Udo Buchholz,

1
Control of tuberculosis analytical methods

• Udo Buchholz,
• WHO/Stop TB/TME

2
Subjects of this talk

• DOTS
• TB and HIV
• DOTS plus
• - but not
• What PPM can contribute
• How poverty can be addressed
• The influence of important, modifiable risk
  factors

3
Evolution of and interaction between evidence,
policy and reality
Modelling
Uncertainty
New parameters, New parameter estimates
Baseline conditions change
Policy
Confirmation? Discrepancies?
New facts of life (war, HIV, economics, collapse
of systems, more/less funding)
New options for intervention (new drugs, new
diagnostics, new vaccines)
Reality
4
"Tuberculosis can be controlled with existing
technologies"1

• Elements of DOTS were shaped through results
  from
• studies, such as that in Kolin,
• analyses of IUATLD-assisted NTPs in Tanzania,
  Malawi, Nicaragua
• theoretical considerations1
• ... and later confirmed by TB models2

1K. Styblo, J.R. Bumgarner. TSRU progress report
1991. 2C. Dye et al. Prospects of worldwide
tuberculosis control under the WHO DOTS
strategy. The Lancet 1998.
5
Estimated prevalence, when incidence 100 ss/100k
and prevalence baseline 200 ss/100k1

• Limitations
• Focus on smear
• positive TB only
• No inclusion of
• HIV and MDR
• Analysis of one
• generation only

1K. Styblo. Tuberculosis can be controlled with
existing technologies. TSRU progress report 1991.
6
Dynamic modelling Fall in incidence depends on
(1) cure rate, (2) failures remaining
infectious1
1C. Dye et al. Prospects of worldwide
tuberculosis control under the WHO DOTS strategy
. The Lancet 1998.
7
How can we measure what DOTS does?
8
Indicators
INPUT Policy environment Human and Financial
Resources Infrastructure
PROCESS Management Training Drug
management Laboratories Communication Advocacy
OUTPUT Diagnostic services Treatment
services Improved knowledge, attitudes, and
practices Reduced stigma
OUTCOME Case detection Treatment success
IMPACT Prevalence of TB infection Prevalence of
TB disease TB morbidity TB mortality
Logical and chronological order
p7, Compendium of Indicators for Monitoring and
Evaluating National Tuberculosis Programs USAID,
MEASURE, CDC, WHO, IUATLD, KNCV, MSH.
WHO/HTM/TB/2004.344, August 2004
9
With the ultimate goal...

• ... to reduce incidence, prevalence and deaths
  from TB

The MDGs
10
But there was a decline, e.g. in TB mortality
long before DOTS...

• ... so how do we know that our achievements are
  due to DOTS?

11
Ideally we would like to test DOTS in a RCT
DOTS
non-DOTS
... but DOTS is a package of (soft)
interventions, not a vaccine or drug
12
Evaluation of DOTS under real-life conditions

• Comparison of districts or provinces (DOTS vs.
  non-DOTS)
• Historical control (DOTS vs. before DOTS)
• Correlation of concurrent changes in
  input/process/output indicators with changes in
  impact indicators (weakest)

13
Example 1 comparison of DOTS with other
districts China
14
Example 2.1 historical comparison El Salvador
15
Example 2.2 historical control (Maldives) DOTS
since 1996
Decline 9/yr
16
Example 3 correlation of concurrent changes in
indicators

• Average number of slides examined per patient
  proportion of TB cases that are smear positive (
  case detection rate)
• Number of persons involved in DOT and defaulter
  tracing treatment outcome
• Number of NGOs involved in NTP programme
  reporting case detection rate
• Etc. ...

17
Return in notified cases as result of increasing
diagnostic efforts ( smears examined), Peru
1Suarez. The dynamics of tuberculosis in response
to 10 years of intensive control effort in Peru.
JID 2001.
18
TB/HIV HIV/AIDS and TB
In areas with high HIV-prevalence

• HIV negative persons and HIV negative TB patients
• Is HIV increasing MTB transmission?
• How can TB in HIV negative persons be controlled?
• HIV positive patients
• TB treatment
• Prevention
• of HIV
• of TB by giving HAART
• of TB disease by treating latent TB infection
  (TLTI)

19
HIV negative TB

• Study among South African gold miners1
• Combined data from 4 time periods in cohort
  studies
• DOTS programme in place
• Increase of HIV from 1 in 1991 to 24 in 1999
• Increase of TB from 1000 per 100k (1990) to 4000
  per 100k (1999)
• Silicosis a strong risk factor for TB
• ? Changes in MTB transmission should be reflected
  in incidences among HIV negative gold miners

