Childhood TB in North Carolina

Published on June 2016 | Categories: Documents | Downloads: 31 | Comments: 0 | Views: 182
of 5
Download PDF   Embed   Report



Childhood Tuberculosis in North Carolina: A Study of the Opportunities for Intervention in the Transmission of Tuberculosis to Children
Abstract: New cases of tuberculosis in children continue to appear. A retrospective review of the medical records of a representative sample of such cases occurring from 1977 through 1981 was conducted to gain an understanding of why this preventable disease continues to occur in North Carolina children. Three per cent of cases were detected by routine screening; 17 per cent were diagnosed after the child's symptomatic presentation; and 80 per cent after contact screening. However, 14 per cent of cases detected by contact
screening developed disease while receiving isoniazid prophylaxis for a positive tuberculin test; and 19 per cent of cases detected by contact screening developed disease while under surveillance, but not receiving isoniazid prophylaxis, as tuberculin negative contacts of known infectious cases. Adherence to accepted recommendations for prophylaxis should reduce the childhood tuberculosis case rate. (Am J Public Health 1986; 76:26-30.)

Introduction Despite nearly a century of public health intervention and the availability of specific chemotherapy for more than 30 years, tuberculosis remains a relatively unusual but important cause of morbidity in children. 2 New cases of tuberculosis in children continue to appear in North Carolina. In recent years, an average of 47 such cases per year were reported. In an attempt to identify the variables associated with tuberculous disease in children, this study examined the demographic characteristics of reported childhood cases in North Carolina and the circumstances of detection, diagnosis, and management. Certain clinical and epidemiologic features of childhood tuberculosis in North Carolina are described.

Methods Data were obtained in two phases: 1) data on cases statewide were obtained from the North Carolina Tuberculosis Control Division (NCTCD); 2) additional data were then gathered on cases in selected counties. A tuberculosis case is defined as any child in whom the presence of Mycobacterium tuberculosis has been confirmed by bacteriologic studies, or, in the absence of bacteriologic confirmation, any child who meets all of the following criteria: 1) a significant tuberculin skin test reaction; 2) appropriately abnormal unstable chest x-rays or clinical or pathologic evidence of disease or both; 3) receiving a course of chemotherapy with two or more antituberculous medications. These are the standard Centers for Disease Control (CDC) criteria for the case definition of tuberculous disease. The reporting criteria specifically exclude individuals who are recent skin test converters without evidence of active disease and individuals for whom the attending physician believes singie drug therapy for active disease is sufficient. The NCTCD requires the following data to be reported to it for each case: name, age, sex, race (White, Black,
Address reprint requests to Robert J. Nolan, Jr., MD, Department of Pediatrics, University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Drive, San Antonio, TX 78284. At the time this study was conducted, Dr. Nolan was a Robert Wood Johnson Foundation Fellow in General Academic Pediatrics at Duke University Medical Center. This paper, submitted to the Journal March 5, 1985, was revised and accepted for publication August 22, 1985. Editor's Note: See also related editorial p 12 this issue.
© 1985 American Journal of Public Health 0090-0036/85$1.50

American Indian, or Asian origin), county of residence, site of disease, culture status (positive, negative, or not done), and reporting agency (private physician, county health department, state hospital, private hospital, etc.). These data (hereafter referred to as the NCTCD data variables) were made available to the author on all 235 cases 0 through 14 years of age reported in North Carolina from January 1, 1977 through December 31, 1981. They were used to create an incidence map of disease (Figure 1). To select a sample of counties for further data collection, counties were divided into three groups: counties with no cases; low incidence counties, with an average yearly incidence below the statewide mean of six cases per 100,000; and high incidence counties with an average yearly incidence greater than the statewide mean. A pilot study was conducted in two high incidence and one low incidence counties. From their respective high and low incidence county groups in the eastern two-thirds of North Carolina (counties within a 150-mile radius of Durham County), five additional high incidence counties and six additional low incidence counties were randomly selected, creating a total sample of 14 counties evenly divided between high and low incidence counties (Figure 1). Since state policy provides for free diagnostic services and anti-tuberculous medications for suspect cases and their contacts, county health departments (CHD) maintain extensive records on reported cases. Additional patient data on reported cases (history of presentation, family history, symptomatology, chest x-ray status, bacteriologic status, hospitalization history, etc.) were obtained at their local county health departments by an on-site, retrospective, medical record review. Specific data regarding compliance with medication regimens were unavailable. The medical records of all 118 reported cases in the 14 selected counties, representing 50.2 per cent of all cases reported statewide during the survey period, were examined. Rates of change in the incidence of disease during the survey period were determined from the linear regression of the logarithms of the annual incidence rates to ensure comparability with recent CDC analyses of national data. Results Statewide Data There were 235 children 0 through 14 years of age reported as cases from 1977 through 1981; 179 of the children (76.2 per cent) were Black, 48 (20.4 per cent) were White, 6
AJPH January 1986, Vol. 76, No. 1



