Annual risk of tuberculous infection. 1988.

Abstract

I. INTRODUCTION Risk of infection in a population is in many respects the most informative index of the magnitude of the tuberculosis problem (Sutherland, 1976; Pio, 1984). The risk of infection at a particular rime indicates the current magnitude of the incidence and prevalence of infectious cases (Styblo, 1985) and also indicates the magnitude of the tuberculosis problem years into the future. An observed decline in the risk of infection would be the earliest indicator of a decline in the epidemic cycle of tuberculosis, resulting from tuberculosis control activities or improvements in living standards. A rising risk of infection would be an early indicator of changes in the other direction, signalling the introduction of new risk factors, such as the spread of human immunodeficiency virus (HIV) infection. It is known that risk of infection bas been declining for many years in developed countries, but remains at high levels in many developing countries (Styblo, 1984). In developing countries in the last decade, a number of surveys of infection prevalence have been carried out in national populations as well as in smaller populations. Therefore a project was carried out to assess the current level and trend in the risk of infection in developing countries by reviewing and assembling tuberculin skin test survey data available since 1975. II. METHODS Tuberculin skin test data collected since 1975 for populations in developing countries were assembled from reports to the World Health Organization and from the published literature. Data from surveys of childhood age groups that were judged of sufficient quality were selected in order to provide as valid and up to date an assessment as practical of the magnitude of the risk of infection. Comparable prior data for the same countries were also selected in order to judge whether the risk of infection is likely to have declined. Prevalence of infection observed in childhood age groups was used to derive the average annual risk of infection that would have resulted in the observed cumulative prevalence rate. Choosing younger age groups allows the calculated average annual risk to be bracketed within a relatively narrow period of time between the average birth date of the group and the date of the survey (Styblo et al, 1969), Each survey was judged on the basis of available documentation for ability to represent the population and to detect the proportion infected at a particular time. A sample survey was judged to have measured the proportion infected in the target population if the probability sampling design and the estimation method appeared correct in concept and conduct, and skin testing technique appeared adequate to measure the proportion infected in the sampling units. Basis for selection Surveys were selected which met most of the following criteria: Specified the sampling design including the sampling frame, staging, stratification, sampling units, allocation, sampling weights, and estimation formulas and documentation to support that the design was followed with adequate coverage. Provided reaction-size distributions for surveyed age groups and for bacteriologically confirmed cases from the same population in order to assess the definition of infection and in order to judge technique. Documented type and strength of antigen used, technique of administration and reading, and quality control procedures. Specified the methods used to detect and eliminate persons with a history of BCG vaccination, and described BCG policy and actual coverage in the population age group surveyed. Data assembled from selected surveys: The area population represented and the survey period. In the youngest age ranges for which sufficient numbers were tested and read, the number tested and read, the number counted as infected with M. tuberculosis, and the mean age. …