 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
From the Division of Tuberculosis Elimination, National Center for HIV, STD, and TB Prevention, Centers for Disease Control and Prevention, Atlanta, Georgia
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Methods. All verified TB cases reported to the national TB surveillance system from 1993 to 2001 were included. A child was defined as a person younger than 15 years.
Results. A total of 11 480 childhood TB cases were reported. Case rates (TB cases/100 000 population) in all children declined from 2.9 (n = 1663) in 1993 to 1.5 (n = 931) in 2001. Among children, those who were younger than 5 years had the highest rate. California, Texas, and New York accounted for 48% of all childhood TB cases. In 2001, TB case rates were higher for foreign-born (12.2) than US-born children (1.1). Hispanic and non-Hispanic black children accounted for nearly three quarters of all cases. Twenty-four percent of children with TB were foreign-born children, with the largest number originating from Mexico (39.8%), the Philippines (8.6%), and Vietnam (5.7%). Most children had evidence of pulmonary TB disease (78.9%). Among culture-positive cases without previous TB, drug resistance to at least isoniazid was 7.3% and to isoniazid and rifampin was 1.6%. In 1999, 82.9% of children received directly observed therapy for at least part of their treatment and 94.8% completed treatment.
Conclusions. Although the overall TB case number among children is declining in the United States, certain groups of children (eg, younger children, racial and ethnic minorities, foreign-born) are at higher risk for TB. As the United States moves toward the elimination of TB, future efforts should endeavor to prevent all cases of childhood TB.
Key Words: epidemiology • tuberculosis • children • pediatrics
Abbreviations: TB, tuberculosis • LTBI, latent tuberculosis infection • HIV, human immunodeficiency virus • RR, relative risk • CI, confidence interval • CDC, Centers for Disease Control and Prevention • TST, tuberculin skin test • INH, isoniazid • MDR, multidrug-resistant • DOT, directly observed therapy
As a result of the national effort to combat the resurgence of tuberculosis (TB) during the late 1980s and early 1990s, overall TB incidence in the United States has been declining.1,2 Despite this decline, outbreaks continue to occur, some of which involve children.3–5 The renewed national effort to eliminate TB includes 3 high priorities: 1) to identify patients with active TB disease, especially those resulting from recent transmission (eg, children); 2) to ensure completion of therapy; and 3) to identify contacts with latent TB infection (LTBI) who are at greatest risk for progression to TB disease (eg, children).6,7
TB disease in a child is considered a sentinel event because it usually represents disease resulting from recent transmission from an adult with infectious pulmonary TB.8,9 In addition, TB disease in children is complicated by the unique challenges that it poses to the clinician. Unlike in adults, the diagnosis of TB disease in children is more difficult owing to several factors, including nonspecific signs and symptoms of TB and the presence of fewer mycobacteria, which results in fewer positive bacteriologic tests.10–12 Children, especially infants, are also at increased risk for progression from LTBI to active TB disease, which may sometimes be severe, such as tuberculous meningitis.13,14 Therefore, understanding TB epidemiology among children is important to overall TB control and to the eventual elimination of TB in the United States. To better understand TB epidemiology among children, including trends in disease, and to highlight associated clinical characteristics and differences when compared with adult disease, we analyzed national TB surveillance data from 1993 through 2001.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Annual population estimates from the 1990 US Census and the Census 2000 were used to calculate rates for 1993 through 1999 and for 2000 through 2001, respectively.16 Annual population estimates from the Census 2000 for race and ethnicity alone classifications were used to calculate race and ethnicity rates for 2000 and 2001. The US Census Bureau's Current Population Survey was used to calculate rates for TB cases in foreign-born and US-born individuals with TB for 2001.17 Relative risks (RRs), 95% confidence intervals (CIs), and 2-tailed P values were calculated using a Mantel-Haenszel 
2 test, and linear trends were tested using 2 for trend. P < .05 was considered statistically significant. Epi Info 6.04d (Centers for Disease Control and Prevention [CDC], Atlanta, GA) and SAS 8.02 statistical programs (SAS Institute, Cary, NC) were used for analysis.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
6% (6.6% in 1993 and 5.8% in 2001). Overall, case rates in children declined from 2.9 cases/100 000 population in 1993 to 1.5 cases/100 000 population in 2001 (Fig 1), and cases declined from 1663 to 931. This represented a 48.3% decrease in the case rate and a 44.0% decrease in the case number. Among children, the highest annual rates were consistently among those who were younger than 5 years, whose rates were 2.1 to 3.5 times greater than for children aged 5 to 9 and 10 to 14 (Fig 1). In 2001, the TB case rate for children who were younger than 5 years was 2.8 cases/100 000 population, for 5 to 9 years of age was 1.0 cases/100 000 population, and for 10 to 14 years of age was 0.9 cases/100 000 population (Fig 1).
