|IFPMA Press Review 7/30/2004
How much of China's success in tuberculosis control is really due to DOTS? The Lancet By S Bertel Squire, Shenglan Tang China, 7/30/2004 - In today's Lancet, Chris Dye and colleagues, from the China Tuberculosis Control Collaboration, analyse the ...
IFPMA Press Review 7/30/2004
How much of China's success in tuberculosis control is really due to DOTS?
By S Bertel Squire, Shenglan Tang
China, 7/30/2004 - In today's Lancet, Chris Dye and colleagues, from the China Tuberculosis Control Collaboration, analyse the surveys of tuberculosis prevalence done by the Chinese Ministry of Health in 1990 and 2000. Their landmark article shows that the estimated prevalence of pulmonary tuberculosis, in the 13 provinces in which the DOTS programme was implemented with support from the World Bank, was 573 per 100 000 population in 1990, and fell to 298 per 100 000 population in 2000;1 a reduction of 48% (95% CI 35-61). In the remaining 15 provinces, which were not supported by the World Bank (the municipalities of Beijing, Shanghai, and Tianjin were excluded from the analysis), the prevalence fell from 493 per 100 000 to 412 per 100 000; a reduction of 16% (1-35). On the face of it, DOTS brought about a 30% greater fall in tuberculosis prevalence than would have occurred had DOTS not been implemented. This reduction holds true if the same comparison is made for culture-positive or smear-positive tuberculosis instead of all pulmonary tuberculosis (defined radiologically). A great achievement for DOTS is claimed: that 382 000 culture-positive and 280 000 smear-positive cases of tuberculosis have been eliminated from a population of more than half a billion children and adults.
There can be no doubt that DOTS, the internationally recommended strategy for tuberculosis control,2 has spear-headed remarkable improvements in tuberculosis control in China.3 We think, however, that it is important to look closely before allowing DOTS to claim all the credit. First, were the prevalence surveys designed and done in such a way as to make comparisons between 1990 and 2000 straightforward? Sadly, the answer is no. Although the designs were similar, the diagnostic protocols differed. In addition, the 1990 survey examined more people (1 461 190) than the one done in 2000 (365 097). A consequence of this difference is that province-specific estimates of prevalence were less accurate in 2000 than in 1990. This lower accuracy is shown by the fact that the estimate of 29 smear-positive cases per 100 000 population in Beijing (total population about 14 million) is based on the detection of a single smear-positive case from two sampling units totalling about 3000 people.