Paediatrica Indonesiana

VOLUME 51 NUMBER 6 November º 2O11
Original Article
332 º Paediatr Indones, Vol. 51, No. 6, November 2011
Limitations of the Indonesian Pediatric Tuberculosis
Scoring System in the context of child contact
investigation
Rina Triasih
1,2
, Stephen M Graham
2,3
Abstract
Background The Indonesian scoring system for childhood
tuberculosis (1B) was developed to address the problem of
overdiagnosis. The implementation of this scoring system
in children who have household TB contacts has not been
evaluated.
Objective To evaluate the performance of the Indonesian
pediatric tuberculosis scoring system in the context of contact
investigation.
Methods A cross-sectional studv was conducted in the Yo,vakarta
municipalitv, between Au,ust 2O1O and March 2O11. Subjects
were children under the age of 15 years, living in the same house
with an adult who had pulmonarv 1B. Subjects underwent
historv takin,, phvsical examinations, tuberculin skin tests (1S1)
and chest X-ravs. Sputum smear was performed in svmptomatic
children. We compared outcomes of the Indonesian pediatric TB
scoring system to that of rigorous clinical assessment.
Results A total of 116 eli,ible children of o2 source cases were
recruited into the studv. Sixtv-ei,ht (17') children had positive
1S1 tuberculin skin tests. Usin, the scorin, svstem, 17' of the
subjects were dia,nosed to have 1B disease, while onlv 1O'
were diagnosed with TB using rigorous clinical assessment. With
ri,orous clinical assessment, 1O' of the subjects were dia,nosed
as havin, latent 1B infection (l1Bl), while none of the subjects
were diagnosed as LTBI using the scoring system.
Conclusion The use of the Indonesian pediatric TB scoring system
in children with household TB contact may lead to overdiagnosis
of TB disease. [Paediatr Indones. 2011;51:332-7].
Keywords: tuberculosis, scoring system, children,
household contact
lrom the Department of Child Health, Medical School, Oadjah Mada
University Yogyakarta, Indonesia
1
, and Centre for International Child
Health, Department of Child Health, University of Melbourne
2
, and
Murdoch Children’s Research Institute
3
, Royal Children’s Hospital,
Melbourne, Victoria, Australia.
Reprint requests to: Rina 1riasih, Department of Child Health, Medical
School, Oadjah Mada Universitv , Jl.Kesehatan No. 1, Sekip, Yo,vakarta,
Indonesia. l-mail: rina_triasih@yahoo.com
E
stablishin, tuberculosis (1B) dia,nosis in
children is a challenge due to the lack of a
practical gold standard. A positive result by
sputum microscopy, the current acceptable
,old standard, has been observed in onlv 1O-15' of
children with probable TB.
1
Sputum cultures may
provide better vield of positive results (3O-1O').
1-3
Furthermore, collecting sputum in children is difficult,
limiting its use for diagnosing TB in children in clinical
practice. For this reason, TB diagnosis in children
has been commonly based on the presence of clinical
symptoms, such as failure to thrive or malnutrition,
peripheral lymphadenopathy and persistent cough,
combined with chest X-rav (CXR) and historv of
close contact with a TB patient. However, clinical
symptoms and radiological features of TB in children
often overlap with other childhood respiratory diseases,
particularly in settings where HIV infection and
malnutrition are endemic. Clinical case definitions are
Rina Triasih et al: limitations of lndonesian Pediatric 1uberculosis Scorin, Svstem
Paediatr Indones, Vol. 51, No. 6, November 2011 º 333
poorly validated and are likely to result in significant
under- or over-dia,nosis of childhood 1B.
4,5
The difficulties in establishing TB diagnoses
in children have led to development of several
diagnostic approaches, such as scoring systems and
diagnostic algorithms.
