I. General concepts in TB Epidemiology II. Epidemiological indicators of TB and their estimation

Epidemiology of TBand its control

Dr. V. K. ChadhaSr. Epidemiologist

National TB Institute

Bangalore

I. General concepts in TB Epidemiology

II. Epidemiological indicators of TB and their estimation

III. Global epidemiological trends of TB

IV. TB situation in South East Asia

- presentations by Country participants

V. Prospects of TB control

Why do we need to study Epidemiology of TB?

Aims of Epidemiology ?

• To describe natural history of disease• Describe Distribution and relative importance • Measure frequency• To define risk groups• To evaluate interventions• To describe trends • To predict future trends and changes in disease

presentation.

What is Epidemiology ?Epi - among ; Demos - People ; Logos - Study

DEFINITION

Epidemiology is the study of the -

• Frequency

• Distribution - time, place & person

• Determinants - physical, biological, social,

behavioural & cultural

of health problems & health related events and

application of this study to control health problems.

ExposureSubclinicalinfection

Infectioustuberculosis

Non-infectioustuberculosis

Death

Riskfactors

Riskfactors

Riskfactors

Riskfactors

A Model for the Epidemiology of Tuberculosis

Rieder HL. Infection 1995;23:1-4

Risk of exposure ?* Incidence / prevalence of infectious TB in

the community

* Duration of infectiousness

* opportunities for case - contact interactions

-Urban/Rural

-No. of individuals in the house holds

Risk of Infection ?

* No. of infectious droplets produced

* Volume of shared air space

* Length of exposure

* Ventilation

* Climatic conditions

Grzybowski S, et al. Bull Int Union Tuberc 1975;50:90-106

Tuberculous Infection Among Children by Type ofContact and Bacteriologic Status of Index Case,

British Columbia and Saskatchewan, 1966 - 1971

Pe

r ce

nt

infe

cte

d

0

5

10

15

20

25

30

35

40Close

Casual Close

Casual

Smear + Smear -

Household transmission of TB- important epidemiological factor

• Case control study in Malawi

TB among contacts

Cases 770 56/2766

Controls 918 11/3203P<0.001

2 cases of TB

1 Infectious case

20 contacts

1 Non-infectious

-_-_-

Each case leads to two cases

Risk of Infection Among Contacts as a Function of the Proximity of Contact

What is the most important risk factor for TB?

Example of Risk Differences in IndividualsFollowing Infection with M. tuberculosis

Cas

es p

er 1

,000

per

son-

year

s(lo

g sc

ale)

1

10

100

1000

Long-standinginfection

Recentinfection

Super-imposed

HIV infection

UnderlyingHIV infection

??

Risk factors for disease given that infection has occurred ?

[Relative Risk of remotely acquired infection = 1] (0.2% per year)

Risk factor Relative Risk

AIDS 200

HIV Infection 30-40

Silicosis 30

Recent Infection 20

Under-nutrition 2-5

Diabetes mellitus 2-5

Incidence of TB in South Africa per 1000 population

0

5

10

15

20

25

30

General populationGold miners

IJTLD,3(9),1999,791-798

Other High Risk Groups

Populations in war / civil unrest Refugees and migrants Slum dwellers Homeless people/Foot path dwellers Smoking Prisoners

TB in prisons

Studies in Thailand

* TB incidence 90 times higher in prisons

* High HIV sero-positivity in TB cases

* High levels of drug resistance • RFLP studies signify role of recent transmission

Determinants of death?

* Severity of illness

* Smear positivity

* delay in diagnosis

* quality of treatment

* drug susceptibility pattern

Epidemiological indicators of TB and their estimation

Enumerate epidemiological indicators of TB you know of?

Epidemiological indicators of tuberculosis ?

* Prevalence of infection

* Incidence (average annual risk) of infection

(ARI)

* Prevalence of disease

* Incidence of disease

* Tuberculosis mortality rates

How to estimate prevalence of infection?

