Age in antiretroviral therapy programmes in South Africa: a retrospective, multicentre, observational cohort study.

BACKGROUND As access to antiretroviral therapy (ART) expands, increasing numbers of older patients will start treatment and need specialised long-term care. However, the effect of age in ART programmes in resource-constrained settings is poorly understood. The HIV epidemic is ageing rapidly and South Africa has one of the highest HIV population prevalences worldwide. We explored the effect of age on mortality of patients on ART in South Africa and whether this effect is mediated by baseline immunological status. METHODS In this retrospective cohort analysis, we studied HIV-positive patients aged 16-80 years who started ART for the first time in six large South African cohorts of the International Epidemiologic Databases to Evaluate AIDS-Southern Africa collaboration, in KwaZulu-Natal, Gauteng, and Western Cape (two primary care clinics, three hospitals, and a large rural cohort). The primary outcome was mortality. We ascertained patients' vital status through linkage to the National Population Register. We used inverse probability weighting to correct mortality for loss to follow-up. We estimated mortality using Cox's proportional hazards and competing risks regression. We tested the interaction between baseline CD4 cell count and age. FINDINGS Between Jan 1, 2004, and Dec 31, 2013, 84,078 eligible adults started ART. Of these, we followed up 83,566 patients for 174,640 patient-years. 8% (1817 of 23,258) of patients aged 16-29 years died compared with 19% (93 of 492) of patients aged 65 years or older. The age adjusted mortality hazard ratio was 2·52 (95% CI 2·01-3·17) for people aged 65 years or older compared with those 16-29 years of age. In patients starting ART with a CD4 count of less than 50 cells per μL, the adjusted mortality hazard ratio was 2·52 (2·04-3·11) for people aged 50 years or older compared with those 16-39 years old. Mortality was highest in patients with CD4 counts of less than 50 cells per μL, and 15% (1103 of 7295) of all patients aged 50 years or older starting ART were in this group. The proportion of patients aged 50 years or older enrolling in ART increased with successive years, from 6% (290 of 4999) in 2004 to 10% (961 of 9657) in 2012-13, comprising 9% of total enrolment (7295 of 83 566). At the end of the study, 6304 (14%) of 44,909 patients still alive and in care were aged 50 years or older. INTERPRETATION Health services need reorientation towards HIV diagnosis and starting of ART in older individuals. Policies are needed for long-term care of older people with HIV. FUNDING National Institutes of Health (National Institute of Allergy and Infectious Diseases), US Agency for International Development, and South African Centre for Epidemiological Modelling and Analysis.

Canine rabies in Australia: a review of preparedness and research needs

Australia is unique as a populated continent in that canine rabies is exotic, with only one likely incursion in 1867. This is despite the presence of a widespread free-ranging dog population, which includes the naturalized dingo, feral domestic dogs and dingo-dog cross-breeds. To Australia's immediate north, rabies has recently spread within the Indonesian archipelago, with outbreaks occurring in historically free islands to the east including Bali, Flores, Ambon and the Tanimbar Islands. Australia depends on strict quarantine protocols to prevent importation of a rabid animal, but the risk of illegal animal movements by fishing and recreational vessels circumventing quarantine remains. Predicting where rabies will enter Australia is important, but understanding dog population dynamics and interactions, including contact rates in and around human populations, is essential for rabies preparedness. The interactions among and between Australia's large populations of wild, free-roaming and restrained domestic dogs require quantification for rabies incursions to be detected and controlled. The imminent risk of rabies breaching Australian borders makes the development of disease spread models that will assist in the deployment of cost-effective surveillance, improve preventive strategies and guide disease management protocols vitally important. Here, we critically review Australia's preparedness for rabies, discuss prevailing assumptions and models, identify knowledge deficits in free-roaming dog ecology relating to rabies maintenance and speculate on the likely consequences of endemic rabies for Australia.