 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
SUPPLEMENT ARTICLE |
 |
ABSTRACT |
|---|
 TOP ABSTRACT EXECUTIVE SUMMARYINTRODUCTIONSCIENTIFIC RATIONALE FOR...Diagnosis of LTBITreatment of LTBIRECOMMENDATIONSPRIORITIES FOR FUTURE RESEARCHREFERENCES |
|---|
1 risk factor is present. The use of administrative or mandated tuberculin skin tests for entry to day care, school, or summer camp is strongly discouraged. Treatment regimens, suggestions to improve adherence, and methods to monitor toxicities are summarized. Children and adolescents with LTBI represent the future reservoir for cases of TB. Thus, detecting and treating LTBI in children and adolescents will contribute to the elimination of TB in the United States.
Key Words: latent tuberculosis infection • tuberculin skin test • children • adolescents • pediatrics • tuberculosis
Abbreviations: TB, tuberculosis • LTBI, latent tuberculosis infection • USPHS, United States Public Health Service • CDC, Centers for Disease Control and Prevention • TST, tuberculin skin test • HIV, human immunodeficiency virus • AIDS, acquired immunodeficiency syndrome • AAP, American Academy of Pediatrics • TU, tuberculin units • PPD, purified protein derivative • MPT, multiple-puncture test • INH, isoniazid • DOT, directly observed therapy • MDR, multidrug-resistant • BCG, bacillus Calmette-Guérin • TNF-
, tumor necrosis factor • CT, computed tomography • DTH, delayed-type hypersensitivity • NTM, nontuberculous mycobacteria • ESAT-6, early secreted antigenic target 6 kDa • QFT, QuantiFERON-TB • IFN-
, interferon • ELISPOT, enzyme-linked immunospot • OR, odds ratio • CI95, 95% confidence interval
|
EXECUTIVE SUMMARY |
|---|
|
TOPABSTRACTEXECUTIVE SUMMARYINTRODUCTIONSCIENTIFIC RATIONALE FOR...Diagnosis of LTBITreatment of LTBIRECOMMENDATIONSPRIORITIES FOR FUTURE RESEARCHREFERENCES |
|---|
The recommendations in this article have been developed by the Pediatric Tuberculosis Collaborative Group to address the need for specific recommendations for children and adolescents for health care providers serving pediatric populations. The age used to define pediatric TB disease and LTBI varies; for example, the CDC defines pediatric TB as occurring in persons <15 years of age. However, this article addresses the needs of children and adolescents from birth to 18 years of age. In this article, LTBI is defined as a child or adolescent with a positive tuberculin skin test (TST) who has no evidence of TB disease. A glossary of terms used in this article is presented in Table 1.
|
This consensus statement was developed by experts in the care of children and adolescents with TB disease and LTBI. This panel was convened by the co-chairs in consultation with the CDC, and this process was endorsed by the American Academy of Pediatrics (AAP). The multidisciplinary panel included health care professionals from health departments, the CDC, the National Tuberculosis Centers, and academic institutions. Relevant studies and unpublished data sets compiled by the participants were summarized. Evidence-based recommendations were developed to update and supplement the recommendations by the 2003 Report of the Committee on Infectious Diseases.3
The data presented in this article support a paradigm shift and a change in guidelines for tuberculin skin testing. Children and adolescents should be screened for risk factors for TB and LTBI and tested with a TST only if 1 risk factors are present. "Routine" or "mandated" LTBI testing policies for pediatric patients without risk factors are strongly discouraged (eg, entry into day care, school, summer camp, or college).
Targeted Tuberculin Skin Testing
Targeted tuberculin skin testing is intended to identify children and adolescents at risk for LTBI who would benefit from treatment to prevent the progression to TB disease. Targeted testing discourages tuberculin skin testing of low-risk populations and focuses on testing children with risk factors. Several recent studies have delineated risk factors for LTBI in children (Table 2). These studies were conducted in different pediatric populations but found very similar risk factors including foreign birth, foreign travel, and a close association with persons having TB disease or LTBI. Based on these factors, a risk-factor questionnaire was developed by the consensus panel to facilitate screening by pediatric health care providers in a variety of clinical settings (Table 3). The use of the screening questionnaire and the precise questions asked will vary from population to population depending on local epidemiology.
|
|
Administration, Reading, and Interpretation of TSTs
The only recommended TST method is the intradermal injection of 5 tuberculin units (TU) of purified protein derivative (PPD) from M tuberculosis administered by the Mantoux technique. Multiple-puncture tests (MPTs) or the Tine test are not recommended for use. TSTs should be read 48 to 72 hours after placement by a trained health care provider. Results should be recorded as millimeters of induration (eg, 00 mm, 12 mm, etc).
The results of the TST are interpreted in the context of the patient’s risk of M tuberculosis infection, ie, exposure to TB disease or risk of progression to TB disease. Three cutoff levels (5, 10, or 15 mm) are used to improve the sensitivity and specificity of the TST (Table 4).
|
|
|
|
|
|
|
INTRODUCTION |
|---|
|
TOPABSTRACTEXECUTIVE SUMMARYINTRODUCTIONSCIENTIFIC RATIONALE FOR...Diagnosis of LTBITreatment of LTBIRECOMMENDATIONSPRIORITIES FOR FUTURE RESEARCHREFERENCES |
|---|
|
|
The third level of TB control is the identification and treatment of individuals with LTBI. This effort, although conducted in part by health departments, is more likely to be conducted by other pediatric health care providers such as pediatricians, family practitioners, and nurse practitioners. Strategies to accomplish this third level of control include a variety of targeted tuberculin skin-testing programs including screening high-risk children and adolescents for LTBI risk factors during primary care visits or in school through school-based screening programs.