Estimating prevalence of infectionEstimating prevalence of infection

* Study population-sampling

* Registration of eligible age group

- house-to-house / school based.

* Informed consent.

* Examination for BCG scar.

* Tuberculin testing with 1TU/2TU PPD RT23 with tween 80.

* Reading of reaction sizes appx. 72 hours later.

What is the rationale behind tuberculin surveys in children ?

• Extent or recent transmission

• Study trends in TB epidemiology

(Ultimate aim of control programme is to replace older more infected cohorts with younger less infected cohorts)

Analysis of tuberculin surveyction size % of

childrenReaction size % of

children1 mm 16 mm2 mm 17 mm3 mm 18 mm4 mm 19 mm5 mm 20 mm6 mm 21 mm7 mm 22 mm8 mm 23 mm9 mm 24 mm10 mm 25 mm11 mm 26 mm12 mm 27 mm13 mm 28 mm14 mm 29 mm15 mm 30 mm

Frequency Distribution of Tuberculin SkinTest Reaction Sizes, Korea 1975

Induration (mm)

0 5 10 15 20 25 30

Fra

ctio

n re

actin

g

0.00

0.05

0.10

0.15

Korean Institute of Tuberculosis 1976:1-116

0

5

10

15

20

25

30

35

40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34

Reaction in mm

Perc

enta

ge

Frequency distribution of tuberculin reaction sizes among children aged 1-9 years - Kota

N = 3870

Estimation of incidence of infection?

Dual skin testing at two different periods -Conversion -Boosting

Compute average annual risk of infection

(ARTI) = 1-(1-P)1/A

A RT I

• Key epidemiological indicator in developing countries.

• It is the probability of acquiring new tuberculosis infection or re-infection over the course of one year.

A R I expresses the overall impact of various factors influencing the transmission of tubercle bacilli !

- Load of infectious cases

- Efficiency of case finding

- Efficiency of treatment programme

ARI identifies the regions of high transmission

It provides an indirect estimate of size of sources

of infection

Any change in disease burden and programme

implementation is first reflected in the change in

ARI

It holds the key to the study of epidemiological

trends which are more important than exact

estimates of disease prevalence

How to estimate prevalence of disease?

DISEASE SURVEY METHODOLOGYDISEASE SURVEY METHODOLOGY

Sampling of representative population

House to house registration

Screening:

- MMR X-ray of all above five years of age

- Symptomatic screening

X-ray pictures read by two independent readers and by an

umpire reader

Sputum specimens (2/3) collected from persons with abnormal

X-ray shadows & / or chest symptomatics

Sputum examination by direct microscopy (and culture).

How to estimate disease incidence?

Relationship between ARTI and incidence of disease

Styblo derived the following relationship from data of pre- chemotherapy

• Every one percent of ARTI corresponds to 50 new smear positive cases per 100,000 population per year

Relationship between ARI & Incidence of smear positive cases of Pulmonary Tuberculosis

(Indian studies)

Incidence / 100,000pop for every one

percent of ARI

NTI Longitudinalstudy (1961 – 1968)

53

BCG Prevention Trial 42-74 (57)

Relation between ARI and Incidence !

* Situation : Disease incidence remains same

but the risk of infection declines

Q 1. When is this situation likely?

Q 2. What is the impact on equation

(relationship) ?

What happens to the equation in high HIV settings?

The equation is dependent more on number of infections generated per case and not merely on incidence

Disease mortality rates !

* Community based prospective studies

* Death certification

ESTIMATION OF ANNUAL RISK OF TUBERCULOUS INFECTION IN DIFFERENT ZONES OF INDIA

-A CROSS SECTIONAL STUDY

2000-2003

Districts selected for National Sample Survey -ARI

The proportion of children with BCG scar by zone

North Zone 45.3%

South Zone 64.3%

West Zone 52.0%

East Zone 51.5%

The estimated prevalence of infection and ARTI by zone

ZonePrevalence

of infectionARTI

North10.3%

(8.4-12.2)

1.9%

(1.5-2.2)

West9.3%

(6.8-11.8)

1.8%

(1.3-2.3)

South6.1%

(4.9-7.2)

1.1%

(0.9-1.3)

East6.9%

(5.5-8.2)

1.3%

(1.0-1.6)

( ) : 95% C.I.