6
6 ln 2OOO, the lndonesian
Pediatric Societv introduced a Pediatric 1B Scorin,
System to address the problem of overdiagnosis of TB
in children, particularly in the private sector in the
large urban settings. Variables assessed in this scoring
system include contact history, typical symptoms,
tuberculin skin test (1S1) and CXR (Table 1). lt is
recommended that children with a total score of 6 or
more be considered to have TB disease. This scoring
was developed based on scoring systems from other
countries, and not based on Indonesian data. A
validation study on the Indonesian scoring system was
conducted, involvin, 1o1 children from 1 hospitals in
Jakarta. 1heir results show the sensitivitv to be 17',
specificitv 6o', and positive and ne,ative predictive
values to be 11' and 92', respectivelv (unpublished
data from the Division of Pediatric Respirolo,v, Cipto
Man,unkusumo Hospital). 1hese results indicate that
the use of this scoring system for diagnosing childhood
TB in Indonesia should be reconsidered.
Another important issue is the implementation
of the Indonesian scoring system in the context of
contact screening. Contact investigation is the process
of conducting an epidemiological investigation to
identify contacts of a tuberculosis case, to investigate
the presence of infection or tuberculosis disease, and
to provide appropriate treatment. The Indonesian
scoring system gives a score of 3 for a history of contact
with a patient having positive sputum smear and a
score of 3 for a positive TST result. Therefore, children
with household contact but have no symptoms,
normal CXR, and positive 1S1 will have a total score
of 6. These children may have LTBI, but based on
this scoring system, these asymptomatic children will
be considered to have TB disease.
This study was conducted to evaluate the
performance of the Indonesian TB scoring system
in children with household TB contact, in order to
improve the clinical management of these children.
Methods
1his cross-sectional studv was conducted in Yo,vakarta
between Au,ust 2O1O and April 2O11. 1he source
cases were new cases of pulmonary TB (either positive
or ne,ative sputum smears) between Januarv 2O1O
and April 2O11, who were identified from all Primarv
Health Centres, some private hospitals and lung
clinics in the Yogyakarta municipality, including the
lun, clinic at Sardjito Hospital, a teachin, hospital of
Oadjah Mada Universitv, Yo,vakarta. Children less
than 15 years of age who lived in the same house as
the source cases were recruited to the study. Those
receiving past or present TB treatment were excluded.
Written informed consent was obtained from parents
or guardians.
Table 1. The Indonesian Pediatric TB Scoring System
Variable 0 1 2 3
Household contact Unknown Contact with smear
negative TB patient
or unknown sputum
smear result
Contact with smear
positive TB patient
TST Negative Positive (> 10 mm, or in
immunocompromised
children > 5 mm)
Nutritional state BW/age < 80% Severe malnutrition
(BW/age < 60%)
Fever of unknown origin > 2 weeks Present
Cough > 3 weeks Present
Lymph node (cervical, axillary, inguinal)
enlargement
Multiple, non- tender,
diameter > 1cm
Joint swelling (knee, phalanges) Present
Chest X ray Normal Suggestive of TB
Rina Triasih et al: limitations of lndonesian Pediatric 1uberculosis Scorin, Svstem
334 º Paediatr Indones, Vol. 51, No. 6, November 2011
The demographic and clinical data of eligible
children was collected bv questionnaire. Phvsical
examination, 1S1 and CXR were performed to eli,ible
children. Children with TB symptoms underwent
induced sputum collection for Acid Fast Bacilli
(AlB) smear and culture. Su,,estive 1B svmptoms
included non-remittin, and chronic cou,h which
did not improve with antibiotics, prolonged fever of
unknown origin, weight loss or lack of weight gain in
the preceding 3 months or weight loss of more than
1O' (minimum 1 k,) over anv time interval despite at
least 2 weeks of confirmed adequate nutrition, fati,ue
(unexplained perceived decrease in plavfulness/activitv
since the onset of cou,hin, as reported bv parents/
care,iver), hemoptvsis (blood in the sputum, not
hematemesis or nose bleed), and ni,ht sweats (re,ular
sweatin, requirin, a drv set of ni,htclothes).