Increasing Importance of Targeted Tuberculin Skin Testing in the United States
As the rate of TB disease has declined in the United States, accurate identification and completed treatment of persons with LTBI are increasingly critical components of TB-elimination strategies.13 Previous recommendations prioritized the identification of high-risk persons, including children and adolescents, at increased risk of progression to TB disease.14 More recent studies have further delineated risk factors for LTBI in children and adolescents and allow further refinements for targeted tuberculin skin testing in general pediatric populations. Thus, the recommendations in this article will focus exclusively on children and adolescents both to identify those at the highest risk of progression to TB disease and those most likely to have LTBI who would benefit from treatment.
|
SCIENTIFIC RATIONALE FOR RECOMMENDATIONS |
|---|
|
TOPABSTRACTEXECUTIVE SUMMARYINTRODUCTIONSCIENTIFIC RATIONALE FOR...Diagnosis of LTBITreatment of LTBIRECOMMENDATIONSPRIORITIES FOR FUTURE RESEARCHREFERENCES |
|---|
In addition to testing the groups of children listed above, this article presents a paradigm shift in the recommendations for pediatric health care providers to promote the targeted tuberculin skin testing of children and adolescents. Targeted skin testing replaces the concept of a routine TST placed in primary health care settings. "Administrative" or mandated TSTs for entry to day care, school, summer camp, or college are strongly discouraged in the absence of risk factors. Instead, children and adolescents should be screened for risk factors for TB disease and LTBI by using a risk-assessment questionnaire as described below and tested with a TST only if 1 risk factors are present.
Contact and Source-Case Investigations
Pediatric patients who are contacts of a patient with known or suspected TB disease must be evaluated promptly for TB disease or LTBI and undergo immediate tuberculin skin testing as part of the assessment process, which would include testing the contacts of an infectious adult or adolescent (contact investigation) as well as testing the contacts of a child with TB disease (source-case investigation).
Studies continue to emphasize the value of contact investigations to identify children with TB disease or LTBI.12,15,16 Marks et al15 compared the outcomes of contact investigations with and without home visits that were conducted for 1080 infectious adult TB patients. Home visits identified 6.7 close contacts, whereas only 4.7 contacts were identified when home visits were not conducted. The additional contacts identified were likely to be children <6 years of age. In this study, 21% (132 of 618) of children <6 years of age had a positive TST (5 mm), and 5% (35 of 705) of such children had evidence of TB disease. Thus, identifying and evaluating young children during contact investigations of infectious adults are critical components of TB-control efforts.
Similarly, Lobato et al16 assessed the yield of source-case investigations conducted for children <5 years of age with active TB for detecting cases of undiagnosed TB and LTBI in children and adolescents in California.16 In all, 111 source-case investigations were performed, and 31% (254 of 815) of persons with whom the index cases had frequent exposures were <15 years of age. In all, 6% (7 of 141) of children <5 years of age were found to have undiagnosed TB disease. The rates of LTBI were 24% (34 of 141) and 32% (36 of 113) among children <5 and 5 to 14 years of age, respectively. This study confirms the importance of assessing other children for TB and LTBI during a source-case investigation.
Screening Children and Adolescents for Risk Factors for LTBI Using a Questionnaire
Several recent studies have assessed risk factors for LTBI in pediatric populations and provided additional justification for targeted tuberculin skin testing. Rather than the use of a TST as a screening tool, these studies promoted the use of a questionnaire as a screening tool. Although these studies assessed different populations, there were marked similarities in their findings (Table 2). Lobato and Hopewell17 conducted a case-control study in 953 children <6 years of age who had a TST read at health clinics in California. Risk factors for a positive (10-mm) TST included foreign travel within the previous 12 months (defined as a trip of >1 week to a country with a high prevalence of TB disease) or a household visitor from such a country.
In a similar study, Saiman et al18 performed a matched case-control study among children 1 to 5 years of age in northern Manhattan and Harlem (New York) whose TSTs were placed by their health care provider as part of routine primary care. Contact with an adult with TB disease, foreign birth, foreign travel, or a relative with a positive TST were identified as risk factors for LTBI. Besser et al19 performed a similar analysis of risk factors for LTBI among children <6 years of age in San Diego, California. In this population, bacillus Calmette-Guérin (BCG) immunization, a TST within 12 months, and a relative with a positive TST were risk factors for a positive TST (10 mm). Froehlich et al20 performed a study to determine if a risk-assessment questionnaire could predict a positive TST in children in northern California and found that foreign birth, BCG immunization, living outside the United States, Asian or Hispanic ethnicity, or contact with a household member with TB disease or LTBI were independent predictors of LTBI.
Finally, Ozuah et al21 sought to determine the sensitivity, specificity, and predictive validity of a New York City Department of Health questionnaire22 in 2920 children. In all, 14% (413 of 2920) of children had at least 1 risk factor (Table 2), and of these, 6% (23 of 413) had a positive TST (10 mm). In contrast, 0.16% (4 of 2507) of children without risk factors identified had a positive TST. The sensitivity of the questionnaire was 85% and the specificity was 86%; the negative predictive value was 99.9%, but the positive predictive value was only 5%. Notably, the questionnaire failed to detect risk factors in 4 children with positive TSTs, of whom 3 were >11 years of age. This suggested that the questionnaire may not have addressed all risk factors in adolescents such as exposure to individuals outside of the immediate household.
Delineation of High-Risk Adults
Past recommendations have suggested that exposure to adults at high risk of TB disease places a child at increased risk for LTBI and TB disease. However, few studies have characterized the magnitude of risk. The studies detailed above attempted to clarify which populations of adults were "high risk."
In the population studied by Saiman et al,18 contact with adults with illicit drug use or HIV/AIDS or adults who were homeless or incarcerated were not risk factors for LTBI in children, nor were foreign-born parents, visitors from abroad, or foreign travel by parents. In contrast, Lobato et al17 found that a visitor from abroad was a risk factor for LTBI in children in California. Ozuah et al21 found that contact with an adult with HIV or illicit drug use or who was homeless or incarcerated was a risk factor for LTBI in children in the Bronx. Thus, the definition of a high-risk adult varied from population to population.
