ARCHIVED - Tuberculin skin testing of close contacts: recent or longstanding infection?
15 June 2006 Volume 32 Number 12
Introduction
Pulmonary tuberculosis (TB) screening of close contacts is a component of all TB surveillance and control programs in low-incidence countries(1-5). Once a diagnosis has been rendered and treatment of the index case has commenced, every person in that person's immediate environment is offered: the tuberculin skin test (TST), a chest X-ray, as well as preventive treatment if a latent tubercular infection is diagnosed.
In January, the Comité québécois pour la surveillance et le contrôle de la TB undertook to assess the manner in which TB cases and close contacts are managed by Quebec's 18 public health branches (PHBs). The goals of this contact investigation evaluation project were to: (1) calculate the number of contacts identified per index case; (2) calculate the proportion of close contacts who presented a positive screening result; and (3) identify the factors associated with a positive screening result.
Methodology
Quebec (population: 7,421,053) is divided into 18 health regions, each of which has a PHB that oversees the province's TB surveillance and control program. When a physician or laboratory reports a TB case, a public health nurse gathers the usual clinical and epidemiological data, determines whether adequate treatment has been prescribed, and ensures that the patient is complying with his/her treatment regime. In cases of pulmonary TB, the nurse ensures that close contacts are identified, are given the TST, and are referred for medical assessment, as required. The TST is administered by injecting 0.1 mL of 5 TU tuberculin intradermally in the volar aspect of the forearm (Mantoux method) and by reading the millimetres (mm) of induration between 48 and 72 hours after injection. Close contacts with an initial TST (TST1 ) of ≤ 5 mm are given a second test (TST2 ). In 1997, the recommendation in Quebec was to perform the second TST 12 weeks after the last exposure to the index case. Screening information and results are collected on a standardized form(6).
In February 2001, a standardized questionnaire was provided to all PHBs for every TB case reported in 1997 and 1998. The questionnaire was designed to collect the usual epidemiological data, as well as a list of close contacts (for cases of pulmonary TB only), including the latter's sex, age, TST1 result, and TST2 result, if any. Data on index cases were derived from provincial reportable disease records.
A close contact is defined as any person who shares domestic or other quarters several hours a day, over an extended period, with a person who has active tuberculosis of pulmonary or laryngeal origin, at a time when the latter is contagious(6).
In order to be included in the final analysis of factors associated with a positive screening result, a contact was required to have undergone two TSTs, unless the TST1 was ≥ 5 mm or had been given at least 12 weeks after the last exposure to the index case. Tuberculin conversion was defined as a TST2 of ≥ 5 mm and an increase of ≥ 5 mm between TST1 and TST2 . Screening was considered to have produced a positive result in the presence of a TST1 of ≥ 5 mm or in the presence of tuberculin conversion.
The data were entered into a Microsoft Access version 97 database and analyzed with SPSS, version 10.1. Pearson's chi-square test was used to compare discreet variables(7). Logistic regression was used to identify possible associations between a positive screening result and the variables found to be significant through univariate analysis. Rate ratios (RR) with confidence intervals (CI) of 95% were calculated.
Results
In 1997 and 1998, 649 cases of TB were reported in Quebec (average annual incidence rate: 4.4/100,000), including 372 cases of pulmonary TB. Of the latter, 59.4% were men, 6.7% were between 0 and 18 years of age, 54.6% were between 19 and 64 years of age, and 38.7% were 65 years of age or older (average age: 53.1 years). In terms of ethnic origins, 51.3% were born in Canada, 43.5% were born outside Canada, and 5.1% were of Aboriginal origin.
Among this group of 372 cases of pulmonary TB, 55.9% presented an acid-alcohol fast bacillus (AAFB) in their microscopic sputum examination, 81.5% had a positive sputum culture, and 26.4% had chest X-rays that showed cavities.
The epidemiological investigation into these 353 cases of pulmonary TB resulted in the identification of 1,422 close contacts (median: three close contacts per case; range: 0 to 42 contacts per case). The average age of contacts was 35.7 years. The average number of contacts was: 4.4 for cases whose sputum smear was positive for AAFB, compared to 3.5 contacts for cases whose smear was negative; 4.2 contacts for cases in whom cavities were detected, and 3.2 contacts for cases that did not present cavities. In 51 cases (14.4%), no close contacts were identified.
The contact results appear in Figure 1. Of 1,205 contacts who had a TST1 and for whom results were available, 464 (38.5%) had aTST1 of ≥ 5mm. Of the 510 contacts who underwent a TST2 and for whom results were available, 57 (11.2%) showed tuberculin conversion.