|
500 000 people. Almost one third of the total were reported by 5 metropolitan statistical areas (New York City: 7.8%, n = 57; Los Angeles: 7.6%, n = 55; Chicago: 5.1%, n = 37; San Diego: 4.7%, n = 34; and Houston: 4.1%, n = 30).
Race/Ethnicity
Although rates for all race and ethnic groups generally decreased between 1993 and 2001 (Fig 2), children who were racial or ethnic minorities shared a disproportionate burden of TB (Table 1). Hispanic and non-Hispanic black children accounted for 73.9% of all childhood TB cases reported between 1993 and 2001 (Table 1). Children younger than 5 years had higher case rates among all race and ethnic categories than children aged 5 to 14, except for non-Hispanic whites, in whom the rates were similar. In 2001, among children younger than 5 years, case rates were significantly lower for non-Hispanic whites (0.5 cases/100 000 population) than for non-Hispanic Asian/Pacific Islanders (7.7), Hispanics (6.8), non-Hispanic blacks (6.0), and non-Hispanic American Indian/Alaska Natives (2.3).
|
|
|
|
9 years) to have a sputum smear done (10–14 years: 63.5%; 5–9 years: 29.7%; <5 years: 12.8%; P < .001), more likely to be smear-positive (10–14 years: 10.3%; 5–9 years: 1.8%; <5 years: 1.6%; P < .001), and more likely to be sputum culture-positive (10–14 years: 21.3%; 5–9 years: 5.0%; <5 years: 4.2%; P < .001). Only 2.4% of all childhood TB cases had a gastric aspirate smear positive for acid-fast bacilli, and 8.5% had a gastric aspirate culture positive for M tuberculosis. However, only positive gastric aspirate results are reported, so it is not possible to know how many specimens were obtained. The majority of children with positive gastric aspirate test results were younger than 2 years (smear: 77.2%; culture: 73.5%) and were US-born (smear: 88.4%; culture: 91.1%). Children who had pulmonary TB and were aged 10 to 14 had higher rates of cavitary disease than younger children (11.8% vs 3.5%; P < .001). We examined young children with cavitary disease more closely. From 1993 to 2001, 190 children who were younger than 5 years and had pulmonary and cavitary TB disease were reported, including 54 (28.4%) children who were younger than 1 year. Nearly all (90.5%) were US-born. More than one third (39.0%) were culture-positive and 11.6% were sputum smear-positive at diagnosis. Between 1993 and 2000, 3.4% were HIV-positive; however, most (76.3%) did not have a reported HIV test result. Between 1993 and 1999, 90.0% completed therapy and 2.5% died (1999: 100% and 0%, respectively).
From 1993 to 2000, most (83.9%) children did not have a reported HIV test result. Nearly all (95.4%) children had a tuberculin skin test (TST), and most (89.0%) were positive (as defined by CDC guidelines).18 However, children who had a diagnosis of miliary or meningeal TB were less likely to have a positive TST (57.6% and 54.6%, respectively) than those with pulmonary TB (90.6%; P < .001 for both comparisons).
Drug Resistance
Between 1993 and 2001, among culture-positive childhood TB cases without a history of TB, rates of drug resistance were 7.3% to at least isoniazid (INH), 15.2% to at least any first-line drug (isoniazid, rifampin, ethambutol, pyrazinamide, or streptomycin), and 1.6% to at least INH and rifampin (ie, multidrug resistant [MDR] TB; Table 2). Drug resistance rates among foreign-born pediatric TB patients were higher than for US-born patients (P < .05). Overall, rates for INH resistance and MDR TB were relatively stable between 1993 and 2001 (2001: resistance to at least INH, 6.5%, MDR TB, 1.2%).
