Wabah TBC Di India

Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

Prevalence of Pulmonary Tuberculosis among Adults in a

Rural Sub-District of South India


Vineet K. Chadha1*, Prahlad Kumar2, Sharada M. Anjinappa1, Sanjay Singh1, Somashekar Narasimhaiah3,
Malathi V. Joshi3, Joydev Gupta1, Lakshminarayana1, Jitendra Ramchandra4, Magesh Velu1,
Suganthi Papkianathan1, Suseendra Babu3, Hemalatha Krishna5
1 Epidemiology and Research Division, National Tuberculosis Institute, Bangalore, Karnataka, India, 2 Office of the Director, National Tuberculosis Institute, Bangalore,
Karnataka, India, 3 Monitoring and Evaluation, National Tuberculosis Institute, Bangalore, Karnataka, India, 4 Human Resource Development, National Tuberculosis
Institute, Bangalore, Karnataka, India, 5 Laboratory, National Tuberculosis Institute, Bangalore, Karnataka, India

Abstract
Background: We conducted a survey to estimate point prevalence of bacteriologically positive pulmonary TB (PTB) in a
rural area in South India, implementing TB program DOTS strategy since 2002.
Methods: Survey was conducted among persons $15 years of age in fifteen clusters selected by simple random sampling;
each consisting of 512 villages. Persons having symptoms suggestive of PTB or history of anti-TB treatment (ATT) were
eligible for sputum examination by smear microscopy for Acid Fast Bacilli and culture for Mycobacterium tuberculosis; two
sputum samples were collected from each eligible person. Persons with one or both sputum specimen positive on
microscopy and/or culture were labeled suffering from PTB. Prevalence was estimated after imputing missing values to
correct for bias introduced by incompleteness of data. In six clusters, registered persons were also screened by X-ray chest.
Persons with any abnormal shadow on X-ray were eligible for sputum examination in addition to those with symptoms and
ATT. Multiplication factor calculated as ratio of prevalence while using both screening tools to prevalence using symptoms
screening alone was applied to entire study population to estimate prevalence corrected for non-screening by X-ray.
Results: Of 71,874 residents $15 years of age, 63,362 (88.2%) were screened for symptoms and ATT. Of them, 5120 (8.1%) 4681 (7.4%) with symptoms and an additional 439 (0.7%) with ATT were eligible for sputum examination. Spot specimen
were collected from 4850 (94.7%) and early morning sputum specimens from 4719 (92.2%). Using symptom screening
alone, prevalence of smear, culture and bacteriologically positive PTB in persons $15 years of age was 83 (CI: 57109), 152
(CI: 108197) and 196 (CI :145246) per 100,000 population respectively. Prevalence corrected for non-screening by X-ray
was 108 (CI: 82134), 198 (CI: 153243) and 254 (CI: 204301) respectively.
Conclusion: Observed prevalence suggests further strengthening of TB control program.
Citation: Chadha VK, Kumar P, Anjinappa SM, Singh S, Narasimhaiah S, et al. (2012) Prevalence of Pulmonary Tuberculosis among Adults in a Rural Sub-District of
South India. PLoS ONE 7(8): e42625. doi:10.1371/journal.pone.0042625
Editor: Madhukar Pai, McGill University, Canada
Received May 1, 2012; Accepted July 10, 2012; Published August 15, 2012
Copyright: 2012 Chadha et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: Survey was undertaken with available human and capital resources already available in the institute, supplemented by funding of 11700000 Indian
Rupees from World Health Organization (WHO) for supplies and additional manpower. Website of WHO India is whoindia.org/en. The funders had no role in study
design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: vineet2.chadha@gmail.com

a decline of about 50% in prevalence of pulmonary tuberculosis


(PTB) was observed from 1999 to 2006 in a sub-division of
Thiruvallur district, Tamil Nadu state [5]. Since, this observation
pertained to a single geographical site, Government of India
identified six institutions to carry out independent surveys in seven
other sites (districts/sub-division of a district) located in different
geographical regions using a generic protocol. The objective of the
survey at each site was to find out the point prevalence of
bacteriologically positive PTB among adults, which would provide
a baseline data to measure the future trends in different parts of
the country by repeat surveys. The personnel to be involved at
these sites were trained in actual field conditions at a common site
by a team of vastly experienced trainers. One of these surveyed