U QI-6 Average

Lb t ut ide S 6.1+ *oses per IO0qqO0) -cif krn 044 eas iE age 1977 19t1

* Durham County


FIGURE 1-Average Annual Inddence of Tubercukles In ChIldrep 0 through 14 Years of Age, North CaroHlia, 1977-81

(2.5 per cent) were of Asian origin, and 2 (0.9 per cent) were American Indians. No Hispanic surnamed children werereported. Female cases outnumbered male cases slightly (125/235, 53 per cent v 110/235, 47 per cent). The incidence of tuberculosis changed by an average of -1.9 per cent per year from 1977 to 1981 (Table 1). The change in incidence averaged +0.02 per cent per year in the White population and -1.9 per cent per year in the non-White population. When cases of Asian origin were excluded from the data analysis, the rate of change in the total disease incidence increased to -3.6 per cent per year and in the non-White population to -2.5 per cent per year. Nonpulmonary cases represented 20 per cent (47/235) of the total. White children were more likely to have nonpulmonary disease than non-White children (15/48, 31 per cent v 32/187, 17 per cent), a difference due to the greater frequency of lymphatic nonpulmonary disease in White children than in non-White children, (12/48, 25 per cent v 15/187, 8 per cent; difference = 17 per cent, 95 per cent confidence intervals 4 per cent, 30 per cent). The proportion of cases who underwent bacteriologic studies and the rate of positive cultures among those studied declined from 1977 to 1981 (Figure 2). This decline was unrelated to race, age, or sex. Bacteriologic studies were obtained with similar frequency in nonpulmonary and pulmonary cases (34/47, 72 per cent v 141/188, 75 per cent); but when bacteriologic studies were obtained, positive results were more common in nonpulmonary cases than pulmonary' cases (27/34, 79 per cent v 45/141, 32 per cent; difference
1977 1978 1979 1980 1981 Average Annual 1977-81

47 per cent, 95 per cent confidence intervals 31 per cent, 63 per. cent).
County Survey Patient Data The 118 cases whose medical records were reviewed did 'not differ fromn the entire population of cases in the eastern two-thirds of North Carolina with respect to the NCTCD data variables. None of the sample cases were children of migrant farm workers. Medical records were available for all 118 cases and were complete in 109 cases. The nine cases for whom only partial records were available did not differ from the remaining sample cases with respect to any NCTCD data variable. Ten cases (9 per cent of-those with complete charts) were either misreported to the NCTCD or did not meet the CDC case definition (three adult cases were reported as children;'two cases'were double reported by different agencies.; two lymphatic cases had culture isolates identified as Mycobacterium avium-intracellulare; and three were culture negative children with insufficient clinical, roentgenologic, or tuberculin test evidence of disease). The ten noncases, excluding the misreported adults, did not differ from the cases with respect to. any NCTCD data variable except that they: were more likely to be residents of low incidence counties than were the cases (6/7, 86 per cent v 37/99, 37 per cent; difference = 49 per cent, 95 per cent confidence intervals 22 per cent, 76 per cent). Noncases and cases with only partial records were excluded from further analysis leaving a sample of 99 valid cases. All cases had tuberculin skin test reactions (PPD, 5TU by the Mantoux method) greater than 10 mm induration with
Total Non-White Rate
11.4 8.7 7.4 10.7 9.3 9.5

