Setting: Analysis of tuberculin skin test surveys.
Objective: To estimate the annual tuberculous infection rate in Tamanrasset (southern Algeria) by applying on the one hand the classical method of the Tuberculosis Surveillance Research Unit (TSRU) of the International Union Against Tuberculosis, and on the other the study of variations of allergy published by Raj Narain et al.
Methods: 3675 pupils aged 5-18 years were tested in October 1982 or November 1983; 1240 of them were tested on both occasions. The technique of the World Health Organization, with 2 TU PPD RT23 tuberculin was used by trained testers. The TSRU method was applied to the 863 children aged 5-15 years without bacille Calmette-Guerin (BCG) scar, according to three estimates of the prevalence of infection (cut-off points of the distribution of reactions of 10 mm, of 14 mm corrected by dividing by 0.82, and mirror technique with a mode of 17 mm), and to three hypotheses of the decrease in infection risk (1%, 3% and 5%). The study of variations of allergy was used in the 1240 pupils tested twice (BCG-vaccinated or not).
Results: The average annual risk of tuberculous infection estimated by the TSRU method was lower than 1% (0.5%-0.6% according to the most stringent criteria). By the method of variations of allergy, the estimated annual infection rate was 0.9%. For the latter, our results are consistent with those obtained by other researchers.
Conclusions: These findings suggest that the study of variations of allergy might be a good alternative approach to estimate the annual infection rate in countries where children are BCG-vaccinated at birth, which is the case for most developing countries.