Introduction
National Tuberculosis Programme (NTP) based on a cost
effective operational strategy was implemented all over India from
1962; after a nation-wide survey during 19551958 revealed that
tuberculosis (TB) was highly prevalent throughout the country [1].
Surveys carried out thereafter in geographically defined areas
revealed that the prevalence of TB continued to be high, though
varied, in different parts of the country [2]. Taking cues from a
review of NTP, Revised National tuberculosis Control Programme
(RNTCP) adopting DOTS (an internationally recommended
strategy for TB control) was launched in 1997 and expanded in
phases to cover the entire population by 2006 [3]. Implementation
of RNTCP lead to improvements in case detection and high
treatment success rates in most parts of the country [3,4]. Indeed,
PLOS ONE | www.plosone.org

August 2012 | Volume 7 | Issue 8 | e42625

Prevalence of TB in a South Indian District

sites was the Nelamangala sub-division in Bangalore rural district,


the results of which are reported hereunder.

Data collection instruments

Materials and Methods

i.
ii.

The following schedules were useful for data collection:-

Study site and setting


Nelamangala is one of the four sub-divisions in Bangalore rural
district. This district is located adjacent to Bangalore city which is
the capital of Karnataka state in southern India. DOTS strategy is
being implemented in the entire district since 2002. Nelamangala
was chosen for this survey so that the trends from the past could
also be elicited since a similar survey had been carried out there
during the year 1975 [6].
Majority of adults in Nelamangala are engaged in agricultural
occupations and a proportion work in factories located in a
bordering industrial area in Bangalore city.

iii.
iv.

Field Procedures
Field work was carried out during October 2008June 2010.
A planning visit was made to each Panchayat office and each
village to familiarize the officials, village leaders and the
community with the purpose and procedures of the survey and
to seek their cooperation. In each village, a rough sketch of lanes
and hamlets showing approximate number of houses in each was
drawn, after going around and in discussions with village leaders.
Survey in each village began on a mutually agreed date.
Enumerators went to each household and recorded the age, sex
and resident status of each individual. Each eligible person (15
years or more in age and residing for $6 months in the household
or in any other village in Nelamangala sub- division) was registered
into an individual card. Subsequently, a symptom elicitor queried
each eligible individual for presence of symptoms suggestive of
PTB (persistent cough for $2 weeks, fever or chest pain for $1
month, presence of blood in sputum any time during last 6
months) and history of ATT. Field supervisors re-interviewed 10%
of eligible individuals, as a quality control mechanism.
Individuals having pulmonary symptoms suggestive of PTB or a
positive history of ATT were eligible for sputum examination. A
spot sputum specimen was collected at a temporarily setup sputum
collection centre within the village, after briefing by the laboratory
technician (LT) on how to bring out good sputum sample and spit
into a pre-numbered sterilized screw capped sputum cup. After a
satisfactory extraction of spot specimen, a pre-numbered empty
bottle was given for collecting another sample next morning.
Sputum containing bottles marked with ID of each patient were
transported in a sputum box on day of collection to the accredited
laboratory of the institute.
In six of the clusters, each registered person was also screened
by X-ray. A 70 mm photofluorography film of the chest was taken
at the temporarily setup centre using a mobile mass miniature
radiography (MMR unit); pregnant women and bed ridden
individuals were excluded (screening by X-ray could not be
undertaken in other clusters due to breakdown of equipment
during the middle period of the survey). After processing of
exposed MMR rolls in X-ray laboratory of the Institute, each film
was read and classified by two trained independent X-ray readers
into one of the categories - N (normal)/TI (technically inadequate)/A (lung pathology other than tuberculosis)/B (tuberculosis
inactive)/C (tuberculosis active). Persons with their films labeled as
TI/A/B/C by any of the two X-ray readers were eligible for
sputum examination, in addition to those having symptoms
suggestive of PTB or history of ATT.

Study population
Persons $15 years of age residing in Nelamangala sub-division
for $6 months comprised the study population.

Sampling
Simple random sampling (SRS) was adopted for selection of
clusters. Each cluster corresponded to a group of villages in the
jurisdiction of a Gram Panchayat a local self government
consisting of elected members of the community and having
responsibility to implement rural development programs. Office
bearers of the Panchayats visited during planning phase of the
study insisted for political reasons to cover all the villages in their
jurisdiction. There were altogether 25 panchyats in Nelamangala
sub-division.
Since the exact count of adult population in individual clusters
was not known, fifteen (arbitrarily decided number) clusters were
selected and the survey was planned to start in the first selected
cluster covering all eligible persons and proceed to subsequent
clusters in order of their selection till the required sample size was
achieved.

