1. Direct estimates of the burden of disease EURO Workshop November 2003

2. File: # Notes and Examples.doc

3. We want to reduce Morbidity Mortality New infections
Recurrence of old infections
Size of the infector pool

4. Routine programme data
Notifications
Cure rates
Drug resistance
Mortality
Process indicators

5. But...
Biases
Incompleteness
Recording
Reporting
Analysing

6. Four things we can measure
Prevalence of infection (tuberculin surveys)
Incidence of infection (ARTI)
Incidence of disease
Prevalence of disease

7. TB in Czechoslovakia 1940s Mass BCG vaccination campaign.
1950s Notifications ~ 200/100k/year. Deaths ~ 49/100k/yr Meningitis in children ~ 9/100k/yr.
1953 Compulsory BCG vaccination of babies
1955 Government programme to increase treatment and diagnostic capacity; eradicate bovine TB; increase social welfare payments to TB patients.
Notifications ~ 130/100k/year (8% per yr);
Deaths ~ 20/100k/year (16% per yr);
Meningitis in children ~ 0.4/100k/year (52% per yr).

8. 1960: Long term epidemiological and clinical study of tuberculosis in Kolín
Improve screening for active tuberculosis;
Find a way to provide adequate therapy on national scale;
Protect the uninfected population.
Therapeutic targets: 95% cure rates; relapse rate below 0.5% per year.
Assess the impact of mass screening.

9. Prevalence of infection
Tuberculin surveys (1961)

10. Prevalence of infection
7 yrs yrs
14 yrs yrs
24 yrs
Prevalence of infection
Age
Vaccinated
Unvaccinated
7 yrs
0.11  0.02
0.11  0.08
11 yrs
0.29  0.04
0.31  0.07
Vaccinated Unvaccinated

11. from prevalence (P) and age (A)
Calculating ARTI (R)
from prevalence (P) and age (A)
R = 1 – (1 – P)1/A
Probability of being infected per year = R
Probability of not being infected per year = 1 – R
Probability of not being infected after A years = (1 – R)A
1 – P = (1 – R)A
1 – R = (1 – P)1/A

12. Prevalence of infection Annual risk of infection
Age
Vaccinated
Unvaccinated
7 yrs
0.11  0.02
0.11  0.08
0.017
(0.0140.020)
(0.0040.031)
11 yrs
0.29  0.04
0.31  0.07
0.030
(0.0260.035)
0.034
(0.0250.043)

13. Assumptions
The annual risk of infection is constant over time and independent of age.
Surveys in school children, aged about 5 to 10 years: estimate of the ARTI is averaged over about five to ten years.

14. Exponential decline
 = 0.05/yr

15. Rate of decline of ARI

16. Prevalence of infection
Based on routine data 1960 to 1964
Mass X-ray screening 1961 and 1963

17. Estimating sample sizes quickly
If you count N positives then:
 N
95% confidence limits  1.96N
Fractional error  1.96N/N  2/N
Incidence  100 per 100,000
95% confidence limits  ±2/N = ±20 (20%)

18. Decline in prevalence
Men: 20%  10%/yr
Women: 26%  21%

21. Incidence of infection
Mass X-ray screening 1961 and 1963

23. New smear-positive cases
21%  4%/yr

24. Cases of bacillary TB over 4 years
Effectiveness of BCG
Cases of bacillary TB over 4 years
Vaccinated: 12/11,636 = 26/100k/yr
Unvaccinated: 15/4,073 = 92/100k/yr
Odds ratio = 0.28 (p = )

25. Conclusions
Adequate chemotherapy greatly reduced the number of chronic excretors: need good treatment
Most cases identified by smear microscopy: good case finding
One round of mass X-ray screening worth doing to reduce the size of the infector pool.
Project substantially increased the rate of decline
Unlikely to sustain these high rates as the epidemic ages
Relapse rate did not fall: duration of the project too short.
BCG appears to be effective but there are many potential confounders.

26. Rates of decline in TB notifications per 100,000 per year
Czechoslovakia: 1955 to 1965 ~ 8%
Kolín: 1961 to 1964 ~ 20%
Czech. & Slovak.: 1997 to 2001 ~ 6%

27. Exercises
1, 3, 4, 2, 5, 6