Figure 1. Contact investigation results
Of 1,367 contacts whose TST status was indicated, 1,046 (76.5%) underwent complete screening. Screening results based on case and contact characteristics are provided in Table 1. In univariate analysis, the proportion of contacts that presented a positive screening result was statistically larger when the index case was ≥ 65 years of age, was born outside of Canada, was of Aboriginal origin, or had a chest X-ray that showed cavities. The proportion of positive screening results was also significantly greater among contacts aged ≥ 19. In a logistic regression model, the age of the contact, the place of birth of the index case, and the chest X-ray results of the index case were all independently associated with a positive screening result. Contacts aged ≥ 19 had a 5.7 times greater chance (CI 95%: 4.0 to 8.0) of having a positive result than those aged ≤ 18. Contacts of cases born outside of Canada had a 2.8 times greater chance (CI 95%: 2.0 to 3.9) of having a positive result than contacts of cases born in Canada. Finally, contacts of cases whose chest X-rays showed cavities had a 2.0 greater chance (CI 95%: 1.4 to 2.7) of having a positive result than contacts of cases that did not present cavities.
Since the age and ethnic origin of contacts may be associated with an increased risk of prior exposure to TB, we analyzed possible associations between tuberculin conversion and the same variables in the 518 contacts who had undergone two TSTs (Table 2). Conversion was observed in 12.1% of contacts linked to cases who were sputum AAFB positive vs. 8.9% of contacts linked to cases where no AAFB was present ( p = 0.26), and in 10.9% of contacts linked to cases with positive cultures vs. 12.2% of contacts linked to cases with a negative culture ( p = 0.77). The proportion of contacts in whom conversion took place was exactly the same (11.1%), regardless of whether the case to which they were linked had a chest X-ray that showed cavities. The proportion of contact conversions was larger when the index case was born in Canada (12.0%) than when the index case was born outside Canada (8.7%), although this difference is statistically insignificant. The age of contacts was associated with the occurrence of conversion: 4.0% of contacts between the ages of 0 and 18 years experienced conversion vs. 16.9% of those aged ≥ 19 (p < 0.001).
Table 1. Results for 1,046 close contacts having undergone complete screening, by case and contact characteristics
Number of contacts | Contacts with TST ≥ 5mm n (%) |
Contact conversion n (%) |
Positive screening n (%) |
|
Case characteristics | ||||
Sex | ||||
Male | 390* | 170 (43.6) | 25 (6.4) | 195 (50.0) |
Female | 656* | 294 (44.8) | 32 (4.9) | 326 (49.7) |
Age | ||||
0-18 | 49* | 22 (44.9) | 3 (6.1) | 25 (51.0) |
19-64 | 591* | 285 (48.2) | 28 (4.8) | 313 (53.0) |
≥ 65 | 406* | 157 (38.7) | 26 (6.4) | 183 (45.1)† |
Sputum smear | ||||
Presence of AAFB | 660* | 299 (45.3) | 41 (6.2) | 340 (51.5) |
Absence of AAFB | 386* | 165 (42.7) | 16 (4.2) | 181 (46.9) |
Sputum culture | ||||
Positive | 908* | 410 (45.2) | 51 (5.6) | 461 (50.8) |
Negative | 138* | 54 (39.1) | 6 (4.3) | 60 (43.5) |
Chest X-ray‡ | ||||
Presence of cavities | 354* | 193 (54.5) | 17 (4.8) | 210 (59.3)§ |
Absence of cavities | 670* | 263 (39.3) | 39 (5.8) | 302 (45.1) |
Ethnic origin | ||||
Non-Aboriginal Canadian | 676* | 260 (38.5) | 43 (6.4) | 303 (44.8) |
Immigrant | 325* | 172 (52.9) | 13 (4.0) | 185 (56.9) |
Aboriginal Canadian | 45* | 32 (71.1) | 1 (2.2) | 33 (73.3)|| |
Contact characteristics | ||||
Sex¶ | ||||
Male | 466 | 217 (46.6) | 21 (4.5) | 238 (51.1) |
Female | 542 | 235 (43.4) | 33 (6.0) | 268 (49.4) |
Age** | ||||
0-18 | 286 | 70 (24.4) | 8 (2.8) | 78 (27.3) |
19-64 | 562 | 295 (52.5) | 41 (7.3) | 336 (59.8) |
≥ 65 | 97 | 54 (55.7) | 5 (5.1) | 59 (60.8)†† |
Total | 1,046 | 464 (44,4) | 57 (5.4) | 521 (49.8) |
* represents the number of contacts for whom the index case presents the characteristic †χ = 6.02; p = 0.049 ‡ for 22 contacts, case chest X-ray results are missing §χ = 18.81; p< 0.001 ||χ = 23.26; p< 0.001 ¶ for 38 contacts, sex was not indicated ** for 101 contacts, age was not indicated ††χ = 85.17; p< 0.001
Table 2. Results for 518 close contacts having undergone two TSTs, by case and contact characteristics
Number of contacts | Contact conversion n (%) | |
Case characteristics | ||
Sex | ||
Male | 302* | 32 (10.6) |