Sample size
Sample size was originally calculated at 47,828 to estimate the
prevalence within 20% of the true value at 5% level of significance
considering a design effect of 2 to account for cluster sampling.
Expected prevalence of bacteriologically positive PTB (positive for
AFB on microscopy and/or culture) while using both screening
tools symptom screening and chest X-ray, was arbitrarily
considered at 400 per 100,000 population. During the earlier
survey in the area in 1975, prevalence of culture positive PTB in
the age group of $15 years while using both screening tools was
480 per 100,000 population [6]; prevalence of bacteriologically
positive PTB was not reported.
Since the mobile X-ray equipment broke down during the
course of the survey, sample size was revised during the course of
the survey to 68,400 for an expected prevalence of 280 per
100,000 population assuming that 30% of PTB cases would be
missed due to non-screening by X-ray. This assumption was based
from the data of the previous survey in the same area [6].

Key variables
Key study variables were: age, sex, presence of pulmonary
symptoms, history of anti-TB treatment (ATT), presence of
abnormal shadow on X-ray, result of sputum smear microscopy
spot & early morning specimen and result of culture - spot &
early morning specimen.
PLOS ONE | www.plosone.org

Household form: to enlist all residents by age and sex


Individual card, for each eligible person: to record study
identification number (ID), age, sex, symptom status, X-ray
film number, date of sputum collection, results of sputum
smear and culture examination
Laboratory register to record sputum specimen numbers and
results of smear and culture examination
X-ray result form

Laboratory procedures
Sputum specimens were subjected to smear microscopy for acid
fast bacilli (AFB) and culture for M. tuberculosis following standard

August 2012 | Volume 7 | Issue 8 | e42625

Prevalence of TB in a South Indian District

Culture Positive case: An individual with at least one sputum


specimen found to be positive for M. tuberculosis on culture,
irrespective of smear result.
Bacteriologically positive case: An individual with at least one
sputum specimen found to be positive for AFB on microscopy
and/or M. tuberculosis on culture.

laboratory procedures, at a national reference laboratory located


in National Tuberculosis Institute, Bangalore (NTI) [7,8].
Two direct smears were made from each specimen on new
labeled slides under aseptic conditions in a biosafety cabinet.
Each smear was stained using 0.1% auramine-O and 0.5%
potassium permanganate and examined under a fluorescence
microscope at a magnification of 2006.
After taking out the amount required for making smears, each
specimen was homogenized and transferred to a McCartney
bottle. For decontamination, 4% sodium hydroxide was added in a
volume twice that of sputum specimen (Modified Petroffs Method)
and incubated in a shaker for 20 minutes. Sterile distilled water
was then added up to the neck of the bottle and centrifuged at
3000 rpm (revolutions per minute) for 15 minutes. The supernatant was decanted and the deposit was inoculated onto 2 slopes of
Lowenstein-Jensen (LJ) medium. Cultures were incubated at 37uC
and examined for the presence of mycobacterial colonies every
week for 8 weeks. Culture was discarded in case of contamination
or no growth at 8 weeks. The growth if observed was subjected to
Niacin test and incubation on LJ medium containing p-nitro
benzoic acid (PNB) in a concentration of 500 mg/ml. It was
labeled as M. tuberculosis, if Niacin test was positive and no growth
was observed on PNB containing medium.
Quality assurance of sputum microscopy and culture was
undertaken using the existing mechanism of RNTCP [8].

Statistical methods
Digitized data was verified and analyzed using SPSS version
17.0 and STATA version 12.0.
Point prevalence was estimated by three different approaches
[9]:
i)

ii)

Study investigators
iii)

Survey was carried out under the overall supervision of the


Senior Epidemiologist in the institute. Enumeration of residents in
the households, screening by interview, sputum collection and data
entry was undertaken by trained contractual field staff and
monitored by experienced institutional field staff. Subjecting the
study subjects to MMR, reading of X-ray films, examination of
sputum by smear and culture and analysis was undertaken by
experienced staff of the institute.