TST was performed by a trained nurse by
intradermal, volar injection of O.1 ml of 2 1U
(tuberculin units) of tuberculin purified protein
derivate (PPD) R1 23. 1S1 result was measured at
72 hours and considered positive if the induration was
> 1O mm, re,ardless of BCO vaccination, or > 5 mm
in severely malnourished or immunocompromised
subjects. CXR was performed in Sardjito Hospital
in antero-posterior and lateral views. Results were
independently assessed by one radiologist and one
pediatrician blinded to clinical information and each
other’s findings. In case of disagreement, consensus
will be made by a second pediatrician. Induced sputum
was done bv trained research nurses. After fastin, 2-3
hours, subjects were pretreated with 2OO u, salbutamol
via nebulizer, followed bv 5 ml of 6' sterile saline for
15 minutes. Chest percussion was performed over
the anterior and posterior chest walls. Sputum was
obtained by suctioning through the nasopharynx with
a sterile mucus extractor. Specimens were transported
directly to the laboratory for processing.
Outcomes of this investigation were assessed
using both the Indonesian scoring system and rigorous
clinical assessment. In rigorous clinical assessment,
children were diagnosed with TB disease if there was
at least one TB symptom and at least one from the
followin,: bacteriolo,ical or radiolo,ic confirmation
or definite extrapulmonary TB.
We used Epidata for data entry and STATA
software, release 1O (StataCorp, Colle,e Station,
1exas, USA) for all descriptive analvses.
Results
A total of 116 eli,ible children from o2 source cases
were recruited durin, the studv period. Most (72') of
the source cases had positive sputum smears. Subjects'
characteristics are presented in Table 2.
Subjects' total scores are presented in Figure
1. Sixtv-ei,ht (17') children had a total score of
6 or more. After further analysis of symptoms, we
found that most (o2') of these children (with score
> 6) were asvmptomatic and had l1Bl, 16' had 1B
disease, and onlv 2' had no evidence of 1B infection
or disease (Figure 2).
Figure 3 shows the comparison of outcomes/
diagnoses using the Indonesian scoring system versus
rigorous clinical assessment in children with household
TB contact. There was an obvious discrepancy in
determinin, the presence of 1B disease, as 17' of
children were diagnosed to have TB disease with
the scorin, svstem, while onlv 1O' were dia,nosed
to have TB disease by rigorous clinical assessment.
Furthermore, none of the children were diagnosed as
l1Bl with the scorin, svstem, compared to 1O' bv
rigorous clinical assessment.
Sixtv-ei,ht (17') of the subjects had a positive
1S1. Òf these, 56 (o2') had no svmptoms and were
Table 2. Subjects’ characteristics
Characteristic n = 146
Mean age, years (SD) 7.1 (4.7)
Females, n (%) 85 (58)
BCG vaccinated, n (%)
BCG scar, n (%)
141 (96)
116 (79)
Figure 1. Total scores (all subjects)
35
30
25
20
15
10
5
0
7
27
32
12
26
29
10
3
t
w
o
t
h
r
e
e
f
o
u
r
À
Y
H
s
i
x
V
H
Y
H
Q
e
i
g
h
t
Q
L
Q
H
Q
X
P
E
H
U

R
I

F
K
L
O
G
U
H
Q
WRWDOVFRUH
Rina Triasih et al: limitations of lndonesian Pediatric 1uberculosis Scorin, Svstem
Paediatr Indones, Vol. 51, No. 6, November 2011 º 335
household contact may lead to overdiagnosis of TB
disease. Close contact with an infectious case of TB,
particularly with an immediate household member,
may lead to TB infection in young children, identified
by a positive TST result. Therefore, a history of
household contact and positive TST are in a causal
pathway, and should not be weighted simultaneously
in the TB scoring system.
TST has been widely used to identify persons
infected with M. tuberculosis.
7
The triad of known
contact with an adult index case, positive TST,
and suggestive signs on chest radiograph has been
recommended by the International Standard for TB
Care, but its accuracy is reduced in endemic areas.
Since transmission in endemic areas is not restricted
to the household, the diagnostic values of both
documented household exposure and a positive TST
are limited.
o
diagnosed as having LTBI. The overall outcome of all
subjects in our investi,ation is presented in Table 3.