1E. Corbett. Stable incidence rates of TB among
HIV negative South African gold miners during a
decade of epidemic HIV-associated TB. JID 2003.
20
TB notification rates among South African gold
miners
21
Trends in incidence of TB in HIV negative miners
P-value for trend 0.17
22
Further analysis (1)

• Risk factors in univariate analysis
• (1) Age

23
Further analysis (2)

• Risk factors in univariate analysis
• (1) Age
• (2) Previous TB
• (3) Surface work (protective)
• (4) Silicosis grade
• ? Multivariate analysis to adjust for the
  different risk factors

24
Multivariate-adjusted incidence rate ratio (IRR)
for calendar period as risk factor for TB among
HIV neg. gold miners

• DOTS seems to be able to control TB among HIV
  negative persons even in settings with a high
  burden of HIV-associated TB

25
Malawi would there have been a decrease of TB in
the absence of HIV?1
1J. Glynn. Trends in tuberculosis and the
influence of HIV infection in northern Malawi,
1988-2001. AIDS 2004.
26
HIV positive TB patients

• Model by Currie et al.1 to compare the
  effectiveness of
• TB chemotherapy - with -
• Prevention of HIV
• HAART
• Prevention of TB disease through TLTI (6-9
  months)
• Example of Kenya

1CSM Currie. Tuberculosis epidemics driven by
HIV is prevention better than cure? AIDS 2003.
27
Model output impact on TB deaths when each
intervention is increased by 10 from 2002
Rapid impact of TB cure and detection,
but greatest, but late impact of reduction of HIV
incidence
28
TB deaths averted over 10 years with 100
coverage of each intervention
Red HIV prevalence stabilizes at 50 of its
current value Blue ... at 100 of its current
value Green ... at 150 of its current value
29
Efficacy of treatment of latent TB infection in
HIV infected persons

• Cochran review (2004)
• Efficacy effect under controlled conditions
  (RCT)
• Effectiveness effect under real-life
  conditions, e.g. as measured by case-control
  studies or retrospective cohort studies
• Overall RR 0.64 (95CI 0.51-0.81)? Efficacy
  1-RR 36
• Efficacy among tuberculin skin test positive
  persons 0.38 (95CI 0.25-0.57)? Efficacy 62

30
TLTI works also under routine conditions

• South African gold miners
• Objective to collect data on effectiveness under
  routine conditions
• Randomized intervention study each participant
  had the opportunity to receive the intervention
• INH for 6 months, self-administered among HIV
  positive persons without evidence of active TB
• No HAART

1AD Grant. Effect of routine isoniazid preventive
therapy on TB incidence among HIV-infected men
in South Africa. JAMA 2005.
31
Results of INH study, S-A gold miners

• TB incidence rate before vs. after clinic
  enrolment dropped from 11.9 to 9 per 100
  person-years ? IRR 0.78
• Multivariate analysis and restricting analysis to
  persons with no h/o TBIRR 0.54 (0.35-0.83) ?
  Reduction of TB incidence by 46
• Nevertheless additional measures are urgently
  needed

32
MDR and DOTS plus

• National cohort study in Peru
• Feasibility of using second-line drugs to treat
  patients with chronic TB
• Encouraging, but not overwhelming results

1PG Suarez. Feasibility and cost-effectiveness of
standardised second-line drug treatment for
chronic tuberculosis patients a national cohort
study in Peru. The Lancet 2002.
33
Preliminary results of the first five DOTS-Plus
projects1

• More than 3,000 MDR-TB patients enrolled and
  1,047 who have completed treatment
• MDR-TB among new cases ranged from 1.5-17.1
• 65 of the MDR-TB cases treated were resistant to
  both first and second-line drugs
• The overall treatment success was 70
• 77 among new patients
• 69 among previously treated patients
• 3.7 of patients stopped treatment due to adverse
  events

1Slide thanks to Eva Nathanson, STB/THD
34
DOTS-Plus - outcomes of 1047 MDR-TB patients
35
Methodological summary

• The evidence for the effect of control
  interventions in TB includes the following
  methods
• RCT/intervention study - example TLTI in
  HIV-positive S-A gold miners
• Surveillance data ex. DOTS
• Cohort studies ex. TB incidence among HIV
  positive and negative persons
• Modelling the simulation of an experiment
  ex. effect of CT vs. preventive methods on TB
  cases and mortality

36
Thank you