Ethical considerations
Survey was approved by the Institutional Ethics Committee.
Written consent for participation was sought from each
individual, after explaining procedures of the survey and its
benefits to the individuals and community, through material
printed in local language and also personally by field staff. No one
was compelled to participate.
TB cases detected during the survey were counseled and
facilitated to initiate ATT at the nearest RNTCP facility. For the
patients who failed to reach the RNTCP facility, repeated visits
(upto a maximum of nine) were made to their households by the
field staff to persuade them and also some times to accompany
them to RNTCP facility in order to initiate ATT. Follow up
during treatment was the responsibility of routine staff of RNTCP.
Individuals with symptoms but not having TB were advised to
seek treatment at the local health center; however no active
intervention was undertaken by survey team for referral or
otherwise.
Village leaders were informed of the findings of the study
emphasizing their role in raising awareness regarding symptoms of
TB and availability of quality services under RNTCP.

In six clusters where both screening tools were used, individual


level analysis was also performed separately with MMR reading as
an additional explanatory variable after imputing its missing
values. A multiplication factor was calculated as the ratio of
prevalence while using both screening tools to the prevalence using

Definitions
Smear Positive case: An individual with at least one sputum
specimen found to be positive for AFB on smear microscopy,
irrespective of culture result.

PLOS ONE | www.plosone.org

Crude prevalence (P). It was calculated as the total


number of smear, culture or bacteriologically positive cases
detected during the survey divided by the total number (n) of
individuals who were screened by interview and whose both
sputum results were available. Standard error (SE) was
estimated as Standard deviation (SD)/!n where SD = !P
(12P). Confidence intervals (95%) of the estimates were
calculated as the mean of the binomial exact 6 2SE.
Cluster level analysis. Average cluster level prevalence
was calculated as Pcluster level = gpi/c, where pi is the crude
prevalence in ith cluster and c is the number of clusters. SE
was calculated as SD/!c where SD = ![g(pi2Pcluster level)2/
c]. Confidence intervals (95%) were calculated as the mean
prevalence across clusters 6 2SE.
Individual level analysis. To correct for bias introduced
by incompleteness of data, individual level analysis was
performed using logistic regression model with robust
standard error. To include all eligible individuals in analysis,
missing value imputation was undertaken for individuals: (a)
without symptom screening (b) symptoms present but the
result of one or both sputum specimen not available. The
method accounted for between-cluster variability and
uncertainty in estimating SE, under the assumption that
within groups of individuals belonging to same sex, agegroup and TB symptoms, data are missing at random. For
imputation of missing values for each variable, starting values
were assigned to missing data, which in turn was obtained
from a random sample of values from individuals whose data
were available. Model was then fitted with this particular
variable as outcome variable and other variables as
explanatory variables. This was done sequentially, in order
of the proportion of data that were missing starting with
variables with least missing data. Finally, a logistic regression
model with smear/culture/bacteriologically confirmed TB as
the outcome variable and sex, age-group and TB symptoms
as explanatory variables were fitted. Newly imputed values
were used as starting values for next iteration of the process
which was undertaken in ten cycles in order to obtain one
imputed data set. Five such data sets were imputed and the
average of their prevalence was considered as final
prevalence. This model takes account of clustering in the
survey design, variation in the number of individuals per
cluster and the uncertainty introduced by imputation of
missing values when estimating the 95% CI for prevalence of
PTB [9].

August 2012 | Volume 7 | Issue 8 | e42625

Prevalence of TB in a South Indian District

symptoms screening alone. This factor was applied to the entire


study population to estimate prevalence corrected for nonscreening by MMR.
Chi-square test with continuity correction was used to test the
significance of differences between proportions and p-values,0.05
were considered significant.

Table 2. Frequency distribution of persons with different


pulmonary symptoms.

Study participants

Cough with other symptoms

1524 (32.6)

A total of 71,874 individuals across 158 villages in 15 clusters


were eligible for inclusion in the survey. Of them, 63,362 (88.2%)
were screened by interview; 8512 (12%) could not be screened due
to not being available despite repeated visits to their households.
Non-participation rates were higher in males compared to females
in all age groups, difference being more predominant in 1545
years (table 1).
Size of eligible population varied from 3620 to 7473 in different
clusters. Male to female ratio was about 1:1 in each cluster.

Chest pain alone

544 (11.6)

Symptoms

No.