Discussion
We observed that almost half of children living in the
same house as an adult with pulmonary TB had TB
score of 6 or more. However, only a small proportion
of these children had TB disease, as confirmed
bv microbiolo,ical examination or clinical and/or
radiological features. Most of them were asymptomatic
and had LTBI. This finding indicates that the use of
the Indonesian scoring system for children with TB
Table 3. Outcomes of rigorous clinical assessment
Diagnosis n (%)
TB disease 15 (10)
LTBI 58 (40)
No evidence of TB infection or disease 73 (50)
Figure 2. Diagnosis (by rigorous clinical assess-
ment) of children with scores of > 6
No evidence of TB
LTBI
TB disease
o2'
16'
2'
Figure 3. Comparison of diagnoses between the
scoring system and rigorous clinical assessment in
subjects with household TB contact
60%
50%
40%
30%
20%
10%
0%
47%
10%
0%
40%
53%
50%
TB disease LTBI No evidence of TB
Scoring System Clinical assessment
Figure 4. Suggested approach for children in con-
tact with TB cases when chest radiograph and TST
are not readily available
10
Notes:
a
Also consider if the mother or primary caregiver has sputum
smear-ne,ative pulmonarv 1B
b
Svmptomatic: if 1B is suspected, refer to local ,uidelines
on diagnosis of childhood TB
c
Unless the child is HlV-infected (in this case, lNH 1O m,/
k, dailv for 6 months is indicated)
d
lNH 1O m,/k, dailv for 6 months
Children in close contact with a case of sputum smear-positive 1B
a
Less than 5 years More than 5 years
If becomes
symptomatic
e
If becomes
symptomatic
e
No Evaluate for TB disease Preventive therapv
d
Well
c
Well Symptomatic
b
Symptomatic
b
Rina Triasih et al: limitations of lndonesian Pediatric 1uberculosis Scorin, Svstem
336 º Paediatr Indones, Vol. 51, No. 6, November 2011
The requirement of TST for diagnosing TB in
children in Indonesia is also problematic, since the
test is not readily available in most health centers,
particularly in rural areas. With regard to contact
investi,ation, as lon, as 1S1 and CXR remain as
mandatory tests for screening children with household
contact, the pro,ram covera,e and the use of isoniazid
(lNH) preventive therapv in resource-limited settin,s
cannot be expected to improve.
The current WHO guideline advocates symp-
tom-based screenin, in resource-limited settin,s
(Figure 4). 1his approach applies to household
contacts of smear-positive pulmonarv 1B cases, but
screening should also be available for contacts of
smear-ne,ative pulmonarv 1B cases. 1his ,uide-
line recommends that clinical assessment alone is
sufficient to decide whether the contact is well or
asvmptomatic. 1S1 or CXR are not required in rou-
tine assessment of exposed contacts. Asymptomatic
contacts do not require additional tests to exclude ac-
tive 1B. lurther investi,ation and follow-up depend
on national policy and practice. WHO recommends
lNH 1O m,/k, dailv for 6 months for healthv contacts
a,ed less than 5 vears, and bimonthlv follow-up until
treatment is completed. If TB is suspected at the
initial assessment or at subsequent follow-up, further
investigation should be performed to establish or
exclude the TB diagnosis. Referral to a district or
tertiary hospital may be necessary when there are
uncertainties about the diagnosis. Contacts with TB
disease should be registered and treated according
to TB guidelines.
9
We also observed that almost 5O' of children
with TB household contact were latently infected,
with onlv 1O' havin, the disease. Similarlv, previous
TB contact studies reported the proportion of children
with LTBI to be higher than that with TB disease.
Among children aged less than 15 years, the proportion
of l1Bl ran,ed from 25' to 69', whereas 1B disease
ran,ed from 1' to 1O'.1O-17 The discrepancy
between the studies is due to different tuberculin
solutions used for TST and different definitions of
positive 1S1 as an indicator of 1B infection. Salazar-
Ver,ara and collea,ues in the Philippines in 2OO3 used
the standardized, recommended dose of O.1 ml of 2
1U of PPD R1 23 with 1ween oO (Statens Serum
lnstitute, Copenha,en, Denmark). Usin, a low cutoff
point for positive 1S1 (more than 5 mm induration),
the l1Bl proportion of as hi,h as 69' mav be an
overestimation of the infection rate among contacts.
15
In studies which defined TB infection as positive TST
of > 1O mm induration, the proportion of children
with TB contact aged less than 5 years who were
infected ran,ed from 7.o' to 33.o'.