International Adoption of Children
For over a decade, the unique medical needs of internationally adopted children have been recognized, because these children are at risk for infectious diseases acquired in their countries of origin.23 Several investigators have evaluated international adoptees for LTBI and TB disease. Saiman et al24 performed TSTs on 404 internationally adopted children; 19% (75 of 404) had positive TSTs (TST 10 mm) and normal chest radiographs. In contrast, previous rates of LTBI among international adoptees ranged from 0.6% to 5%.23,25–29
The marked differences in the prevalence of LTBI noted in different studies may reflect changes in the epidemiology of internationally adopted children. As the primary countries of origin have changed, the prevalence of prior BCG immunization and possible exposure to TB disease (eg, in orphanages) have both increased. In addition, during the 1990s, the rates of TB disease rose worldwide. In earlier studies, most international adoptees were born in Korea and Romania,25,30 whereas the children evaluated by Saiman et al24 were primarily born in China and Russia. Among 873 Korean adoptees, none had received BCG immunization, and 90% had lived with foster families.25 In contrast, 60% of the children adopted from 1997 to 1998 had received BCG immunization, and 88% had lived in orphanages.24
TB disease is far less common than LTBI among internationally adopted children, but a recent report described extensive transmission of TB disease to close contacts of a child adopted from the Marshall Islands.31 Evaluation with a TST on US arrival and treatment for LTBI may have prevented the development of TB disease in this child who was clinically well at the time of adoption.
In summary, several studies have identified risk factors for LTBI in children, such as contact with an adult with active TB, foreign birth (including internationally adopted children), travel to a country with a high prevalence of TB, and a household member with LTBI. Additional risk factors such as contact with high-risk adults or household visitors from a country with a high prevalence of TB disease may be risk factors in some populations. However, few of these studies addressed risk factors for adolescents. Risk factors should be assessed on an individual basis to determine the need for placement of a TST.
School-Based Screening for LTBI
Routine placement of TSTs at school entry has been used as an opportunity to screen children and adolescents for TB disease and LTBI. A recent study of universal school-based screening throughout the United States has demonstrated low rates of TB disease (<0.02%) and LTBI (<2%).32 However, the prevalence of TST positivity among foreign-born students was 6 to 24 times higher than among US-born students. Thus, it has been recommended that only foreign-born students from countries with high case rates of TB be targeted for assessment for LTBI by tuberculin skin testing.33
As additional support of a targeted approach for school-based screening for LTBI, Mohle-Boetani et al34 evaluated the cost-effectiveness of screening strategies to prevent TB disease. These authors compared a screen-all strategy (ie, testing all kindergarten and high-school entrants) with targeted screening (ie, testing only high-risk students in these age groups, defined as birth in a country with a high prevalence of TB disease). Targeted screening was more cost-effective because it was estimated to prevent 85 cases of TB disease per 1000 persons tested, compared with the screen-all strategy, which only prevented 15 cases per 1000 persons tested. In this analysis, the screen-all strategy would be cost-effective only if the prevalence of LTBI was 20%.
Additional studies have suggested that school-based targeted testing should be focused primarily on foreign-born adolescents. Scholten et al35 reported the prevalence and risk factors associated with positive TSTs among school children in New York City, New York, from 1991 to 1993. Overall, 2.1% (6326 of 298506) of new school entrants had a positive TST (10 mm). However, 0.5% (931 of 199 728) of US-born children had a positive TST compared with 9% (3794 of 41 346) of foreign-born students. Older children had the highest prevalence of LTBI; 11% (1548 of 14067) of adolescents in grades 7 to 12 had a positive TST. Similar findings were observed in Los Angeles County, California, among students in grades kindergarten to 12; 1.4% of US-born students versus 18.3% of foreign-born students had a positive TST.36
Gounder et al37 expanded these previous observations and described the experience in New York City from 1991 to 1998 (Table 11; Fig 2). In 1990, a TST was mandated for all new school entrants, but in 1996 the health code was amended, and a TST was mandated only for new entrants to secondary schools. In this study, 788283 children and adolescents were evaluated for LTBI. The proportion of students with positive TSTs varied by age, race, and birth place; US-born Asian students and foreign-born students were most likely to have a positive TST. Among US-born students, 0.5% (2553 of 515005) had a positive TST, whereas among foreign-born students, 9.3% (10413 of 112081) had a positive TST. Older age, defined as 12 to 16 years of age, was associated with an increased prevalence of positive TSTs in both US- and foreign-born students (Table 11). Unfortunately, changes in the health code did not substantially alter tuberculin skin-testing practices. Moreover, the majority of children tested by this semitargeted strategy were at low risk for LTBI. The authors concluded that improving targeted testing and educating and garnering the support of pediatric health care providers and school personnel were needed to alter tuberculin skin-testing practices.37
|
|
Moser presented additional experience with targeted testing of adolescents in San Diego (K. Moser, MD, MPH, written communication, 2003). To facilitate such screening, a school coordinator was hired in 2001, and several models were developed in high schools and middle schools based on their populations and capacities. One district tested foreign-born high school students and had a 32% (154 of 489) TST positivity rate. One district tested middle and high school students in English-learners’ classes, and another tested high school migrant-education–supported students, yielding a 25% (16 of 64) and 43% (23 of 54) TST positivity rate, respectively. A 3-question risk-assessment questionnaire was used in 2 high schools: (1) Were you born in or have you lived in Asia, Africa, Eastern Europe, and/or Latin America (including Mexico)? (2) Have you visited Asia, Africa, Eastern Europe, and/or Latin America (including Mexico) for >2 weeks? (3) Have you spent time close to someone sick with TB? Among students who answered "yes" to any of the 3 questions, the TST positivity rates were 19% in 1 school and 32% in the other. Combined data from 1073 students tested through targeted efforts in San Diego high schools and middle schools in the 2001 and 2002 academic years demonstrated that foreign-born students, US-born Hispanics, and US-born non-Hispanics had TST positivity rates of 35% (237 of 684), 24% (82 of 335), and 5% (1 of 21), respectively.