Cough alone

2365 (50.5)

Fever alone

47 (1.0)

Haemoptysis alone

91 (1.9)

Combination of symptoms without


cough

110 (2.3)

Total

4681 (100.0)

(): percentages out of total numbers with symptoms.


doi:10.1371/journal.pone.0042625.t002

Overnight specimens were collected from 4719 (92.2%), of which


42 were smear positive and 61 were culture positive. Overall, 51
persons were smear positive and 86 culture positive, on spot and/
or overnight specimen. About 1% of spot and 1.8% of overnight
specimen were contaminated.
There were a total of 110 bacteriologically positive patients. Of
them, 101 were picked up through screening for presence of
symptoms and an additional nine through screening for history of
ATT. Among those with symptoms, 89 (88.1%) had cough with or
without other symptoms while 12 (11.9%) had one or more
symptoms other than cough.
Estimated point prevalence of smear positive, culture positive
and bacteriologically positive PTB estimated by three analytical
methods (using screening by interview for presence of symptoms) is
presented at Table 3. Considering the estimates by individual level
analysis as the best estimates, prevalence of smear positive, culture
positive and bacteriologically positive PTB was 83 (57109), 152
(108197) and 196 (145246) per 100,000 population respectively.

Results
I. Prevalence of PTB based on symptom screening
followed by sputum examination
Of the persons interviewed, 4681 (7.4%) were found to have
symptoms. Of them, about 83% had cough of 2 weeks or moreeither alone or in combination with one or more of the other
symptoms- chest pain, fever, haemoptysis. The remaining 17%
had one or more of the other symptoms (table 2).
Previous history of ATT was present in 653 (1.0%) persons and
56 (0.09%) were currently on ATT. Of the total of 709 persons
with history of ATT, 270 had symptoms at the time point of the
survey while 439 (0.7% of the total interviewed) did not have
symptoms.
Of 5120 persons (4681 with symptoms and an additional 439
with history of ATT) eligible for sputum examinations, spot
specimen were collected from 4850 (94.7%), of which 28 were
positive for AFB on smear microscopy and 65 on culture.

Table 1. Age and sex distribution of participants and non-participants.

Age group

1524

Male
No. registered

No. participated

9324

7506

1818

(80.5)

(19.5)

6506

2036

(76.2)

(23.8)

5147

1420

(78.4)

(21.6)

3825

743

(83.7)

(16.3)

2534

8542

3544

6567

4554

5564

65+

Total

Female
No. not
participated

4568

3326

4079

36406

2973

353

(89.4)

(10.6)

3806

273

(93.3)

(6.7)

29763

6643

(81.8)

(18.2)

No. registered

No. participated

No. not
participated

P value*

8728

8159

569

0.009

(93.5)

(6.5)

7446

502

(93.7)

(6.3)

7948

6181

4839

3937

3835

35468

5911

270

(95.6)

(4.4)

4645

194

(96.0)

(4.0)

3766

171

(95.7)

(4.3)

3672

163

(95.7)

(4.3)

33599

1869

(94.7)

(5.3)

0.002

,0.001

0.32

,0.001

,0.001

,0.001

(): percentages out of numbers registered.


*for test of significance between rates of participation between males and females.
doi:10.1371/journal.pone.0042625.t001

PLOS ONE | www.plosone.org

August 2012 | Volume 7 | Issue 8 | e42625

Prevalence of TB in a South Indian District

Table 3. Prevalence of PTB per 100,000 population, by method of estimation, all clusters-using screening by symptoms alone.

Method of Analysis

Type of PTB

Crude Prevalence

Cluster level analysis

Individual level analysis

Smear Positive PTB

81 (62100) N = 62959

74 (44103) N = 62959

83 (57109) N = 71874

Culture Positive PTB

137 (112161) N = 62840

Bacteriologically Positive PTB

175 (147203) N = 62841

126 (85168) N = 62840

162 (112213) N = 62841

152 (108197) N = 71874

196 (145246) N = 71874

( ): 95% confidence intervals.


N: numbers included in analysis.
: Reason for difference in numbers analyzed by three methods: for crude and cluster level prevalence, only those individuals from whom all actual data is available were
considered for analysis; persons missed from screening for symptoms and persons who were eligible for sputum collection but one or both sputum specimen not
collected or found contaminated on culture were excluded.
doi:10.1371/journal.pone.0042625.t003

Discussion

II. Prevalence of PTB corrected for non-screening by MMR


In six clusters where both screening tools were used, there were
a total of 31, 823 eligible persons of whom 28,398 (89.2%) were
screened by interview and 26,429 (83.0) by MMR. Using
individual level analysis, prevalence of bacteriologically positive
PTB by using screening for symptoms (and ATT) was 298 (CI:
327451) per 100,000 population; it was 389 (CI: 243354) per
100,000 population when screening by MMR was also considered.
Multiplication factor for prevalence to correct for non-screening
by MMR was 1.3.
Considering the estimates by individual level analysis as the best
estimates, prevalence among all persons $15 years of age residing
in 15 clusters, corrected for non-screening by X-ray, was 108 (CI:
82134), 198 (CI: 153243) and 254 (CI: 204301) per 100000
population respectively for smear, culture and bacteriologically
positive PTB.