11,11,16,1o-23
. A
meta-analvsis bv Morrison and collea,ues revealed the
pooled yield of TB disease among children aged less
than 15 years in household TB contact in developing
countries to be 7.O' (95' Cl 6.O to o.O), whereas the
proportion of l1Bl was hi,her at 1O.1' (95' Cl 3o.7
to 12.2').
21
1hese results, however, are also subject
to substantial heterogeneity.
In conclusion, contact investigation of TB in
children is of great value, not only to identify infected
or diseased children, but also for TB control in the
community as a whole. However, using the Indonesian
scoring system for this purpose leads to overdiagnosis.
A svmptom-based screenin, recommended bv WHÒ
provides a simple approach for contact investigation
in children. Implementation of this approach may be
warranted.
References
1. zar HJ, Hanslo D, Apolles P, Swin,ler O, Hussev O. lnduced
sputum versus gastric lavage for microbiological confirmation
of pulmonarv tuberculosis in infants and voun, children: a
prospective studv. lancet. 2OO5,365:13O-1.
2. Starke JR. Pediatric tuberculosis: time for a new approach.
1uberculosis. 2OO3,o3:2Oo-12.
3. Marais BJ, Hesselin, AC, Oie RP, Schaaf HS, lnarson
DA, Beyers N. The bacteriologic yield in children with
intrathoracic tuberculosis. Clin lnfect Dis. 2OO6,12:69-71.
4. Brent AJ, Anderson ST, Kampmann B. Childhood
tuberculosis: out of si,ht, out of mind' 1rans R Soc 1rop
Med Hv,. 2OOo,1O2:217-o.
5. Nelson LJ, Wells CD. Global epidemiology of childhood
tuberculosis. lnt J 1uberc lun, Dis. 2OO1,o:636-17.
6. Hesselin, AC, Schaaf HS, Oie RP, Starke JR, Bevers N.
A critical review of diagnostic approaches used in the
diagnosis of childhood tuberculosis. Int J Tuberc Lung Dis.
2OO2,6:1O3o-15.
7. Coulter JB. Diagnosis of pulmonary tuberculosis in young
children. Ann 1rop Paediatr. 2OOo,2o:3-12.
o. Marais BJ, Van zvl S, Schaaf HS, Van Aardt M, Oie RP,
Bevers N. Adherence to isoniazid preventive chemotherapv:
Rina Triasih et al: limitations of lndonesian Pediatric 1uberculosis Scorin, Svstem
Paediatr Indones, Vol. 51, No. 6, November 2011 º 337
a prospective community based study. Arch Dis Child.
2OO6,91:762-5.
9. Stop 1B Partnership Childhood 1B Sub,roup. Chapter 1:
Childhood contact screenin, and mana,ement |Òfficial
Statement. Ouidance for National 1uberculosis Pro,rammes
on the management of tuberculosis in children. Chapter 4 in
the series]. lnt J 1uberc lun, Dis. 2OO7,11:12-5.
1O. Andrews RH, Devadata S, lox W, Radhakrishna S,
Ramakrishnan CV, Velu S. Prevalence of tuberculosis amon,
close family contacts of tuberculous patients in South India,
and influence of segregation of the patient on early attack
rate. Bull World Health Òr,. 196O,23:163-51O.
11. Almeida lMDMM, Barbieri MAMP, Da Paixao ACMP,
Cuevas LEMDM. Use of purified protein derivative to assess
the risk of infection in children in close contact with adults
with tuberculosis in a population with hi,h Calmette-Ouerin
bacillus covera,e. Paediatr lnfect Dis J. 2OO1,2O:1O61-5.
12. Oessner BD, Weiss NS, Nolan CM. Risk factors for pediatric
tuberculosis infection and disease after household exposure
to adult index cases in Alaska. J Pediatr. 199o,132:5O9-13.
13. Narain R, Nair SS, Rao OR, Chandrasekhar P. Distribution of
tuberculous infection and disease among households in a rural
communitv. Bull World Health Òr,an. 1966,31:639-51.