Directly Observed Therapy and Treatment Outcome
Use of directly observed therapy (DOT) for at least part of treatment in children steadily increased from 34.2% in 1993 to 82.9% in 1999 (2 for trend, P < .001). US-born children more frequently received at least some DOT than foreign-born children (RR: 1.2; 95% CI: 1.1–1.2; Table 3). In 1999, 95.7% of US-born and 93.2% of foreign-born children completed therapy (P = .15).
|
Adolescent TB
From 1993 to 2001, there were a total of 3396 TB cases among adolescents aged 15 to 18 years. The annual proportion of adolescent TB cases increased from 1.6% in 1993 to 2.4% in 2001; the annual number remained stable (396 in 1993 and 386 in 2001). Case rates in this population declined from 2.9/100 000 population in 1993 to 2.4 cases/100 000 population in 2001. The decrease in case rates and case numbers in adolescents (17.2% and 2.5%, respectively) was less than the decrease seen among children (48.3% and 44.0%, respectively). States with the largest number of adolescent cases were similar to those for children and adults with TB, with California (25.0%), New York (13.0%), and Texas (11.2%) accounting for the largest proportion of cases. Adolescents who were Hispanic (35.4%), non-Hispanic black (31.5%), or Asian/Pacific Islander (23.2%) accounted for >90% of all cases. The proportion of Asian/Pacific Islanders among adolescent cases was twice that of children (23.2% vs 11.7%). In 2001, Asian/Pacific Islanders had the highest case rate (11.8 cases/100 000 population) followed by Hispanic (5.9), non-Hispanic black (5.7), American Indian/Alaska Native (1.8), and non-Hispanic white (0.3). Case rates among adolescents declined among all racial and ethnic groups from 1993 to 2001, except among non-Hispanic blacks. Compared with children, more adolescent cases were foreign-born (57.6% vs. 24.2%). Among adolescent foreign-born TB cases, 48.1% were diagnosed within 18 months of arriving in the United States (median: 16.0 months; mode: 1.0 months). More foreign-born (54.2%) adolescents were male than US-born adolescents (45.5%). The most common countries of origin were Mexico (25.7%), Vietnam (9.4%), and the Philippines (9.0%). Adolescents with TB were more likely than children to be sputum smear-positive (31.3% vs 3.2%) and sputum culture-positive (54.2% vs 7.4%). Most (78.4%) adolescents had pulmonary disease and were more likely than adults (P < .01) or children (P < .0001) to have pleural disease (5.2% vs 4.1% and 1.2%, respectively). More (26.7%) adolescents had evidence of cavitary disease on chest radiograph than adults (24.7%; P < .01) or children (4.4%; P < .0001). Most (82.2%) cases had a positive TST at diagnosis. From 1993 to 2000, 1.2% of adolescents with TB were HIV-positive; however, most (64.9%) adolescents did not have a reported HIV test result. Between 1993 and 2001, among culture-positive adolescents without a history of TB, rates of drug resistance were 10.5% to at least INH, 15.3% to at least any first-line drug, and 1.4% to at least INH and rifampin (ie, MDR TB). These rates were similar to adults and children with TB. In 1999, 94.5% of adolescents completed therapy.
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
From our analysis, we observed that TB case rates among foreign-born children are significantly higher than in US-born children. Mexico, the Philippines, and Vietnam have been reported as the most common countries of birth of foreign-born children and adults in the United States with TB; recently, countries in Sub-Saharan Africa (eg, Somalia, Ethiopia) and Eastern Europe (eg, Russia) have also increasingly been reported. Fluctuations in the proportion of countries of birth for people with TB are probably directly related to changing immigration patterns. Immigration from regions with higher rates of drug-resistant TB (eg, Eastern Europe) and from regions with high rates of HIV infection (eg, sub-Saharan Africa) has important implications for TB control among children.
