Setting: The first tuberculin survey conducted in Kenya by the World Health Organisation in 1958-1959 found an annual risk of tuberculosis infection (ARTI) of 2.5%.
Objective: To estimate the ARTI and its trend in recent years and to compare the estimated incidence rates with the notification rates.
Design: A tuberculin survey was held in 12 randomly selected districts in the period 1986-1990. Tuberculin testing with 2TU PPD RT 23 + Tween 80 was performed in 40,365 primary schoolchildren aged 6-13.
Results: Of 14,984 non BCG-vaccinated children, 1,380 (9.2%) had indurations of > or = 10 mm. Double testing with PPD RT 23 and PPD-scrofulaceum in 980 non BCG-vaccinated children revealed a high level of infections due to mycobacteria other than tuberculosis (MOTT). Therefore, the prevalence of tuberculous infection was based on the sum of 50% of the indurations of 17 mm and all indurations of 18 mm or more multiplied by two. The prevalence of tuberculous infection in schoolchildren aged on average 8.4 years, 'weighted' for the population size according to the provisional results of the 1989 census, was calculated at 5.5%. The corresponding ARTI is 0.6%.
Conclusion: The ARTI has declined by an average 4.6% per year. The tuberculosis problem differs from one area to the next, with the highest prevalences of infection on the coast and in Eastern Kenya, and the lowest in Western Kenya. Although the average ratio of observed and estimated incidences indicate that 70% of incident cases are notified, considerable inter-district variations are observed.