TABLE 1-Tuberculosis Incidence Rates* for Children In North Carolina, 1977481
Total Rate
4.0 3.1 2.5 4.0 3.2 3.3

White Rate
1.1 1.1 0.6

Black Rate

Armerican lndian Rate

Asian Rate



1.4 0.8 1.0

Total Cases 0




7.5 10.2... 9.6 9.6

0.0 0.0 4.9

9.0 0.0 9.0 26.9 9.0

19 (2)



l r ana. hmun,h 14 vaars. uIuI I-9 yavuo. J'('o s nr r I18, V oIl. 76 n No 1 asH pJn iVVnn ..hilefran

AJPH January 1986, Vol. 76,

NJo. 1



cases, in which disease developed while the child was upder surveillance, did not differ from other cases in the NCTCD data variables examined, or in ultimate severity of chest x-ray or clinical involvement. They were uniformly distributed over the period of the survey.
The circumstance of detection (CS v SPP v RS) was unrelated to race, age, sex, or county of residence. Cases with nonpulmonary involvement were more likely to have been detected after SPP than CS (6/17, 35 per cent v 5/79, 6 per cent; difference = 29 per cent, 95 per cent confidence intervals 6 per cent, 52 per cent). Chest x-ray involverhent at diagnosis was more often mild or moderate than severe. Mild chest x-ray involvement



0-e 40






FIGURE 2-Percentage of Cases 0 through 14 Years of Age with Bacteriologic Studies Obtained -4; Percentage of Positive Studies among those Cases Studied A A; and Percentage of Cases with Positve Bacteriologic Studies

an increase of at least 6 mm over any previous reaction. For those cases in which exact measurements were noted (n = 78), the mean cutaneous induration was 18.5 mm, (range 10-70 mm); 81 per cent (63/78) of the reactions were 15 mm or greater. Three cases (3 per cent) were detected and diagnosed after routine screening (RS); 17 cases (17 per cent) were diagnosed after the child's symptomatic presentation to a

(significant unilateral or bilateral hilar or paratracheal adenopathy) was present in 42 (42 per cent); moderate involvement (hilar or paratracheal adenopathy with segmental infiltrates or collapse) in 37 (37 per cent), and severe
involvement (miliary or cavitary disease or moderate involvemient with recurrent pleural effusions) in 8 (8 per cent). Seven chest x-rays (7 per cent) were read as "positive" or "consistent with the clinical diagnosis" without further elaboration; 5 (5 per cent) were read as negative (all lym-

phatic nonpulmonary cases). Severity of chest x-ray involvement was unrelated to circumstance of diagnosis, race, age, or sex.

physician (SPP) with clinical signs referrable to tuberculosis, and 79 cases (80 per cent) were detected after contact screening (CS) of the child by CHD personnel in response to a reported case of disease. All cases detected by routine screening and 77 per cent (61/9) of cases detected by contact screening were asymptomatic. Medical history disclosed the identity of the source case in 10 of the children diagnosed after symptomatic presentation to a physician. Medical history failed to reveal the source of disease in seven cases detected by SPP and all three cases detected by RS. In four instances in which there was no source revealed by history, the diseased child (one detected
by RS and three detected by SPP)
served as an index case

Specimens were obtained for culture with similar frequency in cases with mild chest x-ray involvement as opposed to moderate or severe involvement (26/42, 62 per cent v 35/45, 78 per cent). Cultures were positive with similar frequency in the two groups (8/26, 31 per cent v 13/35, 37 per
Cases under 5 years of age were as likely to have cultures obtained as those over 5 years (34/51, 67 per cent v 35/48, 73 per cent) and those cultures were as likely to be positive in the younger age group as in the older (11/34, 32 per cent v 14/35, 40 per cent). Culture rate and results were unrelated to race, sex, or county of residence. Cases detected by SPP were more likely to have had bacteriologic cultures obtained than cases detected by CS (16/17, 94 per cent v 51/79, 65 per cent; difference = 29 per cent, 95 per cent confidence intervals 14