Prevalence of bacteriologically positive PTB in persons $15


years of age corrected for non-screening by MMR in the entire
sample of selected clusters was estimated at 254 per 100000
population, after imputing the values for missing data. Prevalence
of smear positive and culture positive PTB was estimated at 108
and 198 respectively.
Prevalence was not estimated for abacillary (bacteriologically
negative but radiologically active) TB, due to low specificity of
diagnosis of PTB when based on single X-ray picture especially in
field conditions [10]. Similarly, tools for diagnosing EPTB are too
cumbersome to apply in field conditions.
A common generic protocol was used for simultaneously carried
out surveys during 200810 at eight sites in the country including
Nelamangala. However, screening by symptoms only was undertaken at five sites; MMR as additional screening tool was used in
the study populations at two sites and a proportion of study
population in Nelamangala. Using only symptom screening,
prevalence of bacteriologically positive TB at these eight sites
varied between 129 (CI: 92165) to 399 (325469) per 100,000
population [11]. Prevalence at three sites where both screening
tools were used varied between 199 (CI: 147251) to 391 (CI: 352
430) [11].
A number of surveys had earlier been carried out in different
parts of India since mid-twentieth century. In 1956, a nation-wide
survey using screening by MMR followed by sputum examination
by smear and culture among those with abnormal shadow on Xray film revealed the prevalence of bacteriologically positive PTB
at 400 per 100,000 population [1]. During subsequent surveys in
different geographical locations at different points of time,
prevalence of bacteriologically positive TB varied between 182
1270 per 100,000 [2]. These surveys not strictly comparable due
to variations in definition of symptoms, screening tools (symptoms
and/or MMR), case definition and analytical methods, nonetheless revealed that TB continued to be a high burden disease in
India. Male to female ratio in these surveys has been found to vary
between 2:1 to 5:1 [2]. It was 6:1 during the present survey.
The earlier survey in Nelamangla sub-division during 1975 was
carried out in 55 villages selected by SRS method [6]. A total of
12,105 persons $15years of age were screened for pulmonary
symptoms as well as by MMR. As in the present survey, a spot and
an early morning sputum sample were collected from those with
symptoms and/or any abnormal shadow on MMR. However,
there were differences in age structure with higher proportion of
persons in elder age groups during the present survey. We
compared the prevalence at the two surveys in the age structures
as they existed at given times since TB prevalence has been found

III. Relationship of Prevalence with sex and age


Estimated crude prevalence by symptom screening alone was
used to find out the relationship with sex and age. Prevalence was
about six times higher among males compared to females and
generally increased with age (Table 4).

Table 4. Prevalence (crude) of bacterologically positive PTB


per 100,000 population, by age-group and sex.

No.

Prevalence

Odds ratio

Age group
(Yrs)*
1524

15602

57.7 (25.989.5)

1.00

2534

13866

108.2 (62.0154.4)

1.88 (0.77, 4.63)

3544

10971

182.3 (115.0249.6)

3.16 (1.37, 7.49)

4554

8388

154.9 (83.9225.9)

2.69 (1.08, 6.81)

5564

6648

420.2 (288.8551.6)

7.33 (3.31,16.7)

65+

7366

339.4 (227.6451.2)

5.9 (2.63, 13.6)

Female

33403

50.9 (30.671.2)

1.00

Male

29438

315.9 (261.8370.0)

6.23 (3.63, 10.82)

Sex**

*x for linear trend = 41.4, P value,0.001.