Hsu et al40 examined the correlation with self-reported risk factors and recent TSTs to determine if at-risk adolescents were being screened for LTBI in Boston public schools. Although the majority of 9th-grade students surveyed (75% [436 of 578]) did report at least 1 risk factor, only 40% (231 of 578) had been tested for LTBI. Notably, 81% reported that they had an annual checkup. The authors concluded that screening and testing for LTBI was not occurring appropriately among adolescents in Boston attending public schools and that school-based programs were needed.
Thus, data suggest that, in some communities, middle school and high school may be ideal settings to screen and test adolescents for LTBI because of the higher prevalence of infection. To be effective, a risk-factor questionnaire should consider local TB epidemiology. The increased risk of developing reactivation and infectious TB among adolescents also makes school-based screening, targeted testing, and treatment desirable.41
Associate Investigations as a Targeted Tuberculin Skin-Testing Strategy
Associate investigations traditionally are performed by health departments whereby the close contacts of children with LTBI (ie, their associates) are tested to detect undiagnosed cases of infectious TB. However, associate investigations may detect greater numbers of associates with LTBI and thus may be considered a form of targeted testing for LTBI. The AAP currently recommends that the associates of children with a positive TST undergo tuberculin skin testing.3 In general, most health departments perform associate investigations for children <4 years of age with LTBI because young children are likely to have been infected recently and have a limited number of associates, which theoretically makes the likelihood of finding an active case of TB among their associates high.
The yield of associate investigations has been evaluated in several studies. Sullam et al42 conducted associate investigations for 297 children with LTBI <8 years of age. The associates were largely foreign-born, primarily Asian, and resided in San Francisco, California. Associate investigations detected undiagnosed cases of TB disease in 0.36% (3 of 831) of associates, but more striking is that 40% (330 of 831) of associates had positive TSTs and were considered candidates for LTBI treatment.
Soren et al43 studied 659 associates of 187 children and adolescents 
21 years of age with LTBI in northern Manhattan. This study population was largely Hispanic immigrants, primarily from the Dominican Republic. No cases of TB disease were detected among the associates, but 32% (210 of 659) had positive TSTs (10 mm).
Driver et al44 examined the yield of associate investigations conducted in New York City by the Department of Health. In all, 980 associates of 207 children 3 years of age were evaluated, and 26% (255 of 980) had a positive TST. However, the yield was higher among household associates: 30% (198 of 668) had a positive TST, compared with 18% (57 of 312) of nonhousehold associates (P < .01). This associate-testing effort detected TB disease in 0.3% (3 of 980) of those assessed.
The Health Department in San Diego performed associate investigations among 234 children 5 years of age reported from January 2001 to March 2002 (K. Moser, MD, MPH, written communication, 2003). In all, 910 associates of these primarily Hispanic children were identified, and 78% (713 of 910) were evaluated. No cases of TB disease were detected, but 41% (292 of 713) of associates had a positive TST.
The Tarrant County (Texas) Health Department conducted targeted associate investigations from January 1999 to December 2001.45 Associate investigations in Tarrant County are targeted to associates of non–BCG-immunized children <6 years of age because such children are hypothesized to be more likely to have a positive TST from community transmission of M tuberculosis. Overall, 16% (38 of 232) of children with LTBI met these criteria, and 259 of their associates were tested (median: 7.8 associates per investigation). Undiagnosed, culture-confirmed TB disease was detected in 3% (n = 8) of associates, all of whom were foreign-born, yielding a rate of 21 new cases of TB disease per 100 investigations performed. In addition, 43% (110 of 259) of associates had LTBI, of whom 72% (n = 79) were foreign-born.
In summary, among high-risk populations (eg, foreign-born persons), associate investigations can identify associates with a high prevalence of LTBI. Some health districts have further refined associate investigations by targeting efforts to non–BCG-immunized children. These strategies also may enhance efforts to detect new cases of TB disease. The cost-effectiveness of associate investigations compared with other methods of targeted testing has not been studied.
Underlying Medical Conditions and Concomitant Medications
Several medical conditions and concomitant medications increase the risk of progression to TB disease in persons infected with M tuberculosis. Thus, children and adolescents with such conditions or receiving such medications are candidates for LTBI screening. These medical conditions include HIV infection, diabetes, organ transplantation, chronic renal failure, and malignancies. The use of high-dose steroids, chemotherapy,8–11 or agents with activity against tumor necrosis factor (TNF-) (eg, infliximab [Remicade]) has also been associated with progression to TB disease. Although the published reports linking TNF- antagonists with active TB have been in adults,46 these agents are being increasingly used for the treatment of joint, skin, and gastrointestinal diseases in pediatric patients. The manufacturers of these agents recommend assessing patients for LTBI before use. A review of the risks associated with these agents, proposed mechanism of action, and clinical management has been published.47
There are few published reports evaluating the risk of progression to TB disease in children and adolescents with LTBI who are receiving inhaled corticosteroids. Bahceciler et al48 studied the effect of inhaled budesonide in 32 asthmatic children with positive TSTs (10 mm) and normal chest radiographs. The children were treated for a mean of 10 months with budesonide (mean cumulative dose: 275 mg) but did not receive INH. All 32 children had high-resolution computed tomography (CT) of the chest, and 22% (7 of 32) were thought to have detectable mediastinal lymph nodes that were unchanged on high-resolution CTs performed 9 months later. The authors concluded that inhaled steroids did not effect the progression to TB disease in patients untreated for LTBI. However, this report described a limited number of children followed for a relatively short period of time. Thus, larger studies with longer follow-up are needed.