leading to the detection of diseased adults. Nearly all (98 per cent, 58/59) of the index cases leading to the CS detection of children and the historically revealed source cases were adults; 88 per cent (51/58) of these adults were household

per cent,

have been difference

Sixty-seven per cent of the CS-detected cases (53/79) had positive tuberculin tests and clinical or chest x-ray evidence
of disease at their initial encounter for contact screening. additional 14 per cent (11 /9) had positive tuberculin tests at their initial encounter without clinical or chest x-ray signs of disease. These 11 children were placed on secondary isoniazid prophylaxis but developed signs of disease later (mean = 97 days, range 22-191 days). Nineteen per cent of the CS-detected cases (15/9) had

per cent) and those cultures were more likely to positive (10/16, 63 per cent v 15/51, 29 per cent; 34 per cent, 95 per cent confidence intervals 7 61 per cent). No culture isolate demonstrated per cent, primary drug resistance to isoniazid, rifampin, or paraaminosalicylic acid.

Most cases (51 per cent, 51/99) were hospitalized at shortly after the time of diagnosis. Seventy-five per cent



these (38/51) were hospitalized at the North Carolina State Sanitorium and the remainder at community or at teaching hospitals. Although the two hospitalized groups did not differ with respect to demographic characteristics, circumstances of detection, presence of symptoms, chest x-ray or culture


the mean






a negative tuberculin test at their initial encounter. These children were followed with skin tests at three-month intervals, or when clinically indicated, without receiving primary

longer at the State Sanitorium than at other hospitals; days 37 days, 95 per cent confidence v 9 days, difference
intervals 15 days, 59 days. In no case was disease of reactivation, and there was no mortality.
the result

isoniazid prophylaxis while remaining in close contact with the diseased individual who had prompted the contact screenhave positive skin tests and evidence of disease (mean = 113 days, range: 37-154 days) after initial contact screening. Thus 33 per cent (26/79) of the cases detected by contact screening (26 per cent [26/991 of all cases) developed disease while on secondary prophylaxis, or after failure to receive primary prophylaxis.

ing. All were subsequently discovered to

In the 15 years prior to 1977,1,2 the national tuberculosis case rate for children 0 through 14 years of age relatively more rapidly (8.8 per cent per year) than through case rate (4.4 per cent per year). However, from




1981, the last five years for which complete

data are available,


AJPH January 1986, Vol.




national childhood case rates increased 0.2 per cent per year2 while the total case rates continued to decline nearly 3.5 per cent per year. The stability of childhood case rates has been attributed to an increased incidence of disease in Hispanic children and the influx of Indochinese refugees in the late 1970s.2 Childhood tuberculosis case rates in North Carolina declined an average of 1.9 per cent per year from 1977 through 1981. Although subtracting the Hispanic cases from the national White case rates and the Indochinese refugee cases from the national non-White case rates for the period 1977-81 returned the average annual rate of decline to the pre-1977 average of 8.8 per cent per year, an effect of similar magnitude was not evident in North Carolina. The indigenous childhood population ofNorth Carolina experienced a slower decline in the case rate than the indigenous national population. This slower rate of decline is unlikely to have been due to an increase in overdiagnosis. Noncases among the children reported were uniformly distributed during the study period. Although the proportion of cases with bacteriologic confirmation declined during the study period partially as a result of a decrease in the procurement of cultures, it averaged 31 per cent, which is comparable to the 20 per cent-30 per cent range found nationally during the same period.2 Diagnostic confusion with disease caused by atypical Mycobacteria (to which skin test sensitivity is common in eastern North Carolina),3 was probably minimal in view of the clinical characteristics of the cases reported. Moreover, the mean skin test induration was 18.5 mm, similar to the mean indurations of 16 to 17 mm reported for series of Mycobacterium tuberculosis culture positive patients.4 Routine screening was only of minor importance in the detection of cases (3 per cent, or about 0.9 cases per 1,000,000 children per year). These data do not address the continuing controversy as to the efficiency and effectiveness of routine screening with regard to converter detection and prevention of disease.5 It is interesting to speculate, however, since childhood converters and childhood cases probably acquire their infection from the same group of diseased adults, that routine screening may also be less important than contact screening in the detection of converters. The vast majority of cases, 97 per cent, were detected by means other than routine screening. Contact screening was responsible for the detection of 80 per cent of cases. Through contact screening, most cases were detected prior to the onset of symptoms. Thus prompt reporting of diseased adults to public health authorities was the single most important factor in the early detection of childhood cases. Although only 17 per cent of cases were diagnosed after the child's symptomatic presentation to a physician, many ofthese cases were nonpulmonary and required familiarity with the multiple presentations of tuberculosis. A small group of children, 14 per cent of those cases detected by contact screening, developed disease while on secondary isoniazid prophylaxis for a positive tuberculin test. Since the total number of tuberculin positive children detected by contact screening and placed on prophylaxis is unknown, it is possible that the cases that failed prophylaxis represent the biological failure rate of isoniazid prophylaxis, which has been estimated by Hsu as no greater than 0.3 per cent.6 A more likely explanation, however, is lack of compliance with the prescribed prophylaxis. Of greater concern is the development of disease in tuberculin negative contacts (19 per cent of all cases detected
AJPH January 1986, Vol. 76, No. 1