Prevalence of infecton by zone and stratum (1-9 years)

9.1

4.5

7.8

6.5

14.1

9.8

12.6

9.0

10.3

6.1

9.3

6.9

0

2

4

6

8

10

12

14

16

North South West East

%

Rural Urban Zone

Does higher ARTI in urban areas indicate higher incidence of

smear positive cases

Programme inputs The survey findings provide baseline estimates of ARI for

- evaluation of TB Control Measures.

- Study of epidemiological trends in years to

come High rate of ARI indicates high load of infectious cases of TB in

most parts of India. Prolonged and sustained efforts required to control TB.

There are significant inter-regional differences in tuberculosis situation.

Intensification of TB control services in urban areas with higher ARI rates to be taken up on priority basis.

Case finding expectations cannot be applied uniformly all over the country

Other Epidemiological indicators of Tuberculosis

* Ratio of prevalence and incidence

* Age distribution of cases

* Case fatality rates

* Force of MDR cases

* TBM notification rates

* Disability adjusted life years (DALY)

Epidemiological trends of TB

Tuberculosis Mortality in Three European Cities,Modeled From Available Data, 1750 - 1950

Year

1750 1800 1850 1900 1950

Dea

ths

per

100,

000

0

200

400

600

800

1000

Grigg ERN. Am Rev Tuberc Pulm Dis 1958;78:151-72

London Stockholm

Hamburg

Tuberculosis Mortality Rates in Germany, 1892 - 1940

Year

1890 1900 1910 1920 1930 1940

Dea

ths

per

100,

000

0

50

100

150

200

250

Redeker F. In: Handbuch der Tuberkulose (Hein J, et al, eds) 1958;1:473

Secular Trend in Annual Risk of Infection,Selected European Countries

Calendar year

1900 1920 1940 1960 1980

Per

cen

t ris

k (lo

g sc

ale)

0.01

0.1

1

10

Norway

PolandSlovenia

FranceNetherlands

England and Wales

Waaler H, et al. Bull Int Union Tuberc 1975;50:5-61Sutherland I, et al. Bull Int Union Tuberc 1971;45:75-114Lotte A, et al. Int J Epidemiol 1973;2:265-82

Sutherland I, et al. Tubercle 1983;64:241-253Styblo K, et al. Bull Int Union Tuberc 1969;42:5-104

Vynnycky E, et al. Int J Tuber Lung Dis 1997;1:389-96

Slope reference:% decline / year

Serbia

0%

5%

10%

15%

TB trends in EuropeMedian age in Finland

39.4

31.1

55.7

65.1

0

10

20

30

40

50

60

70

Males Females

19541986

TB trends in EuropeNetherlands

Year ARI Median Age

1950 0.53% 28.9 year

1980 0.21% 43.7 year

Global drug resistance surveillance

D.R. among newcases

D.R. amongpreviously treatedcases

1994-96 1.4% 13%

1996-99 1.0% 9.3%

Centers for Disease Control and Prevention. Reported Tuberculosis in the United States 1996:1997:5Centers for Disease Control and Prevention. MMWR 1998;47:253-7

Reported Tuberculosis Cases in the United States, 1953 - 1997

Year of notification

1950 1960 1970 1980 1990 2000

Num

ber

of c

ases

(lo

g sc

ale)

20000

40000

80000

Annual Risk of Tuberculous InfectionWHO South-East Asia Region

Year

50 60 70 80

Ris

k of

infe

ctio

n (%

)(lo

g sc

ale)

0.1

0.2

0.5

1

2

5

Slope reference:% decline / year

Cauthen GM. WHO Document 1988;WHO/TB/88.154:1-34

India

Indonesia

Thailand

1%

5%

10%

Trends in ARI-Chingleput

At intake in 1969 : 1.8%After 4 years in 1973 : 1.8%After 10 years : 1.9%After 15 years : 1.7%

How does HIV pandemic influence TB epidemic

• Higher rate of progression from latent infection to disease (5-10% per year compared to 10% per year among HIV negative)

• Previously HIV infected persons when exposed to TB rapidly develop the disease.