Thus, children receiving medical treatments or recently diagnosed with conditions known to predispose adults to progression to TB disease should have a TST and begin treatment immediately if LTBI is diagnosed.
|
Diagnosis of LTBI |
|---|
|
TOPABSTRACTEXECUTIVE SUMMARYINTRODUCTIONSCIENTIFIC RATIONALE FOR...Diagnosis of LTBITreatment of LTBIRECOMMENDATIONSPRIORITIES FOR FUTURE RESEARCHREFERENCES |
|---|
Mantoux Skin Test
History of PPD Preparations.
Koch prepared the first tuberculin from concentrated filtrates of heat-sterilized tubercule bacilli, but the heterogeneity of the filtrate caused unreliable and nonspecific reactions.49 Thus, Seibert developed PPD tuberculin in 1934 by using a protein precipitation of culture filtrates that reduced the amount of polysaccharides and nucleic acids in the preparation.49 In 1939, PPD-S was prepared and continues to serve as the international reference to ensure equal biological potency among various lots of PPD.2,50
Administration of the TST by the Mantoux Method.
The recommended TST is administration of the standardized PPD by the Mantoux method in which 0.1 mL of 5 TU of PPD tuberculin is injected intradermally to form a wheal 
6 to 10 mm in diameter.51,52 Other concentrations (1 or 250 TU per dose) are not well standardized, less sensitive and specific, and not recommended.53 Two tuberculin PPD preparations, Aplisol and Tubersol, are available in the United States.2
DTH Reaction.
DTH reaction to a TST manifests as an indurated area at the site of the intradermal injection and usually begins within 5 to 6 hours of administration of the PPD as previously sensitized lymphocytes, monocytes, and macrophages infiltrate the site. The DTH reaches a maximum size by 48 to 72 hours and subsides over the subsequent few days.51,54 Proper reading of the TST includes measuring and recording the diameter of the area of induration in millimeters 48 to 72 hours after TST placement.51 An immediate wheal-and-flare reaction may occur but usually disappears by 24 hours and should not be interpreted as a positive reaction to a TST.49 Rarely, the immediate reaction may be severe, and experts suggest that it may be prudent not to retest such individuals.52 Although the area is frequently erythematous at 48 to 72 hours, only the area of induration should be measured. A negative TST should be recorded in millimeters (eg, 00 mm) and not as "negative." TSTs read after 72 hours of placement can underestimate the size of the initial DTH response, and if the TST is <10 mm, it should be repeated immediately. However, if a TST is read after 72 hours and is 10 mm, it can be considered positive if risk factors for LTBI are present. Duboczy and Brown55 followed TST reactions for 7 days in adults with TB disease and found that 4.5% (14 of 239) of those with a TST >5 mm at 48 hours had no induration when read at 5 days. Thus, a TST must be read within 72 hours after placement to accurately determine the diameter of the area of induration.
There are several Web sites and educational materials that describe proper administration and reading of TSTs, including ones from the CDC Division of Tuberculosis Elimination (www.cdc.gov/nchstp/tb/pubs/slidesets/core/Chapter4/test8.htm and https: //www2.cdc.gov/nchstp_od/piweb/tborderform.asp) and the New Jersey Medical School National Tuberculosis Center (www.umdnj.edu/ntbcweb/pr_frame.html).
MPTs
MPTs (eg, Tine, Aplitest, Mono-Vacc test, and the Heaf test) introduce tuberculin antigen into the skin through prongs coated with dried tuberculin or puncture the skin through a liquid film of tuberculin. There are several limitations associated with MPTs including: (1) the amount of antigen introduced is not precise, and reaction sizes are not standardized51; (2) all potentially positive reactions must be followed by a Mantoux test, which increases the cost and complexity of follow-up and prolongs the time until diagnosis and treatment; (3) MPTs may increase the potential for boosting; (4) MPTs have greater variability of sensitivity and specificity than the Mantoux method; and (5) the practice of allowing parents to interpret MPTs in non-health care settings further diminishes the accuracy of the test.56
Sensitivity and Specificity of TSTs
Unfortunately, there is no "gold standard" to diagnose LTBI. Thus, the sensitivity and specificity of the TST is difficult to calculate. The estimated sensitivity of currently available TSTs is based on the use of these tests in patients with TB disease and ranges from 80% to 96%.51 Approximately 10% of immunocompetent children with TB disease have a negative TST.56 False-negative and false-positive TSTS may be caused by several factors (Table 12).
|
Live, Attenuated Vaccines.
Live, attenuated vaccines such as measles, mumps, rubella, varicella,61 oral polio,62 BCG, and oral typhoid (TY21a) may temporarily suppress the DTH response to a TST.2 Kupers et al63 found a 50% decrease in the millimeters of induration in 13 of 17 TST-positive children 1 to 4 weeks after mumps immunization. Similarly, Berkovich et al64 noted a decrease in millimeters of induration in 22% (4 of 18) of children with TB disease after mumps immunization. In another study of 24 children with TB disease conducted by Berkovich et al65 to assess the impact of rubella immunization, 56% (10 of 18) of rubella-immunized children and 33% (2 of 6) of unimmunized children had a reduction in the size of their TST. A decrease in the size of a TST has been described 4 to 6 weeks after polio vaccine62 and 1 month after smallpox vaccine.66
Brickman et al67 sought to examine the impact of live viral vaccines administered at the same time as a TST. These authors administered measles, mumps, and/or rubella vaccines with TSTs to 100 children with previously positive TSTs. A control group consisted of 29 unimmunized children with previously positive TSTs. Overall, 3% (3 of 100) of immunized children and 3.6% (1 of 29) of unimmunized children had negative TSTs, supporting the recommendation that live vaccines and TSTs can be administrated at the same time. If the TST is indicated after a live, attenuated vaccine, it will likely be most accurate if 6 weeks have passed since vaccine administration.