**x2 = 62.7, P value,0.001 ( ): 95% confidence intervals.
doi:10.1371/journal.pone.0042625.t004

PLOS ONE | www.plosone.org

August 2012 | Volume 7 | Issue 8 | e42625

Prevalence of TB in a South Indian District

to increase with age [2]. Thus increase in life expectancy per se


can lead to increase in prevalence and can offset any declining
trends. We compared the two surveys for crude prevalence of
culture positive PTB using symptom screening only as authors of
previous survey have reported only crude prevalence of culture
positive PTB and that a proportion of study population during
present survey was not screened by MMR. This crude prevalence
was 311 per 100,000 population in 1975 and 137 (112161) in the
present survey [6]. This denotes a decline in prevalence by 56%
between the two survey periods Longitudinal studies in other
areas during NTP era had revealed no significant change in
prevalence [2]. Thus assuming that prevalence in the present study
area remained steady upto 1990, there was 56% decline in crude
prevalence between base year of Millennium Development Goals
(MDGs) of the United Nations and 2009 (mid-point of present
survey). This decline in Nelamangla can be attributed to overall
strengthening of TB control activities due to political commitment,
implementation of DOTS strategy with quality assured sputum
microscopy and administration of a rifampicin based treatment
regimen under observation with high rates of treatment success
[3]. During the year 2010, case notification rate of PTB (new+retreatment) under RNTCP in the Nelemangala Sub-division was
203 per 100,000 population in $15 years age group [unpublished
quarterly case finding reports of Nelamangala Tuberculosis Unit,
published data available only at district level]. This corresponds to
80% of the estimated point prevalence of bacteriologically positive
PTB at 254 per 100,000 population. Prior to DOTS, TB
programme performed poorly due to lack of adequate infrastructure and treatment with a self-administered regimen as in the rest
of the country [12]. The progress towards MDGs from the base
year of 1990 was thus far available only from Tiruvallur where
three rounds revealed a decline in prevalence by about 50%
between 1999 and 2006 [5]. Trends in prevalence of PTB in other
Asian countries have revealed declines of 32% in China between
19902000, 30% in Philippines between 19972007 and 67% in
Indonesia between 19802004 [1315].
A major limitation of the present survey was that mobile odelca
camera based MMR unit broke down mid way and could be put
back to use only after a long gap. This underlies the necessity of a
back up machine. We worked out a multiplication factor of 1.3 to
correct for non-screening by X-ray from the data in a subset of the
eligible population in which both screening tools were used. This
factor has been assumed to be homogenous across the study

population which could be a limitation of this approach given that


the prevalence even by symptom screening alone was higher in the
six clusters where both screening tools were used compared to
other clusters. Incidentally, these six clusters were located in closer
proximity to an urban area and had higher population density
than the rest. These six clusters were however not representative of
the study area. Same correction factor of 1.3 has been estimated in
Wardha district from the simultaneously carried out survey in
200809 (unpublished data). During earlier surveys in different
parts of the country, this correction factor was found to vary
between 1.5J and 1.7 [1619]. This variation may be attributed to
variation in quality of symptom screening by field surveyors.
Another limitation could be that missing value imputation
might not fully account for the missing cases in the event of higher
prevalence among the absentees compared to those surveyed.
Finally, the estimated prevalence reveals that TB is still a major
problem considering that much lower levels of prevalence have
been observed in countries with good TB control programs in
place [20]. To further reduce the disease prevalence, further
strengthening of the program may be undertaken to ensure
universal access to diagnosis, early detection of all incident TB
patients including multi-drug resistant (MDR) patients and to
achieve higher rates of treatment success. Repeat surveys if carried
out in all eight recently surveyed sites after a reasonable time
interval of approximately 78 years would provide information on
future trends in different parts of the country.

Acknowledgments
Authors are thankful to Umadevi, Jameel Ahmed, Ramesh Srivastav,
Narayana Prasad (Field Investigators), Dr. Alpana Mishra (Bacteriologist);
Surendra, Hemalatha, VijayLakshmi, Shivashankari, Raghunandan,
Manjunath (LTs); Ravindra, Padmesh, Anandababu (X-ray Technicians),
Dr L Suryanarayana, Dr (Mrs.) Sophia Vijaya (X-ray readers), Venkatachallappa, Nagaraja (Statistical Assistants) and Babis Sisimandis (for
guidance in statistical analysis).

Author Contributions
Conceived and designed the experiments: VKC SMA SS MVJ. Performed
the experiments: SS SN JG L MV SP VKC. Analyzed the data: SMA JR
MVJ VKC SB. Contributed reagents/materials/analysis tools: HK. Wrote
the paper: VKC SMA JR. Final approval, provision of resources and funds:
PK.