11. Rathi SK, Akhtar S, Rahbar MH, Azam Sl. Prevalence and
risk factors associated with tuberculin skin test positivity
amon, household contacts of smear-positive pulmonarv
tuberculosis cases in Umerkot, Pakistan. lnt J 1uberc lun,
Dis. 2OO2,6:o51-7.
15. Salazar-Ver,ara RMl, Sia lO, 1upasi 1l, Alcares M, Òrillaza
RB, Co V, et al. Tuberculosis infection and disease in children
livin, in households of lilipino patients with tuberculosis: a
preliminarv report. lnt J 1uberc lun, Dis. 2OO3,7:S191-5OO.
16. 1eixeira l, Perkins MD, Johnson Jl, Keller R, Palaci M, Do
Valle Dettoni V, et al. Infection and disease among household
contacts of patients with multidru,-resistant tuberculosis. lnt
J 1uberc lun, Dis. 2OO1,5:321-o.
17. Tornee S, Kaewkungwal J, Fungladda W, Silachamroon
U, Akarasewi P, Sunakorn P. Risk factors for tuberculosis
i nfecti on among househol d contacts i n Bangkok,
1hailand. Southeast Asian J 1rop Med Public Health.
2OO1,35:375-o3.
1o. Ouwatudde D, Nakakeeto M, Jones-lopez lC, Ma,anda A,
Chiunda A, Mugerwa RD, et al. Tuberculosis in household
contacts of infectious cases in Kampala, Uganda. Am J
lpidemiol. 2OO3,15o:oo7-9o.
19. lin X, Chon,suvivatwon, V, lin l, Oeater A, lijuan R.
Dose-response relationship between treatment delav of
smear-positive tuberculosis patients and intra-household
transmission: a cross-sectional studv. 1rans R Soc 1rop Med
Hv,. 2OOo,1O2:797-oO1.
2O. Maciel ll, Vieira lW, Molina lP, Alves R, Do Prado 1N,
Dietze R. Juvenile household contacts a,ed 15 or voun,er
of patients with pulmonary TB in the greater metropolitan
area of Vitoria, Brazil: a cohort studv. J Bras Pneumol.
2OO9,35:359-66.
21. Sinfield R, Nvirenda M, Haves S, Molvneux lM, Oraham
SM. Risk factors for TB infection and disease in young
childhood contacts in Malawi. Ann 1rop Paed: lnt Child
Health. 2OO6,26:2O5-13.
22. Sin,h M, Mvnak Ml, Kumar l, Mathew Jl, Jindal SK.
Prevalence and risk factors for transmission of infection
among children in household contact with adults having
pulmonarv tuberculosis. Arch Dis Child 2OO5,9O:621-o.
23. lskandar H, Nataprawira HMD, Oarna H, Djais J1B.
Tuberculosis prevalence among under five children in
household contact with negative acid fast bacili adult
pulmonarv tuberculosis. Paediatr lndones. 2OOo,1o:1o-22.
21. Morrison J, Pai M, Hopewell PC. 1uberculosis ad latent
tuberculosis infection in close contact of people with
pulmonrv tuberculosis in low-income and middle-income
countries: a svstematic review an meta-analvsis. lancet lnfect
Dis. 2OOo,o:359-6o.