Use of Corticosteroids.
Corticosteroids may affect both the size of a TST and the progression of LTBI to TB disease. In adults, 15 mg of daily prednisone may cause suppression of previously positive TSTs, but the exact risk is unknown.2 Bovornkitti et al68 placed serial TSTs on adults with TB disease (n = 58) or adults with positive TSTs (5 mm) who had other illnesses requiring steroid treatment (40 mg/day of prednisone). The vast majority (97% [68 of 70]) reverted their TSTs to negative (00 mm) a mean of 14 days after starting steroids (treatment duration: 1–4 weeks). These adults reconverted to a positive TST a mean of 6 days after cessation of steroid treatment. In contrast, MacGregor et al69 found no evidence of TST suppression in 12 adults with inflammatory diseases treated with alternate-day prednisone (average: 62 mg/day). Schatz et al70 sought to examine the prevalence of positive TSTs among 132 patients with asthma (range: 9–76 years of age; mean: 47 years of age) receiving long-term steroids (mean duration of treatment: 4.7 years). The investigators placed TSTs on these study subjects and 28% (37 of 132) self-reported positive TSTs (10 mm). Those with negative TSTs received a significantly higher mean daily dose of corticosteroids than those with positive TSTs: 18 vs 11.6 mg/day, respectively (P < .001). However, the dose, dosing frequency, and length of treatment with corticosteroids that confer risk for a false-negative TST have not been defined for children and adolescents.
Anergy Testing.
"Control" skin-test antigens such as Candida, mumps vaccine, diphtheria, or tetanus toxoid have been used to assess a patient’s ability to mount a DTH response. This strategy was used in an attempt to improve the detection of a false-negative TST reaction, particularly among HIV-infected individuals with low CD4 lymphocyte counts. However, the use of control skin-test antigens has several limitations and is not recommended by the CDC as routine practice71: (1) the antigens administered and the reproducibility of the DTH have not been standardized72; (2) the diagnosis of anergy has not been associated with a high risk of developing TB disease; and (3) no demonstrable benefit from empiric INH therapy to prevent TB disease has been noted for anergic HIV-infected persons.73
Factors Associated With False-Positive TSTs
Previous BCG Immunization.
Children born in countries with high case rates of TB disease are likely to have received BCG immunization in infancy. The World Health Organization estimates that 79% of the world’s population has received a BCG vaccine. Twenty-two countries account for 80% of the world’s TB cases and include India, China, Indonesia, Bangladesh, Nigeria, Pakistan, South Africa, the Philippines, Russia, Ethiopia, Kenya, Democratic Republic of the Congo, Vietnam, United Republic of Tanzania, Brazil, Thailand, Zimbabwe, Cambodia, Myanmar, Uganda, Afghanistan, and Mozambique (www.who.int/gtb/Country_info/index.htm). These nations recommend vaccination of children with BCG at birth, and some countries (eg, Brazil and Russia) revaccinate children during the school years. Mexico requires all children to receive BCG once between birth and 14 years of age, and the majority of children receive BCG by 5 years of age.74,75 Thus, the impact of previous BCG immunization on TSTs is of great interest to pediatric health care providers in the United States caring for foreign-born children.
Numerous studies have assessed the relationship between the size of the TST and BCG immunization to determine the extent of false-positive reactions associated with BCG vaccine (Tables 13 and 14). Multiple studies have assessed the size of a single TST after a single BCG immunization. No significant effect of BCG immunization as a risk factor for LTBI was noted among children in New York,18 northern Brazil,76 Uganda,77 or Botswana,78 but the number of children in these studies was modest; only a few hundred children per study were assessed. Larger surveys conducted in Malawi79 and Tanzania80 consisted of >50000 children and found a higher prevalence of positive TSTs (10 mm) in children with a BCG scar when compared with children without a scar. It is somewhat difficult to compare these studies because (1) different methods were used to document BCG immunization, including immunization records and the presence of scars, (2) different vaccine strains and doses were administered, and (3) different TST methods were used.
|
|
15 mm.
Other studies examined the impact of age on the prevalence of TST positivity after BCG immunization. Rates of positive TSTs (10 mm) varied by age: 12% to 31% of 3-month-olds, 3% to 13% of 4-month- to 1-year-olds, and 0% to 18% of children over 1 to 5 years of age had positive TSTs.76–78,81–84 Among older children, 4% to 36% of those 6 to 12 years of age and 7.5% to 15.5% of those 13 to 18 years of age had positive TSTs.81,85–88
Finally, the size of the TST after BCG immunization has been shown to correlate with the risk of developing TB disease. In Singapore, 17% (45 727 of 266 005) of school children who were vaccinated at birth had a TST 10 mm at 12 years of age.89 These children then were followed for 4 years and found to have a 5- to 48-fold increased risk of developing TB disease when compared with children whose TST had been <5 mm at 12 years of age.
In summary, BCG immunization has a variable affect on TSTs. A minority of vaccinated children have a TST 10 mm, and older children are more likely to have a positive TST, suggesting the cumulative effect of exposure to TB disease and the risk of acquiring LTBI. Children who receive BCG after infancy or those who receive >1 BCG immunization also have an increased rate of positive TSTs (Table 14).81,82,84,87,88,90,91 BCG immunization, especially if >1 BCG vaccination is given, is associated with boosting of the DTH response to TST.53,92 Unfortunately, reactivity from BCG cannot be distinguished from reactivity from true infection with M tuberculosis, but data support the conclusion that children from countries with high case rates of TB disease are more likely to have a positive TST from LTBI than from BCG immunization.
Nontuberculous Mycobacteria.
