Estimating mortality based on maternal orphanhood in populations with HIV: a simulation study

(2021) Annual conference of the British Society for Population Studies — Location: Virtual (14.September.2021)

Files

No attached file found for this publication.

Details

Authors
Abstract
In countries lacking comprehensive systems of death registration, adult mortality rates are regularly estimated from parental survival through the orphanhood method. The HIV epidemic has introduced breaches in assumptions underpinning this method, for example due to increased correlation between maternal and child survival. This study evaluates the magnitude of HIV-related bias in orphanhood-based estimates and introduces new adjustments. A set of 576 populations with substantial mortality from AIDS are generated with microsimulations. Orphanhood estimates are compared to the underlying mortality levels. Revised coefficients to convert proportions of respondents with surviving parents into survivorship probabilities are calculated and applied to survey and census data from 16 African countries facing severe HIV epidemics. Without adjustment for HIV-related biases, the orphanhood method produces downward-biased mortality levels, with mean percentage errors reaching 37% in reports from young children aged 5-14 in simulations. Unbiased mortality rates can be obtained by combining two sets of proportions of parents surviving, and using revised coefficients accounting for trends in seroprevalence and coverage of antiretroviral therapy. In most countries with large HIV epidemics, this new approach provides estimates that are in agreement with those of the United Nations. Orphanhood-based estimates can fill data gaps on adult mortality in low- and middle-income countries, including those with high HIV prevalence. Further research is needed to correct for biases in male mortality inferred from paternal orphanhood.
Affiliations

Citations

Masquelier, B. (2021). Estimating mortality based on maternal orphanhood in populations with HIV: a simulation study. Annual conference of the British Society for Population Studies, Virtual. https://hdl.handle.net/2078.5/212449