence tuberculosis settings THE CHOICE of a specific test to assess a health outcome is driven by multiple factors, including test characteristics, operational issues and costs. The conventional test characteristics are assessed in rigorously conducted validation studies estimating sensitivity, specificity and predictive value. Comparing the characteristics of different tests is not without pitfalls: the setting in which tests are evaluated influences sensitivity and specificity, while the predictive value is influenced by the prevalence of the health status in the study population. These issues are complicated by the potential misclassification of the health status (poor definition of a gold standard), or the test outcome (variation around the level of quantification), and additional biases.1 In this issue of the Journal, Grinsdale et al. describe a comparison of different tests beyond the conventional test characteristics.2 They compared the programmatic outcomes of contact tracing using the QuantiFERON®-TB Gold (QFT-G) test and the tuberculin skin test (TST) for the assessment of Mycobacterium tuberculosis infection. The authors show that identified contacts of culture-confirmed TB patients were more likely to complete evaluation, and to start and complete isoniazid preventive therapy (IPT), when they were tested using the QFT-G test compared to the TST. Moreover, the use of QFT-G in this setting (in San Francisco) was supported by the observation of a stronger association with exposure than TST. These results might be surprising at first glance. Why would the type of test have a differential effect on willingness to start therapy or adherence? The answer to that question might be in one of the limitations that the authors describe. The choice of test used was not random, but was decided by the Disease Control Investigator (DCI) who performed the contact tracing. Although the analyses were adjusted for the different DCIs, and effect modification was excluded, there will be room for residual confounding. It might well be that the DCIs put more confidence in a positive QFT-G as a marker for latent tuberculosis infection than in a positive TST. This increased confidence in the test result might have made the DCIs more persuasive towards the contact to start IPT, and instigate a more rigorous adherence strategy. It is well known that provider-patient interaction is strongly associated with adherence to medication.3
Another important finding was the staggering loss to follow-up between the first and second visit. A second test is required when the first test is negative within 8 weeks after exposure, and it is important, as converters between the first and second test are at high risk of tuberculosis. Although the proportion of loss to follow-up was lower when using QFT-G (50%) than TST (66%), there is also enormous scope for improvement with QFT-G. These results provide interesting arguments for public health programmes in low-incidence countries where contact tracing plays a vital role in tuberculosis control. An anticipated fear of low IPT adherence rates in these settings is often used as an argument to provide no preventive treatment at all. The study by Grinsdale et al. shows that such fears may be misplaced. Rather than fears about adherence rates, the real challenge lies in improving the completeness of contact follow-up. Frank van Leth*† Martien Borgdorff‡¶ *Department of Global Health Academic Medical Center University of Amsterdam Amsterdam Institute for Global Health and Development † KNCV Tuberculosis Foundation ‡ Public Health Service Amsterdam ¶ Department of Clinical Epidemiology and Biostatistics Academic Medical Center University of Amsterdam Amsterdam, The Netherlands e-mail:
[email protected]
References 1 Whiting P, Rutjes A W S, Dinnes J, Reitsma J, Bossuyt P M M, Kleijnen J. Development and validation of methods for assessing the quality of diagnostic accuracy studies. Health Technol Assess 2004; 8 (25): 1–234. 2 Grinsdale J A, Ho C S, Banouvong H, Kawamura L M. Programmatic impact of using QuantiFERON®-TB Gold in routine contact investigation activities. Int J Tuberc Lung Dis 2011; 15: 1614–1619. 3 Munro S A, Lewin S A, Smith H J, Engel M E, Fretheim A, Volmink J. Patient adherence to tuberculosis treatment: a systematic review of qualitative research. PLoS Med 2007; 4: e238.