More than 200 M tuberculosis antigens are found in the precipitates of PPD preparations. Many of these antigens are common to Mycobacterium bovis, BCG, and nontuberculous mycobacteria (NTM) (eg, Mycobacterium avium, Mycobacterium intracellulare, Mycobacterium fortuitum, Mycobacterium abscessus, and Mycobacterium kansasii), which can result in cross-reactivity and false-positive reactions to TSTs.93–95 However, a true positive TST can result from disease caused by M bovis. Some of the better-studied mycobacterial antigens include the 65-kDa heat-shock protein, the 38-kDa species-specific protein of M tuberculosis, and the early secreted antigenic target 6 kDa (ESAT-6).96 Some of these antigens form the basis of newly developed tests to improve the specificity of the diagnosis of LTBI, as will be described below.
Boosting Effect.
Over time the DTH to mycobacterial antigens may wane, and thus a TST could be negative. However, with subsequent TSTs, the DTH response may be stimulated by PPD and result in a positive reaction. Such a reaction can be misinterpreted as a recent TST conversion. This phenomenon is known as boosting, ie, an increase in TST size caused by repetitive TSTs in an individual previously sensitized to mycobacterial antigens, particularly BCG and NTM. Boosting is minimized if TSTs are placed <1 week apart.53 However, if a person has not been infected with mycobacterial antigens, boosting will not occur.
Positive Predictive Value of TSTs
The positive predictive value of the TST is influenced by the specificity of the test and the prevalence of true LTBI in the population being tested. The lower the prevalence of LTBI in a given population or the higher the prevalence of exposure to NTM or BCG vaccine, the more false-positive TSTs will occur, which results in lower specificity and lower positive predictive value. Conversely, the positive predictive value of a TST is high when the prevalence of LTBI is high, such as among contacts of a case of TB disease.53
The use of 3 cutoff levels (5, 10, and 15 mm) to define a positive TST in different populations improves the positive predictive value of a TST. Thus, the definition of a positive TST depends on risk factors present in the individual being tested.3 The interpretation of a TST is stratified based on the millimeters of induration (Table 4). A smaller TST (5 mm) is interpreted as positive in children in whom the risk of LTBI (or TB disease) is higher. This lower cutoff level yields a higher sensitivity of the TST (ie, fewer false-negatives). Conversely, in children at lower risk for LTBI or TB disease, a larger cutoff level improves specificity by reducing the number of false-positive interpretations. Notably, testers in California only use 2 cutoff levels (5 or 10 mm) (California Tuberculosis Controllers Association [
www.ctca.org/guidline/combined%20ltbl%20guide2002.pdf]). Targeted tuberculin skin testing should dramatically reduce testing of children at low risk for LTBI and TB and further improve the positive predictive value of TSTs.
Interpretation of the TST by Trained Health Care Workers
Several studies have emphasized that trained health care professionals must place, read, and interpret TSTs. Ozuah et al97 showed that patients can reliably detect the presence or absence of induration but cannot reliably measure or interpret the TST reaction. Howard and Solomon98 demonstrated that 63% (133 of 212) of patients with positive TSTs did not report induration, although 99% (520 of 525) of those with negative TSTs correctly interpreted their skin test as negative. Froehlich et al20 compared TST readings by parents and health care professionals. Parents failed to detect 9.9% of positive TSTs when using the 10-mm cutoff level (1% of cohort) and 5.9% of positive TSTs when using the 15-mm cutoff level (0.5% of cohort). Similarly, Colp et al99 found that only 6% (1 of 18) of patients correctly identified a TST with 10 to 20 mm of induration as 10 mm; 56% (10 of 18) considered the test negative, and 39% (7 of 18) were unable to make a judgment. Cheng et al100 correlated parents’ readings with those of a visiting nurse. In all, 6% (5 of 89) of parents did not note induration observed by the nurse, whereas 3% (3 of 89) reported induration for a negative TST.
These observations extend to untrained health care workers. Carter and Lee101 studied pediatric providers with no specific training in interpreting TSTs to determine if they could interpret a 15-mm TST reaction correctly. Twenty-three percent (13 of 57) read the TST as <10 mm, and 18% (10 of 57) read it as <5 mm. In a similar study, Kendig et al asked 107 health care professionals to interpret a 15-mm TST.102 Overall, 33% (17 of 52) of practicing pediatricians misinterpreted the 15 mm of induration as <10 mm, and only 7% (8 of 107) measured the induration correctly.
In summary, laypersons and untrained health care workers frequently misinterpret TSTs. Only trained health care workers should plant, read, and interpret a TST.
Newer Assays to Diagnose LTBI
In efforts to address the technical limitations of the TST and improve sensitivity, specificity, and convenience, newer assays have been developed that rely on cellular responses to specific antigens of M tuberculosis.
QuantiFERON-TB
QuantiFERON-TB (QFT) (Cellestis Limited, Carnegie, Victoria, Australia) is a Food and Drug Administration–approved diagnostic test for M tuberculosis that quantifies interferon (IFN-) released by sensitized lymphocytes. Whole blood containing lymphocytes is incubated with proteins from M tuberculosis, M avium, and control antigens. After exposure to M tuberculosis complex, lymphocytes that have been sensitized release IFN- that can be quantified. This assay is approved for use in adults.103 Guidelines for using QFT for diagnosing LTBI in adults were published by the CDC in December 2002 and are summarized in Table 15.
|
was 85% (
= 0.55). Among persons being screened for LTBI who had (n = 157) and had not (n = 770) received BCG immunization, a positive TST and a negative QFT assay for M tuberculosis occurred in 22% (35 of 157) and 4% (33 of 770) of persons, respectively. Of the 33 unvaccinated subjects with a positive TST and negative QFT assay for M tuberculosis, 21% (7 of 33) had detectable IFN- for M avium complex. Factors found to be associated with a positive TST and negative QFT for M tuberculosis included a history of BCG immunization, Asian race, study site, and evidence of M avium complex by QFT assay.
