Glynn, J. RWarndorff, D. KFine, P. E. MMsiska, G. KMunthali, M. MPonnighaus, J. MAwusabo-Asare, KofiPisani, ElisabethBoerma, J. TiesZaba, Basia2003-03-122004-05-192011-01-052004-05-192011-01-051997http://hdl.handle.net/1885/41321http://digitalcollections.anu.edu.au/handle/1885/41321Since the mid-1980s tuberculosis (TB) case numbers and HIV seroprevalence have both risen sharply in sub-Saharan Africa. Estimates for the relative risk of TB in those infected with HIV have ranged from less than five to more than 20. The proportion of TB cases attributable to HIV (the population attributable fraction) has been calculated for several populations but is difficult to interpret if no account is taken of the age and sex distribution of the cases. In a rural area of Malawi we have studied the proportion of TB attributable to HIV over time. Nearly 40 per cent of smear-positive TB cases in this rural area of Malawi can now be attributed directly to HIV. The actual effect of HIV on TB is even greater than this because increased case numbers increase transmission of tuberculosis infection to both HIV-infected and non-infected sections of the population. We compare our findings with others from sub-Saharan Africa and discuss reasons for the differences, and methodological issues in interpretation53612 bytesapplication/pdfen-AUAfricaMalawiHIV/AIDSmorbiditymortalitytuberculosisTB1997