Therefore.tender.Rina Triasih et al: poorly validated and are likely to result in significant 4. and Table 1. particularly in the private sector in the large urban settings. some private hospitals and lung clinics in the Yogyakarta municipality. The Indonesian scoring system gives a score of 3 for a history of contact with a patient having positive sputum smear and a score of 3 for a positive TST result. in order to improve the clinical management of these children. November 2011 333 . This study was conducted to evaluate the performance of the Indonesian TB scoring system in children with household TB contact. non. Those receiving past or present TB treatment were excluded. Methods cases were new cases of pulmonary TB (either positive the use of this scoring system for diagnosing childhood TB in Indonesia should be reconsidered. No. phalanges) Chest X ray Normal Paediatr Indones. Contact investigation is the process of conducting an epidemiological investigation to identify contacts of a tuberculosis case. typical symptoms. Table 1 recommended that children with a total score of 6 or more be considered to have TB disease. Vol. Another important issue is the implementation of the Indonesian scoring system in the context of contact screening. Written informed consent was obtained from parents or guardians. including the than 15 years of age who lived in the same house as the source cases were recruited to the study. 1 2 Contact with smear negative TB patient or unknown sputum smear result 3 Contact with smear positive TB patient TST Negative Positive (> 10 mm. The Indonesian Pediatric TB Scoring System Variable Household contact 0 Unknown Health Centres. This scoring was developed based on scoring systems from other countries. 6. to investigate the presence of infection or tuberculosis disease. children with household contact but have no symptoms. or in immunocompromised children > 5 mm) BW/age < 80% Present Present Multiple. diameter > 1cm Present Suggestive of TB Severe malnutrition (BW/age < 60%) Nutritional state Fever of unknown origin > 2 weeks Cough > 3 weeks Lymph node (cervical. of 6. A validation study on the Indonesian scoring system was to provide appropriate treatment. Variables assessed in this scoring system include contact history. and not based on Indonesian data.6 System to address the problem of overdiagnosis of TB in children. 51. These children may have LTBI. inguinal) enlargement Joint swelling (knee. but based on this scoring system. these asymptomatic children will be considered to have TB disease.5 The difficulties in establishing TB diagnoses in children have led to development of several diagnostic approaches. axillary. such as scoring systems and diagnostic algorithms.

November 2011 ei gh re t . children were diagnosed with TB disease if there was at least one TB symptom and at least one from the or definite extrapulmonary TB. Induced sputum 7 12 10 3 ur six o e tw th fo 15 minutes. did not improve with antibiotics.1 (4. we Table 2. n (%) BCG scar. Sputum was obtained by suctioning through the nasopharynx with a sterile mucus extractor. 51. years (SD) Females. n (%) BCG vaccinated. Vol. After further analysis of symptoms. Children with TB symptoms underwent induced sputum collection for Acid Fast Bacilli characteristics are presented in Table 2. none of the children were diagnosed as rigorous clinical assessment. Subjects’ characteristics Characteristic Mean age.Rina Triasih et al: The demographic and clinical data of eligible Results children. consensus will be made by a second pediatrician. Outcomes of this investigation were assessed using both the Indonesian scoring system and rigorous clinical assessment. In rigorous clinical assessment. n (%) n = 146 7. There was an obvious discrepancy in children were diagnosed to have TB disease with to have TB disease by rigorous clinical assessment. Total scores (all subjects) Figure 3 diagnoses using the Indonesian scoring system versus rigorous clinical assessment in children with household TB contact. We used Epidata for data entry and STATA Figure 1. 6. 334 Paediatr Indones. No.7) 85 (58) 141 (96) 116 (79) TST was performed by a trained nurse by > or disease (Figure 2 > > 5 mm in severely malnourished or immunocompromised 35 30 25 20 15 10 5 0 32 27 26 29 independently assessed by one radiologist and one pediatrician blinded to clinical information and each other’s findings. weight loss or lack of weight gain in the preceding 3 months or weight loss of more than Figure 1 6 or more. Furthermore. Specimens were transported directly to the laboratory for processing. Chest percussion was performed over the anterior and posterior chest walls. In case of disagreement. prolonged fever of unknown origin.

