Tom Brake: To ask the Secretary of State for International Development (1) what representations his Department has made to UNAIDS
about changes to future exercises estimating HIV/AIDS prevalence rates in sub-Saharan Africa, to include population-based surveys as
opposed to surveys which indicate trends in HIV prevalence; [156438]
(2) what representations his Department has (a) received from non-governmental organisations and (b) made to UNAIDS on
over-estimates of HIV/AIDS rates in sub-Saharan Africa by UNAIDS; [156784]
(3) what research his Department (a) has commissioned and (b) plans to commission into differences between data based
on population-based surveys and on ante-natal clinic-based surveys in giving reliable estimates of the level of HIV/AIDS infection
in sub-Saharan Africa; [156397]
(4) what representations his Department (a) has made and (b) plans to make to UNAIDS to include population-based HIV
survey data in UNAIDS' next report on the global HIV/AIDS epidemic. [156400]
Mr. Gareth Thomas: We have not received any representations from non-governmental organisations on over-estimates of HIV/AIDS rates
in sub-Saharan Africa. This Department does not plan to commission research into the differences between data based on
population-based surveys and on ante-natal clinic- based surveys in giving reliable estimates of the level of HIV/AIDS infection in
sub-Saharan Africa. UNAIDS has the mandate to provide information on the status of the epidemic.
UNAIDS included information about methodological issues relating to differences between data taken from these two different sources
in the Global Epidemic update, which was released last World AIDS day. They concluded that population based surveys are likely to
underestimate true HIV prevalence in most cases, because some respondents are absent from the household at the time of a survey and
some refuse to participate. UNAIDS also recognise that most antenatal clinic-based surveillance systems have limited geographical
coverage, which can lead to wide variations in the quality of the national estimate of HIV prevalence.
UNAIDS is now doing some analysis linking behavioural data to the surveys, to see whether and how that has affected the result of
the population based surveys. They will convene a meeting in May to analyse different results across a number of countries in sub-
Saharan Africa from the two sources of data. They aim to report findings at the World AIDS Conference in Bangkok in 2004.
We have not made any specific representations to UNAIDS about changes to future exercises estimating HIV/AIDS prevalence rates in
SA, to include population based surveys as opposed to surveys which indicate trends in HIV prevalence. DFID is in dialogue with them
on issues around data sources for HIV estimates. UNAIDS contend that conducting ante-natal clinic- based surveys and population
surveys at three to five year intervals can serve as valuable components of surveillance systems and can help improve estimates of
the levels and trends in HIV prevalence. This Department supports UNAIDS in their efforts to analyse the different results from
different sources, rather than relying on one source alone. All HIV estimates need to be assessed critically. Using different
sources for data and interrogating the results will always provide more robust analysis than relying on one source.