by contact screening) who were followed without primary isoniazid prophylaxis. The American Academy of Pediatrics and the American Thoracic Society recommend isoniazid prophylaxis for three months for tuberculin negative children remaining in a potentially infectious environment, with further skin testing at the end of the three-month period.7'8 The recommendations are- based upon several facts: contacts can be infected and be initially tuberculin negative since skin test sensitivity requires three to eight weeks to develop after exposure9; children progress from infection to disease with greater frequency than adults10; and the severe forms of disease-meningitis and miliary disease-usually occur within the first several months after the infecting exposure.'1 Isoniazid prophylaxis prevents progression to disease in infected tuberculin negative contacts and ,decreases the frequency of development of infection and skin test conversion in uninfected tuberculin negative contacts.9"12"13 Our data clearly support the use of isoniazid prophylaxis for tuberculin negative contacts as a possible means of reducing the childhood case rate. The need for prophylaxis of tuberculin negative contacts has been increased by the recent trend toward greater use of outpatient care for adults with positive sputum cultures.'4 The results of this study indicate that the decision to obtain cultures should not be -influenced by the severity of chest x-ray presentation or the age of the case at diagnosis. Although there were no cases of primary drug resistance among the survey cases, the importance of vigorous culturing is emphasized by the nationalprimary drug resistance rate of 14 per cent among children." In summary, childhood tuberculosis case rates in North Carolina continue to decline slowly. Investigation of contacts of adult cases is the most important method of childhood case detection. However, two groups of cases detected by such investigation developed disease while under surveillance. Careful monitoring of compliance with isoniazid prophylaxis and adherence to accepted recommendations for isoniazid prophylaxis for tuberculin negative contacts should reduce the childhood tuberculosis case rate.
This work was supported, in part, by the Robert Wood Johnson Foundation Program for Academic Development in General Pediatrics. The author thanks Jim Jones, North Carolina Tuberculosis Control Division, and the County Public Health Nurses without whose cooperation this study could not have been done. The author acknowledges Dr. Laura T. Gutman for suggestions, Sandra Funk for knowledgeable statistical consultation, and Dr. Thomas E. Frothingham for his suggestions, advice, criticism, and review of the manuscript. This study was presented in part at the General Pediatrics Academic Development Program annual meeting sponsored by the Robert Wood Johnson Foundation, Princeton, NJ, June 7-9, 1984.