In our analysis, the effect of immigration on childhood TB is probably underestimated, because the national TB surveillance system currently does not collect information on travel history or countries of birth of parents or guardians of childhood TB patients. Several studies have shown that family immigration patterns or travel history put a US-born child at risk for TB.24,25 Therefore, many US-born children with TB may have been exposed to increased risks associated with TB in the foreign-born, although they were born in the United States.
For identifying TB disease among people who enter the United States, a medical examination, including TB screening, is performed overseas on all refugees or people who apply for an immigrant visa; however, screening requirements vary depending on age. Unless a child (<15 years) is suspected of having TB disease or has a history of contact with a known TB case, no TB screening is performed.26 Our analysis indicated that >60% of TB cases in foreign-born children were diagnosed within 18 months of arrival in the United States. Therefore, when caring for immigrant children, including those adopted internationally,27 and US-born children with foreign-born parents or guardians, it is important to consider their risk for TB, especially during their initial medical examination in the United States. The American Academy of Pediatrics recommends a comprehensive medical examination, including TB screening, within 2 weeks of arrival in the United States.9 TB screening should include obtaining a thorough history, specifically asking about contact with a known infectious TB case, and, if warranted, a TST (Mantoux method) to identify children who have LTBI and TB disease.28,29 In addition, several studies have documented the advantages of using a TB risk assessment questionnaire as a screening tool to identify children who should receive tuberculin skin testing.30–32 New variables are planned for inclusion in the upcoming revision of the national TB case report form, including travel history, parental/guardian birth country, and immigration status of TB patients, which, in turn, will allow for better assessment of TB burden and monitoring of trends among foreign-born children, including those who are immigrants and refugees and potentially those born in the United States of immigrant and refugee parents.
Only 1.1% of childhood TB patients were reported to have a positive HIV test result. Reporting of HIV test results for pediatric patients was poor despite the documented association between TB and HIV infection in children33–35; the CDC recommends that all patients with TB disease receive voluntary HIV counseling and testing at diagnosis.36
Our findings indicate that drug-resistance patterns for INH and MDR TB in children closely mirror those of adults. This is consistent with a previous study.37 Increased risk for drug resistance has been associated with previous treatment for TB disease, contact with a patient who has drug-resistant TB, being born in or emigrating from areas where drug-resistant TB is endemic, and remaining sputum smear- or culture- positive or not clinically improving despite adequate treatment.9 Therefore, even in the absence of culture confirmation, drug resistance should always be considered in children with TB who are at risk as described, especially in children who are not responding to TB treatment. In addition, given that fewer than one quarter of children had culture-positive disease, determining the drug-resistance pattern of the adult source case is critical in the management of a child with TB. Ensuring that children take their anti-TB medications through the use of DOT is also critical to successful treatment.38 In our analysis, DOT use has increased since 1993, and in 1999, >80% of all children received DOT during at least part of their therapy.
The diagnosis of TB in children, particularly in children younger than 5 years, remains difficult. Sputum specimens are frequently difficult to obtain and often have a low yield.12,13,39 Gastric aspirates remain useful in children younger than 5 years, although yields on culture are generally <50%.10,40 The low proportion of children reported with sputum culture-positive disease reflects, in part, that sputum cultures may not be attempted in many children. Bronchoscopy and nucleic acid amplification tests (eg, polymerase chain reaction), have generally shown disappointing results for the diagnosis of childhood TB.39–43 Confirming a bacteriologic diagnosis of TB in a child is particularly important if drug resistance is suspected. In the absence of a better diagnostic test, the optimal diagnosis of TB in children still rests on a thorough history, especially a history of contact with a known TB case (eg, epidemiologic link), and a history of being born in, traveling to, or having a parent or guardian born in a country where TB is endemic. Tuberculin skin testing is currently the foremost tool for identifying LTBI, even if a person was immunized with Bacille Calmette-Guerin.44 Together with other findings (eg, clinical, history), TST can also support a diagnosis of TB disease.