51. 6. Close contact with an infectious case of TB. may lead to TB infection in young children. tuberculosis. Paediatr Indones. Diagnosis (by rigorous clinical assessment) of children with scores of > 6 60% 50% 40% 30% 20% 10% 0% TB disease Scoring System 10% 0% LTBI No evidence of TB d b 53% 47% 40% If becomes symptomatice 50% If becomes symptomatice Notes: a Also consider if the mother or primary caregiver has sputum on diagnosis of childhood TB c Clinical assessment Figure 3. but its accuracy is reduced in endemic areas. particularly with an immediate household member. No.7 The triad of known contact with an adult index case. TST has been widely used to identify persons infected with M. The overall outcome of all Table 3. the diagnostic values of both documented household exposure and a positive TST are limited. as confirmed radiological features. only a small proportion of these children had TB disease. Therefore. Suggested approach for children in contact with TB cases when chest radiograph and TST are not readily available10 diagnosed as having LTBI. and suggestive signs on chest radiograph has been recommended by the International Standard for TB Care. Vol. Table 3. This finding indicates that the use of the Indonesian scoring system for children with TB household contact may lead to overdiagnosis of TB disease. and should not be weighted simultaneously in the TB scoring system. Since transmission in endemic areas is not restricted to the household. a history of household contact and positive TST are in a causal pathway. November 2011 335 . Outcomes of rigorous clinical assessment Diagnosis TB disease LTBI No evidence of TB infection or disease n (%) 15 (10) 58 (40) 73 (50) Discussion We observed that almost half of children living in the same house as an adult with pulmonary TB had TB score of 6 or more. However. Most of them were asymptomatic and had LTBI. positive TST. Comparison of diagnoses between the scoring system and rigorous clinical assessment in subjects with household TB contact Figure 4. identified by a positive TST result.Rina Triasih et al: a No evidence of TB LTBI TB disease Less than 5 years More than 5 years Well Symptomaticb Symptomaticb Wellc d Evaluate for TB disease No Figure 2.

If TB is suspected at the investigation should be performed to establish or exclude the TB diagnosis. Contacts with TB disease should be registered and treated according to TB guidelines. contact investigation of TB in children is of great value. Childhood 5. with TB household contact were latently infected. A pooled yield of TB disease among children aged less than 15 years in household TB contact in developing cannot be expected to improve. WHO recommends to substantial heterogeneity. Among children aged less than 15 years. particularly in rural areas. Implementation of this approach may be warranted. the proportion The discrepancy between the studies is due to different tuberculin solutions used for TST and different definitions of 4. 51. Anderson ST. In conclusion. Global epidemiology of childhood A critical review of diagnostic approaches used in the diagnosis of childhood tuberculosis. since the test is not readily available in most health centers. References sputum versus gastric lavage for microbiological confirmation DA. 7. November 2011 . The current WHO guideline advocates symp (Figure 4 screening should also be available for contacts of line recommends that clinical assessment alone is sufficient to decide whether the contact is well or tine assessment of exposed contacts. not only to identify infected or diseased children. Beyers N. With regard to contact mandatory tests for screening children with household overestimation of the infection rate among contacts. TB contact studies reported the proportion of children with LTBI to be higher than that with TB disease. Kampmann B.15 In studies which defined TB infection as positive TST of > with TB contact aged less than 5 years who were . Coulter JB. No. 6. Wells CD. The bacteriologic yield in children with Brent AJ. but also for TB control in the community as a whole. Int J Tuberc Lung Dis. Nelson LJ. using the Indonesian scoring system for this purpose leads to overdiagnosis. Diagnosis of pulmonary tuberculosis in young 336 Paediatr Indones. Referral to a district or tertiary hospital may be necessary when there are uncertainties about the diagnosis. treatment is completed. However.Rina Triasih et al: The requirement of TST for diagnosing TB in children in Indonesia is also problematic. provides a simple approach for contact investigation in children. Asymptomatic contacts do not require additional tests to exclude ac on national policy and practice. Vol.

Mugerwa RD. 17. Tuberculosis infection and disease in children Tuberculosis prevalence among under five children in household contact with negative acid fast bacili adult tuberculosis infection in close contact of people with Valle Dettoni V. Infection and disease among household Paediatr Indones. Am J close family contacts of tuberculous patients in South India. Kaewkungwal J. Vol. Risk factors for TB infection and disease in young tuberculous infection and disease among households in a rural risk factors associated with tuberculin skin test positivity among children in household contact with adults having RB. Silachamroon infection among household contacts in Bangkok. et al. Use of purified protein derivative to assess the risk of infection in children in close contact with adults of patients with pulmonary TB in the greater metropolitan tuberculosis infection and disease after household exposure SM. on the management of tuberculosis in children. No. Tornee S. Fungladda W. 51. Chapter 4 in Chiunda A. et al. November 2011 337 . and influence of segregation of the patient on early attack Cuevas LEMDM. et al. 6. Arch Dis Child. Uganda. Tuberculosis in household contacts of infectious cases in Kampala. Co V.Rina Triasih et al: a prospective community based study.