References
1. Indian Council of Medical Research (1959) Tuberculosis in India- A National
Sample Survey 19551958. ICMR special report series No. 34. New Delhi.
2. Chadha VK (2011) Epidemiology of pulmonary tuberculosis. Text book of
pulmonary and critical care medicine- First edition. Jaypee Brothers Medical
Publishers. New Delhi: 489510.
3. Central TB Division, Directorate General Health Services, Ministry of Health
and Family Welfare, Government of India (2012) TB India 2012- Revised
National TB Control Programme Annual Status Report. Available at http://
www.tbcindia.nic.in. Accessed 2012 June 4.
4. Khatri GR, Frieden TR (2000) The status and prospects of tuberculosis control
in India. Int J Tuberc Lung Dis 4: 193200.
5. Subramani R, Radhakrishna S, Frieden TR, Kolappan C, Gopi PG, Shanta T,
et al. (2008) Rapid decline in prevalence of pulmonary tuberculosis after DOTS
implementation in a rural area of South India. Int J Tuberc Lung Dis 12: 916
20.
6. Gothi GD, Narayan R, Nair SS, Chakraborty AK, Srikantaramu N (1976)
Estimation of prevalence of bacillary tuberculosis on the basis of chest x-ray
and/or symptomatic screening. Indian J Med Res 64: 11509.

7. Rieder HL, Van Deun A, Kam KM, Kim SJ, Chonde TM, et al. (2007)
Priorities for tuberculosis bacteriology services in Low Income Countries. The
international union against tuberculosis and lung diseases -Second edition. Paris.
8. Central TB Division, Directorate General Health Services, Ministry of Health
and Family Welfare, Government of India (2012) Revised National Tuberculosis
Control Programme training manual for culture and drug susceptibility testing.
Available: http://www.tbcindia.nic.in/pdfs/Training manual M. tuberculosis
CDST.pdf. Accessed 2012 June 4.
9. World Health Organization (2011) Tuberculosis Prevalence Surveys- a hand
book. WHO/HTM/TB/2010.17. Geneva.
10. Gothi GD, Chakraborty AK, Banerjee GC (1974) Interpretation of photofluorograms of active Pulmonary TB patients found in epidemiological surveys and
their five year fate. Indian J Tuberc 21: 907.
11. World Health Organization - Regional Office for South East Asia (2012).
Tuberculosis Control in the South East Asia Region. SEA/TB/338. New Delhi.
12. World Health Organization (1992) Tuberculosis programme review-India.
Geneva.
13. China Tuberculosis Control Collaboration (2004) the effect of tuberculosis
control in China. Lancet 364: 41722.

decline was calculated using the formula: (P22P1) *100/P1, where P1 and P2
represent the prevalence at the previous and present survey respectively

J
A correction factor of 1.5 implies that 33% of the cases will be missed if only
symptom screening is employed.

PLOS ONE | www.plosone.org

August 2012 | Volume 7 | Issue 8 | e42625

Prevalence of TB in a South Indian District

17. Datta M, Gopi PG, Appegowda BN, Bhima Rao KR, Gopalan BN (2000)
Tuberculosis in North Arcot District of Tamilnadua sample survey.
Indian J Tuberc 47: 14754.
18. Gopi PG, Vallishayee RS, Appegowda BN (1997) A tuberculosis prevalence
survey based on symptoms questioning and sputum examination. Indian J Tuberc
44: 17180.
19. Gopi PG, Subramani R, Govindaraj MS (1999) Quality of symptom elicitation
in an epidemiological survey on tuberculosis. Indian J Tuberc 46: 2614.
20. Styb1o K (1991) Epidemiology of tuberculosis- selected papers vol. 24. Royal
Netherlands. TB Association. Holland.

14. Tupasi TE, Radhakrishna S, Chua A, Mangubat HV, Guilatco R, et al. (2009)
Significant decline in the tuberculosis burden in the Philippines ten years after
initiating DOTS. Int J Tuberc Lung Dis 13: 122430.
15. Soemantri S, Senewe FP, Tjandrarini, Manissero D, Day R, et al. (2007) Threefold reduction in the prevalence of tuberculosis over 25 years in Indonesia.
Int J Tuberc Lung Dis 11: 398404.
16. Gopi PG, Subramani R, Sadacharam K, Narayanan PR (2006) Yield of
pulmonary tuberculosis cases by employing two screening methods in a
community survey. Int J Tuberc Lung Dis 10: 3435

PLOS ONE | www.plosone.org

August 2012 | Volume 7 | Issue 8 | e42625

You might also like