• Excess cases due to the above lead to increased transmission of infection

• Higher case fatality due to HIV infection

Evidence of association between HIV and TB

* Increase in TB in areas worst affected by HIV

* Higher increase in age group affected by HIV.

* 50 to 70% AIDS cases develop TB in SEAR.

* HIV positivity higher among TB cases than

general population.

-Northern Thailand: HIV positivity in TB cases :

40%

: Malawi : 75%

Total population

Infected withM. tuberculosis

Infected with HIV

Determinants for the Frequency of HIV-Associated Tuberculosis in a Community

Prevalence of infection with M. tuberculosis

Prevalence and incidence of HIV infection

Overlap of the two respective population segments

Impact of HIV Infection on Tuberculosis Notificationsin Chiang Rai, Thailand, 1985 - 1994

Year of notification

85 90 95

No.

of c

ases

(lo

g sc

ale)

200

300

400

500

All cases

HIV-neg cases

Yanai H, et al. AIDS 1996;10:527-31

TB trends in Africa (countries with high HIV rates)

0

50

100

150

200

250

300

350

1980 1985 1990 1995 2000

Sta

nd

ard

ize

d n

oti

fic

ati

on

ra

te

Estimated TB incidence vs HIV prevalence

0

200

400

600

800

0.0 0.1 0.2 0.3 0.4HIV prevalence, adults 15-49 years

Esti

mat

ed

TB

inci

de

nce

(p

er

100K

, 199

9)

Notification Rates of Sputum Smear-Positive Tuberculosis,by Age, Tanzania Mainland, 1984 and 1995

Age group (years)

0 15 25 35 45 55 65

Not

ifica

tions

per

100

,000

0

50

100

150

200

Tanzania NTLP / IUATLD. Progress Report 1996;No. 36

1995

1984

20% of all patients in Russia have MBR TB

TB morbidity rates in Russia

0

10

20

30

40

50

60

70

80

90

1970 1980 1990 1997 1998 1999

per l

akh

pop.

0

5

10

15

20

25

30

35

1987 1997

%

Case fatality rates in Russia

Increase in CFR attributable to increase in drug resistance cases

Culture Positive cases, Prevalence:Incidence - Chingleput

0

1.5

3

4.5

1968-70 1971-73 1973-75 1976-78 1979-81 1981-83

Average - 3.4

(3.6 for smear pos)

In your opinion, what should be the practical methods of monitoring epidemiological trends in any given community

Global picture

• 3rd largest cause of death (2.8%) and loss of DALYs in 15-59 year age group

• Incidence all cases - 8.8 million (2002)-141/100000

• in 22 HBCs - 7.0 million (80%)

• Smear + - 3.9 (63/100000) million

• Case notifications of smear positive cases increasing @ 4% per year- 5% in eastern Europe and 7% in high HIV African countries.

Epidemiological situation of TB in South East Asian countries

Format for Country presentationsEstimated incidence of New smearpositive casesLatest estimates of ARTI

Population mortality rates

Any information on disease trends

HIV sero-prevalence among TBcasesMDR in new cases

MDR in previosly treated cases

Any other epidemiologicalinformation eg, age sex distributionof cases, TB in prisoners etc.

TB in South-East Asia

WPR25%

AFR18%

EMR8%

EUR6%

AMR5%

SEAR38%

Incidence: 3 millDeaths : 1 mill (1500/day)

India, Bangladesh, Indonesia, Myanmar & Thailand contribute 95% of regional burden

HIV-TB in SEAR

* Second largest number of HIV positives after SSA

SSA:60% SEAR:30%

* 6 million HIV positives in SEAR

India :4 mill

Thailand :1 mill

Myanmar :0.5 mill

* Low sero-positivity in Bangladesh, Maldives, Bhutan, Indonesia and Sri lanka

* Nepal : Low in antenatal women, high among IDUs.