Enzyme-Linked Immunospot
Enzyme-linked immunospot (ELISPOT) is an investigational immunoassay that detects IFN- molecules secreted by ESAT-6-specific T cells. ESAT-6 is a secreted antigen specifically expressed by the M tuberculosis complex but absent in strains of M bovis BCG vaccine and most NTM.94 Among patients with culture-confirmed TB disease, 96% (45 of 47) had ESAT-6-specific T cells.93 Lalvani et al93 compared ELISPOT with a multiple-puncture TST (Heaf test) in an effort to diagnose LTBI in contacts of newly diagnosed smear-positive cases of pulmonary TB. ELISPOT identified slightly more infected contacts (73% [16 of 22]) than the Heaf test (65% [13 of 20]). There was a strong positive association between ELISPOT results and increased exposure defined as proximity to the index case and duration of contact (odds ratio [OR]: 9.0 per unit increase in level of exposure; 95% confidence interval [CI95]: 6.0–31.6; P = .001). None of the 19 contacts with BCG immunization and little or no exposure to case patients had a positive ELISPOT, whereas 31% (6 of 19) had a positive Heaf test.
In summary, these newer diagnostic assays show great promise and can differentiate T cell response to M tuberculosis, NTM, or BCG. Second-generation QFT tests are currently being evaluated and may prove more specific than the currently approved assays. There are no published studies in children to date.
Medical History
To diagnose and treat children and adolescents with LTBI correctly, a medical history must be obtained to elicit symptoms of TB disease and the presence of coexisting medical conditions that could complicate treatment of LTBI (Table 5). The most common symptoms of TB are cough, fever, wheezing, and failure to gain weight.58 Infants and adolescents with pulmonary TB are generally more symptomatic than older children. Children with TB disease identified by contact investigations or targeted tuberculin skin testing are often asymptomatic.58 Before initiating treatment for LTBI, other factors such as previous treatment for LTBI or TB, a possible infectious source case, concomitant medical conditions or medications, and maternal and child HIV status may guide treatment and monitoring.
Physical Examination
A directed physical examination in children and adolescents with a positive TST can identify signs of pulmonary or extrapulmonary TB disease (Table 6). Such an examination requires a short time to perform. Particular attention should be given to palpating the cervical lymph nodes, because this is a common site of TB disease in children.
Radiographic Studies
Chest Radiographs
Chest radiographs are considered essential to assess children and adolescents with positive TSTs for pulmonary TB. Chest radiographs in LTBI are usually normal, but findings may include dense nodules with calcifications (ie, a Ghon complex), calcified nonenlarged regional lymph nodes, or both, or pleural thickening (ie, scarring).2,3 Patients with these lesions can be treated for LTBI, because these isolated findings are not associated with an increased risk of progression to active TB compared with radiographs with no abnormalities.2 In contrast, findings consistent with TB disease include enlargement of hilar, mediastinal, or subcarinal lymph nodes and parenchymal changes such as segmental hyperinflation, atelectasis, alveolar consolidation, interstitial infiltrates, pleural effusion, or a focal mass.52 Cavities are rare in young children but may occur in adolescents with reactivation disease. Patients with noncalcified nodular lesions and fibrotic scars may be at higher risk of progression to TB disease and may require additional evaluation for active TB.
Younger children are more likely to have intrathoracic lymphadenopathy than adolescents. Of 4607 children with TB disease studied in California from 1985 to 1995, 6% (157 of 2778) of children 0 to 4 years, 8% (150 of 1829) of children 5 to 14 years, and 0.5% (8 of 1615) of adolescents were reported to have intrathoracic adenopathy.105 Smuts et al106 demonstrated that lateral chest radiographs considerably improved the accuracy of detecting hilar adenopathy in children 1 month to 12 years of age. Among 176 culture-confirmed cases of TB disease, 46% (81 of 176) had adenopathy visible on chest radiographs. Adenopathy was visible on both frontal and lateral views in 49% (40 of 81), on only the frontal view in 24% (19 of 81), and on only the lateral view in 27% (22 of 81) of patients. Furthermore, hilar adenopathy was detected only on the lateral view of 19% (27 of 140) of children diagnosed with probable TB disease who had negative cultures for M tuberculosis.
CT Scans
In recent years, the role of CT scans in pediatric patients with TB disease has been studied. Because of increased sensitivity when compared with chest radiographs, a chest CT scan may show enlarged or prominent mediastinal or hilar adenopathy that is not demonstrable on chest radiographs and is thought to be of no clinical significance.107 CT scans may prove useful in children with equivocal chest radiographs or may help further define an alternative pathologic process. CT scans can demonstrate endobronchial disease, pericardial invasion, early cavitation, or bronchiectasis. Neu et al108 found that 31% (6 of 19) of chest CTs demonstrated mediastinal or hilar adenopathy in children with equivocal or absent adenopathy on chest radiographs. In addition, CT scans provided an alternative diagnosis (eg, a bronchogenic cyst) in some children. However, a pediatric patient with presumptive LTBI generally should not undergo a chest CT.
In summary, there are limited studies demonstrating the yield of lateral chest radiographs for children >6 years of age including adolescents. However, lateral views and chest CTs have been shown to be useful in the assessment of pediatric patients whose frontal views are equivocal for TB diseases.
Cultures for M tuberculosis
If TB disease is suspected, respiratory specimens should be collected. Gastric aspirates or induced sputum may be useful for children who cannot produce sputum. By definition, children with LTBI have a low organism burden, and occasionally such children may have a positive culture from the respiratory tract.58 However, cultures are not recommended to assess children or adolescents with LTBI.
Testing for HIV
It is recommended that all patients with TB disease be offered HI