One possible limitation of our analysis of national TB surveillance data is completeness of reporting, especially among children, because <25% of childhood TB cases are culture confirmed. A previous study had found good completeness of the national TB surveillance system.45 However, 1 recent study from New York City highlighted many of the challenges to assessing the completeness of reporting of childhood TB cases.46
Understanding TB epidemiology among children is critical to overall TB control and especially to the goal of TB elimination in the United States.8,9 Most TB disease may be effectively prevented through the treatment of LTBI. Thus, as preventable cases, infected children should be an important target in the efforts to treat LTBI.47 Although the overall TB case number among children is declining in the United States, the numbers for certain groups (eg, racial and ethnic minorities, foreign-born) are unchanged or even increasing. Adolescents with TB may also represent a vulnerable group and are generally more infectious than younger children. Clinicians, even in low-incidence areas, need to remain aware of the possibility of TB in these high-risk groups and screen appropriately, using available risk assessment questionnaires to help identify which children need tuberculin skin testing.30–32 Thorough contact investigations for all TB cases, especially for adult cases, that result in identification of childhood contacts is imperative in preventing TB disease and in achieving TB elimination. Interrupting transmission by identifying and successfully treating infectious TB patients and identifying those with LTBI is essential to eliminating TB among children and eliminating TB in the United States. Future efforts to decrease the incidence of childhood TB should endeavor to prevent all cases of childhood TB, particularly in populations at highest risk, because each childhood case of TB could be preventable.
|
ACKNOWLEDGMENTS |
|---|
We thank the state and local tuberculosis control officials in health departments throughout the United States who collected and reported the national surveillance data presented in this article and the surveillance staff at the Division of Tuberculosis Elimination who maintain the national database. In addition, we thank Kenneth G. Castro, MD; Michael F. Iademarco, MD, MPH; Kayla F. Laserson, ScD; Thomas R. Navin, MD; and Jeffrey Starke, MD, for critical review of this manuscript.
|
FOOTNOTES |
|---|
Reprint requests to (L.J.N.) Division of Tuberculosis Elimination, Centers for Disease Control and Prevention, 1600 Clifton Rd, MS E-10, Atlanta, GA 30333. E-mail: lbn9{at}cdc.gov
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  I. Trehan, J. K. Meinzen-Derr, L. Jamison, and M. A. Staat Tuberculosis Screening in Internationally Adopted Children: The Need for Initial and Repeat Testing Pediatrics, July 1, 2008; 122(1): e7 - e14. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. I. Flores, W. B. Bilker, and E. A. Alessandrini Effects of Continuity of Care in Infancy on Receipt of Lead, Anemia, and Tuberculosis Screening Pediatrics, March 1, 2008; 121(3): e399 - e406. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. H. Baumann, R. Nolan, M. Petrini, Y. C. G. Lee, R. W. Light, and E. Schneider Pleural Tuberculosis in the United States: Incidence and Drug Resistance Chest, April 1, 2007; 131(4): 1125 - 1132. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. E. Stout, K. K. Saharia, S. Nageswaran, A. Ahmed, and C. D. Hamilton Racial and Ethnic Disparities in Pediatric Tuberculosis in North Carolina Arch Pediatr Adolesc Med, June 1, 2006; 160(6): 631 - 637. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. A. Jereb, L. J. Nelson, and K. G. Castro Commentary on the Risk of Active Tuberculosis Arch Pediatr Adolesc Med, March 1, 2006; 160(3): 317 - 318. [Full Text] [PDF]
|
|
|
|
|
|
S. H. Wootton, B. E. Gonzalez, R. Pawlak, L. D. Teeter, K. C. Smith, J. M. Musser, J. R. Starke, and E. A. Graviss Epidemiology of Pediatric Tuberculosis Using Traditional and Molecular Techniques: Houston, Texas Pediatrics, November 1, 2005; 116(5): 1141 - 1147. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S Howie, L Voss, M Baker, L Calder, K Grimwood, and C Byrnes Tuberculosis in New Zealand, 1992-2001: a resurgence Arch. Dis. Child., November 1, 2005; 90(11): 1157 - 1161. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Schneider Tuberculosis Among American Indians and Alaska Natives in the United States, 1993-2002 Am J Public Health, May 1, 2005; 95(5): 873 - 880. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
|
||
Home | My Pediatrics | Journal Information | Current Issue | Past Issues | Subscriptions & Services | Contact Us Click here for faster international access © 2004 American Academy of Pediatrics. All rights reserved. |
||
|
||
