128
Early Detection of Effect on Adult Mortality of a Government ART Program in Rural KwaZulu Natal, South Africa
Till Barnighausen*1,2, G Cooke1,3, K Herbst1, A Kany Kany1, and M L Newell1,4
1Africa Ctr for Hlth and Population Studies, Univ of KwaZulu Natal, Somkhele, South Africa; 2Harvard Sch of Publ Hlth, Boston, MA, US; 3Imperial Coll, London, UK; and 4Univ Coll London, UK
Background: The HIV epidemic has been associated with large increases in
mortality in South Africa, reducing average life expectancy to the late
forties. The benefit of ART to an individual with advanced HIV infection is
well established, and the widespread introduction now ongoing in Sub-Saharan
Africa thus has the potential to reduce the substantial increases in
HIV-related deaths. We studied changes in population mortality following the
introduction in late 2004 of an ART programme into rural South Africa.
Methods: We used data from demographic (from 2000) and separate HIV (from
2003) surveillance within the Africa Centre research portfolio in the Mpukunyoni
area, Hlabisa sub-district, KwaZulu Natal. Age standardized mortality rates
were calculated for the total population by gender. We investigated in a subset
of 15- to 50-year-old individuals (17,254) with known HIV status whether,
independent of age and separately for men and women, survival risk between HIV+
and HIV– individuals was modified by the period of observation, using
Cox regressions.
Results: In 234,740 person-years of observation between January 2003 and
December 2006, 3593 deaths occurred. The age-standardized mortality rate for 25
to 49 year olds, almost halved from 0.0338 (95%CI 0.030 to 0.038) deaths per
person year in the first semester of 2003 to 0.0186 (95%CI 0.014 to 0.023) in
the second semester of 2006. Controlling for age the HIV effect on risk of
dying in women was 3 times higher prior to than following the initiation of the
ART program (hazard ratios 19, 95%CI 13 to 29 vs 6, 95%CI 3 to 13). The smaller
decline seen in HIV+ men before and after ART program introduction was
not statistically significant.
Conclusions: We show a significant effect on adult mortality within 2 years of
ART program start in a high-prevalence community, even with incomplete coverage
of all those eligible for treatment; thus allowing a clear public health
message to encourage all who need treatment to access care. The lesser impact
on male mortality may be due to lower than expected uptake of ART by men, AIDS
mortality in men occurring in older age groups (for whom we currently do not
have HIV data) and a larger proportion of non-HIV-related deaths.
|