Female | 216* | 25 (11.6) |
Age | ||
0-64 | 306* | 31 (10.1) |
≥ 65 | 212* | 26 (12.3) |
Sputum smear | ||
Presence of AAFB | 338* | 41 (12.1) |
Absence of AAFB | 180* | 16 (8.9) |
Sputum culture | ||
Positive | 469* | 51 (10.9) |
Negative | 49* | 6 (12.2) |
Chest X-ray† | ||
Presence of cavities | 153* | 17 (11.1) |
Absence of cavities | 351* | 39 (11.1) |
Ethnic origin | ||
Non-Aboriginal Canadian | 357* | 43 (12.0) |
Immigrant/Aboriginal | 161* | 14 (8.7) |
Contact characteristics | ||
Sex‡ | ||
Male | 221 | 21 (9.5) |
Female | 273 | 33 (12.1) |
Age§|| | ||
0-5 | 86 | 2 (2.3) |
6-18 | 111 | 6 (5.4) |
19-34 | 77 | 12 (15.6) |
35-64 | 159 | 29 (18.2) |
≥ 65 | 36 | 5 (13.9) |
Total | 518 | 57 (5.4) |
* represents the number of contacts for whom the index case presents the characteristic † for 14 contacts, case chest X-ray results are missing ‡ for 24 contacts, sex was not indicated § for 49 contacts, age was not indicated ||χ = 19.7; p = 0.001
Discussion
Management of TB close contacts is an issue that has recently drawn the attention of health professionals engaged in the fight against tuberculosis(3,8). Although the importance of contact tracing has been acknowledged for some time, the methods employed for this purpose tend to vary from program to program and, deplorably, there are no standardized definitions of contact duration and proximity to determine the scope of screening activities.
In our study, 44.4% of close contacts who had undergone complete screening had a TST1 of ≥ 5 mm compared to 34% in the study by Reichler et al.(9) and 44.5% in the study by Behr et al.(10) Varying definitions of what constitutes a close contact may explain these differences. With our definition, only persons who live under the same roof as the index case are identified as close contacts in most instances(6). The definition used by Reichler was broader : members, visitors or employees of the index case's household, as well as friends and family were defined as close contacts(9). In the Behr study, only persons who had spent at least 40 to 100 hours with the index case during the 3 months preceding diagnosis or during the infectious period were considered to be close contacts(10).
The characteristics of close contacts (age, sex, place of birth, BCG), as well as those of index cases (sputum AAFB result, chest X-ray showing cavities, etc.), and the environment (ventilation, duration of exposure) can influence screening results(11-13) .Inour study, 12.0% of close contacts were ≥ 65 years of age, compared to 7% in the Reichler study(9). Although data on contacts' place of birth are not available, 43.5% of our cases were born outside of Canada, whereas in the United States, 31% of cases were born elsewhere. A larger proportion of contacts born in regions where tuberculosis is endemic may have been infected before they were ever exposed to the index case. On the other hand, the proportion of cases who were sputum smear positive for AAFB was smaller in our study (55.9% vs. 58.6%) than in the Reichler study, as was the proportion of cases whose chest X-rays showed cavities (26.4% vs. 30%)(9).
Among close contacts who had undergone complete screening, 521 (49.8%) had positive screening results. It is difficult to separate positive results that represent an infection transmitted by the index case from those attributed to an M. tuberculosis infection, an atypical mycobacterium acquired earlier, or a BCG reaction. The addition of initial tuberculin skin tests of ≥ 5mm (24.4%) and conversions (2.8%) in those aged 0 to 18 years (27.2 %) likely provides a more accurate estimate of the proportion of contacts who present a recent infection than the 395 (59.9%) contacts aged ≥ 19 who had a positive screening result. Greenaway et al. has estimated that in populations similar to immigrant communities living in Canada (i.e. vaccinated in childhood with BCG, frequently exposed to M. tuberculosis or other mycobacteria) the probability that a TST of ≥ 10 mm represents a recent infection was only 24% among individuals exposed for 100 hours to a pulmonary TB case with a positive smear(14). The probability of recent infection is far greater in a population rarely exposed to M. tuberculosis or other mycobacteria and not vaccinated with BCG, as is the case with most adults born and residing in Canada, although the more generalized use of BCG in Quebec up to 1976 may exert a downward influence on this probability.