1. Centers for Disease Control: Tuberculosis Statistics, States and Cities. Atlanta: CDC 1981; 35. 2. Poweli KE, Meador MP, Farer LS: Recent trends in tuberculosis in children. JAMA 1984; 251:1289-1292. 3. Edwards LB, Acquaviva FA, Livesay VT, et al: An atlas of sensitivity to tuberculin, PPD-B and histoplasmin in the United States. Am Rev Respir Dis 1969; 99 (suppl):1-132. 4. Comstock GW: Epidemiology of tuberculosis. Am Rev Respir Dis 1982; 125 (suppl):8-15. 5. Edwards PQ: Tuberculin testing of children. Pediatrics 1974; 54:628630. 6. Hsu KHK: Thirty years after isoniazid-its impact on tuberculosis in children and adolescents. JAMA 1984; 251:1283-1285. 7. American Academy of Pediatrics: Report of the Committee on Infectious Disease, 19th Ed. Evanston, IL: AAP, 1982. 8. American Thoracic Society: Control of tuberculosis. Am Rev Respir Dis



1983; 128:336-342. 9. Smith MHD, Marquis JR: Tuberculosis and other mycobacterial infections. In: Feigin RD, Cherry JD (eds): Textbook of Pediatric Infectious Diseases, Vol. 1. Philadelphia: W. B. Saunders, 1981; 1016-1060. 10. Comstock GW, Livesay VT, Woolpert SF: The prognosis of a positive tuberculin reaction in childhood and adolescence. Am J Epidemiol 1974; 99:131-138. 11. Wallgren A: The timetable of tuberculosis. Tubercle 1948; 29:245-251. 12. Ferebee SH: Controlled chemoprophylaxis trials in tuberculosis: a general review. Adv Tuberc Res 1969; 17:28-106. 13. Comstock GW, Ferebee SH, Hammes LM: A controlled trial of community-wide isoniazid prophylaxis in Alaska. Am Rev Respir Dis 1967; 95:935-943. 14. Poweli KE, Brown ED, Seggerson JJ, et al:Evaluation of tuberculosis control programs: some national trends. Am J Public Health 1984; 74:344-348. 15. Primary resistance to antituberculous drugs-United States. MMWR 1983; 32:521-523.

IATPM Announces the Luther L. Terry Preventive Medicine Fellowshipl
The Association of Teachers of Preventive Medicine has announced a two-year senior level position for a physician to provide national leadership in the development of preventive medicine in primary care. The felloWship is sponsored by the Association of Teachers of Preventive Medicine (ATPM), in cooperation with the Society of Teachers of Family Medicine (STFM) and the Society for Research and Education in Primary Care Internal Medicine (SREPCIM). The Fellow will serve as senior technical and policy advisor to the Deputy Assistant Secretary for Health (Director, Disease Prevention and Health Promotion), US Public Health Service, in matters relating to prevention priorities and programs in clinical settings. The Fellow will also serve as a national liaison and counsel for initiatives in preventive medicine in primary care which major national professional organizations and the Office of Disease Prevention and Health Promotion (ODPHP) consider most important. In this role, the Fellow will work closely with staff and leadership of the ATPM, SREPCIM, and the STFM, and other national preventive medicine/primary care organizations. Qualifications are as follows: Required: MD degree with Board Certification in primary care and/or preventive medicine disciplines; current faculty appointment in a health sciences school at associate or full professor level (clinical or tenure-track) or an equivalent level position in a health organization; superior oral and written communication skills; demonstrated professional leadership skills, and interest in health policy. Desirable: additional postgraduate training such as MPH, MS, PhD, or fellowship; clinical experience; research and teaching experience in preventive medicine areas. The application deadline is January 15, 1986. Salary is $70,000 per annum with ample travel funds. The fellowship term will be two years and it is understood that the fellow will have assurance from his/ her parent institution that he/she may return to that original (or a comparable) position following the fellowship. The starting date is between June 1 and Sept. 1, 1986. For further information and application procedures about the Luther L. Terry Fellowship position, please contact:
Dennis J. Barbour, JD, Executive Director ATPM 1030 15th St., NW, Ste. 1020 Washington, DC 20005 (202) 682-1698


A0JPH January 1986, Vol. 76, No. 1

Sponsor Documents

Or use your account on


Forgot your password?

Or register your new account on


Lost your password? Please enter your email address. You will receive a link to create a new password.

Back to log-in