A recent study published in the Lancet seeks to give a statistical picture of HAART coverage in China and its impact on mortality from 2002 to 2009. It is based on a correlation of data from the Chinese national epidemiological database, which records positive HIV results, and a “treatment” database.
HIV affects an estimated 740,000 people living in China. The nature of the available data means that this study is restricted to the 145,000 individuals who are either “treatment eligible” because their CD4 count falls below a certain level (<350 cells/mm3 as from 2008; <200 prior to 2008), or reported as having AIDS or WHO stage 3 or 4 HIV/AIDS. Coverage here refers to coverage of this treatment eligible group.
Correlation of mortality with a range of risk factors reveals late diagnosis (as indicated by low CD4 count on diagnosis) and no HAART as by far the most significant predictors of mortality (adjusted hazard ratio of 7.92 for CD4 <50, 3.54 for CD4 50-199, 4.35 for no HAART). The study focuses on the correlation of mortality with HAART coverage, defined as proportion of person-years from treatment eligible date spent on HAART. The overall figures show a sharp increase in coverage (as defined above) between 2002 and 2009 from near 0% to 61.7%, and of a concomitant decline in mortality from 39.3 deaths per 100 person-years to 14.2 deaths.
Particularly interesting, however, is the picture which emerges from the stratification of the results by infection route (categorised as: blood/plasma transfusion; sexually infected; injecting drug user) – especially when these figures are set in the context of what this study tells us about central government interventions. A coverage of 80.2% in the blood/plasma transfusion group (as of 2009), concomitant with a mortality of 6.7 per 100 person-years, indicates the relative success of a raft of national government interventions directed at this population, including: wide-scale HIV screening in areas with large populations of former plasma donors; free antiretroviral treatment to patients with AIDS living in rural areas and to those in urban areas without insurance; free drugs to HIV-infected pregnant women; HIV testing of newborn babies. Coverage of 42.7% in injecting drug users and 61.7% in the sexually infected group (as of 2009), with concomitant mortality of 15.9 and 17.5 deaths per 100 person-years, evidently reflects the lower penetration of interventions into these groups.
The study provides statistical corroboration of the effectiveness of interventions in the blood/plasma donation group, and suggests the potential of HAART to improve outcomes among injecting drug users and those sexually infected, especially when offered before patients become severely immuno-suppressed.

Fujie Zhang et al., “Effect of earlier initiation of antiretroviral treatment and increased treatment coverage on HIV-related mortality in China: a national observational cohort study”, The Lancet, published online 19th May 2011

http://www.thelancet.com/journals/laninf/article/PIIS1473-3099(11)70097-4/fulltext