TB situation in India

Prevalence of sputum positive pulmonary TB

Area Year Preval. Rate per 1000 pop.

National Sample Survey 1955-58 4

Tumkur 1960-611979

4.14.4

Rural Bangalore 1960-611967-681974-751984-86

4.13.93.24.4

Chingleput 1968-711973-751979-811984-861999-2001

10.78.97.76.96.9

Raichur 1988-89 10.7

Morena, M.P. 1991-94 12.7

A R T I i n I n d i a

0

0.5

1

1.5

2

2.5

3

3.5

4

Tum

kur D

istt,

1960

-61

Tum

kur D

istt,

1972

-73

Rura

l Ban

galo

re, 1

961

Rura

l Ban

galo

re, 1

970

Bang

alor

e(Ru

ral),

197

7-78

Bang

alor

e Ru

ral,

1984

Per

i urb

an B

anga

lore

, 199

2

Bang

alor

e Ci

ty, 1

997

Chin

glep

ut, T

N 1

969

Chin

glep

ut, T

N 1

979

Chin

glep

ut, T

N 1

984

Car N

icob

ar Is

land

, 198

6

Triv

andr

um, 1

991-

92

Bika

ner

Raj

, 199

2

Mor

ena,

MP,

1989

Tiru

vallu

r, 19

99-2

001

Nor

th Z

one-

Indi

a 20

00-0

2

Wes

t Zon

e-In

dia

2000

-02

Sout

h Zo

ne-In

dia

2000

-02

East

zon

e In

dia

2001

-03

Annu

al R

isk

of I

nfec

tion

(%)

INCIDENCE OF PULMONARY INCIDENCE OF PULMONARY TUBERCULOSIS IN INDIATUBERCULOSIS IN INDIA

Study Period Method Incidence

1961-62 1.361962-64 0.80

Bangalore RuralAge 5years 1964-68

Repeated Surveys1.04

Def : Culture +ve

1968-71 3.83

1976-78 2.30

BCG.TRIAL,ChingleputAge > 15 years

1981-83

Repeated Surveys,passive case

finding , selectivecase finding 3.00

Def: Culture positive &/or Microsopy +ve

CMC – VelloreAge> 10 yrs

1981-83 Active case finding 1.10

Def:Smear +ve

HIV Sero-prevalence among TB Cases

Year of study % HIV +ve

Govt Hospital Tanjavur, TN 1999 8.9

General Hospital, Pune, MH 2000 28.8

TB & Chest Hospital,Goa 2000 10.9

AIIMS, Delhi 2000-02 9.4

Medical College, Lucknow 2000-01 4.3

Medical College, Aligarh 2000-01 2.8

Multi Drug Resistance in new TB cases

Year of study % MDR

23 districts of Tamilnadu 1997 3.4

DOT centres, Bangalore 1999 2.2

DTP centres, Raichur, Karnataka 1999-2000 2.5

Wardha, Maharashtra 2000-01 0.5

Jabhalpur, Madhya Pradesh 2001-02 1.0

Hoogli, West Bengal 2000-01 3.0

Mayurbhanj, Orissa 2000-02 0.7

Multi Drug Resistance in previously treated TB cases

Year of study % MDR

TB Sanatorium, Chennai, TN 1997-2000 54.8

DOT centres, Tiruvallur, TN 1999-2000 18.3

State TB Centre, Ahmedabad, GJ 2000-01 33.0

ARI in other countries

2.32.2 2.2

1.5

1

2

2.6

1.1

0.9

0.6

0.3

2

0

0.5

1

1.5

2

2.5

3

Series1

Incidence of allcases

Country Pop. inmillion

Globalrank

%contribution

Total(000)