Our study presents certain limits. It is based on a retrospective analysis carried out by different stakeholders in 18 regions in which practices are not necessarily uniform. Information on case contacts was very limited: the place of birth was not indicated in their records, information on BCG was lacking, quantification of exposure was far from precise. Moreover, we limited our study to the assessment of close contacts.
Conclusion
A large number of TB cases and their close contacts who live in industrialized countries were born in other countries where TB is highly prevalent. In Quebec, the BCG vaccine was administered to school-age children until the mid-1970s. Given these factors, accurately interpreting TST results in TB case contacts can be very challenging. Still, it is important that we continue to evaluate the results of contact investigations and that we attempt to identify factors that may facilitate accurate interpretation of TST results.
Acknowledgements
The authors wish to thank all the PHB professionals in Quebec who participated in our study.
References
-
Menzies D, Tannenbaum TN, FitzGerald JM. Tuberculosis: 10. Prevention. CMAJ 1999;161(6):717-24.
-
Horsburg CR, Feldman S, Ridzon R. Practice guidelines for the treatment of tuberculosis. Clin Infect Dis 2000;31:633-9.
-
Centers for Disease Control and Prevention. Guidelines for the investigation of contacts of persons with infectious tuberculosis; recommendations from the National Controllers Association and CDC. MMWR 2005;54(No. RR-15): 1-47.
-
Centers for Disease Control and Prevention. Progressing toward tuberculosis elimination in low-incidence areas of the US : recommendations of the Advisory Council for elimination of tuberculosis. MMWR 2002;51(No. RR-5):1-18.
-
Etkind SC. Contact tracing in tuberculosis. In: Reichman L, Hershfield E, eds. Tuberculosis: A comprehensive approach. New York, NY: Marcel Dekker Inc, 1993:275-89.
-
Direction générale de la santé publique, Ministère de la Santé et des Services sociaux. Protocole d'intervention - La tuberculose. Québec: La Direction. 1998:1-134.
-
Daniel WW. Biostatistics - A foundation for analysis in the health sciences. 6th edition. New York: John Wiley & Sons Inc., 1995:180-1.
-
Reichler MR, Etkind S, Taylor Z et al. Tuberculosis contact investigations. Int J Tuberc Lung Dis 2003;7(12): S325-S7.
-
Reichler MR, Reves R, Bur S et al. Evaluation of investigations conducted to detect and prevent transmission of tuberculosis. JAMA 2002; 287:991-5.
-
Behr MA, Hopewell PC, Paz EA et al. Predictive value of contact investigation for identifying recent transmission of M. tuberculosis. Am J Respir Crit Care Med 1998;158(2):465-9.
-
Bailey WC, Gerald LB, Kimerling ME et al. Predictive model to identify positive tuberculosis skin test results during contact investigations. JAMA 2002;287:996-1002.
-
Menzies RI, Vissandjee B. Effect of Bacille Calmette-Guérin vaccination on tuberculin reactivity. Am Rev Resp Dis 1992;145:621-5.
-
Sepulveda R, Ferrer X, Latrach C et al. The influence of Calmette-Guérin Bacillus immunization on the booster effect of tuberculin testing in healthy young adults. Am Rev Resp Dis 1990;142:24-8.
-
Greenaway C, Palayew M, Menzies D. Yield of casual contact investigation by the hour. Int J Tuberc Lung Dis 2003;7(12):S479-S85.
Source : Paul Rivest, MD, Direction de santé publique de Montréal, Département de médecine social et préventive, Université de Montréal, Institut national de santé publique du Québec; Serge Déry, MD, Nunavik Public Health Directorate; Monique Douville-Fradet, MD, Institut national de santé publique du Québec; Terry-Ann Tannenbaum, MD, Direction de santé publique de Montréal, Institut national de santé publique du Québec; Paul Brassard, MD, Direction de santé publique de Montréal, Département de médecine social et préventive, Université de Montréal, Division of Clinical Epidemiology, McGill University Health Centre, Montreal, Quebec.
Page details
- Date modified: