Health chief says HIV figures were not manipulated

CAPE TOWN – The health department has denied suggestions that officials manipulated the 2007 antenatal HIV survey to paint a rosier picture of SA’s epidemic. On Monday, the South African Medical Journal published a letter from two of SA’s leading demographers, saying the department introduced a new methodology to analyse the 2007 figures, but did not apply it retrospectively to 2006 before comparing numbers. The report said overall HIV prevalence among pregnant women fell from 29,1% in 2006 to 28% last year. The scientists reassessed 2007 data, stripping out the methodological change they identified, and calculated a new prevalence of 29,4%, marginally up on the previous year. Health director general Thami Mseleku said that any claim the figures were changed to manipulate information is an insult to the integrity and ethics of the department. We stand by our methodology, which has not changed (from 2006 to 2007), he said. Officials consulted United Nations and Medical Research Council experts in analysing the data, he said. The only methodological change in the survey, conducted yearly since 1990, was expanding the sample size in 2006. Mseleku said the health department would convene an expert meeting, including the authors of the letter, University of Cape Town professors David Bourne and Rob Dorrington, to discuss their claims that the report was flawed. Dorrington said he would be more than happy to attend. Mseleku would not explain why the report’s figures for Western Cape appeared contradictory. He not want to discuss it in the press. The report said HIV prevalence among Western Cape pregnant women fell from 15,1% to 12,6% from 2006 to last year, a fall of 2,5 percentage points. Bourne has said that as no Western Cape district recorded a drop greater than one percentage point, the decline was a mathematical impossibility.