Rate/100000

India 1045 1 20 1761 168

Indonesia 217 3 6 557 256

Bangladesh 144 5 4 318 221

Thailand 62 19 1 80 128

Myanmar 49 22 1 75 154

Country wise Epidemiology situation

Country wise Epidemiology situation - Continued

Incidence of ss +CountryTotal(000)

Rate/100000

Prevalence(ss +)

/100000

TBMortality/

100000

HIV +TB

cases

%casesMDR

India 787 75 156 37 4.6

(0.4-28)

3.4

Indonesia 250 115 272 59 0.6 0.7

Bangladesh 143 99 188 520 0.1 1.4

Thailand 35 57 254 86 24 0.5

Myanmar 33 68 83 26 11 1.5

Country DOTSpopulationcoverage

(%) - 2002

Treatmentsuccess (%)

– 2001cohort

DOTS detectionrate

(ss +) - 2002 (%)

India 52 85 31

Indonesia 98 86 30

Bangladesh 95 84 34

Thailand 100 56 47

Myanmar 88 81 73

Progress of DOTS in high burdened countries

High treatment success (>70%)Low treatment

successCase detection under

DOTS 10-49%>50%

Brazil, Russia,South Africa,Uganda

Afghanisthan,Bangladesh, China,Euthopia, India,Indonesia, Kenya,Mozambique, Nigeria,Pakistan, Tanzania,Zimbabwe

Cambodia, Cango,Myanmar, Philipines,Thailand, Vietnam

What is meant by control ?

• To move from high to low endemicity or elimination

Objectives of TB control programmes

• Decrease transmission of infection by:-

- Rapidly identifying cases

- Adequate treatment• Decrease deaths due to TB.• Cure of maximum number of cases.• To prevent relapse.• To prevent emergence of drug resistance.• To reduce TB in children by preventive treatment.• IEC - Purpose ?

2 cases of TB

1 Infectious case

20 contacts

1 Non-infectious

-_-_-

Each case leads to two cases

How does DOTS strategy help control TB?

DOTS

• Decreases deaths

• Decreases duration of infectiousness

• Increased case detection plus high cure rate decreases transmission of infection that will ultimately lead to decline in incidence.

• Prevents emergence of MDR

A good programme like DOTS reduces disease burden

• Case fatility rate reduced to <5% compared to 60%-70% in a few years among untreated cases.

• Cure of every case under DOTS with about 4 months diagnostic delay prevents 0.7 new smear positive cases.(further prevention possible by reducing diagnostic delay)

• Preventive treatment to each child prevents 0.03 new case and 0.007 deaths.

How does a poor programme worsen the TB situation

• Poor programme with low cure rate (<50%) and low detection rate worsen TB situation by decreasing case fatility rates leading to increased prevalence and transmission of infection.

HIV prevention and control is of major importance towards TB

control

Priority to smear positive cases

• To reduce transmission of infection. A good DOTS programme would reduce transmission of infections by about 73%

• Cost per DALY highest for treating smear positive cases.

The Cuba example

• Very low levels of MDR in Cuba

• Cuba is a low HIV country

HIV-TB Vs. DOTS - TB trends in Tanzania

0

10

20

30

40

50

60

70

80

90

100

1978-82 1983-87 1988-92 1993-97

0

0.2

0.4

0.6

0.8

1

1.2

smear positivenotif icationrate/ 100,000popTreatmentcompletionrates

ARI

• Increased case detection will decrease transmission rapidly provided cure rates are high.

• It has been estimated that achievement of 70% case detection and 85% cure rate by 2010 will result in greatest benefits in cases and deaths averted in regions with highest burden - South East Asia, Africa and Western Pacific.

• Longer the time taken to reach targets, incidence will decrease more slowly.

• The proportion of deaths averted by DOTS would be greater than the proportion of cases– Non curative treatment can prevent death without

eliminating infectiousness.

– Programme will treat non-infectious cases also

Control TB since every breadth counts (World TB day 2004 theme)

Business as usual will not eliminate TB

It is time for business unusual