Petrelis Files

CDC Admits Error: AIDS Over-Estimated for 2001 - 2005

Some very interesting HIV/AIDS revisions out of the Centers for Disease Control and Prevention in Atlanta this morning.

The federal agency is acknowledging over-estimating infections/full-blown cases in recent years, while also claiming the higher numbers did not affect federally-funded assistance programs for persons with HIV/AIDS.

Not sure how this was possible, since local health departments of public health and AIDS agencies surely used the original numbers to apply for and potentially receive more federal dollars. Knowing that the San Francisco DPH, how shall I put this, can be very creative when counting HIV infections and AIDS cases, it would not surprise me to learn other local DPHs have been as creative in tallying up such things, given how many millions of dollars are at stake.

I simply don't accept the CDC's claim that the over-estimates didn't mean more federal money for the DPHs, but I'll leave it to other government watchdogs to ask tough questions about the original AIDS stats and the downward revisions.

From the current CDC Weekly Morbidity and Mortality Report:

Each year the Division of HIV/AIDS Prevention of the National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention at CDC publishes an HIV/AIDS surveillance report. On June 28, 2007, CDC published online a revised 2005 report to correct an error in the calculation of the number of estimated AIDS cases for the period 2001--2005; this error resulted in an overestimate of 2% for the period 2001--2005 and 8.3% for the year 2005 . . .

CDC identified the error when reviewing computer programs in preparation for analyzing data for the 2006 report. An error in the computer code used to adjust for reporting delays in AIDS cases led to overestimation of AIDS cases for the period 2001--2005. Because 20% of AIDS cases typically are reported >1 year after diagnosis with AIDS, estimates are used to adjust for reporting delays and to allow for evaluation of data from recent years. The error has been corrected, and a complete external review of procedures for data estimation and validity checks has been conducted to ensure that this type of error does not occur again.

The error affected only estimated AIDS cases for the period 2001--2005, including those AIDS cases for which HIV and AIDS were diagnosed concurrently . . .

The error did not affect data used to allocate funding for major HIV/AIDS assistance and treatment programs administered by the Health Resources and Services Administration and the Department of Housing and Urban Development; only reported cases are used for allocation of resources administered via the Ryan White HIV/AIDS Treatment Modernization Act and Housing Opportunities for Persons with AIDS.

The overestimate of AIDS cases was 2% for the period 2001--2005 and 8.3% for 2005. Because some AIDS cases are diagnosed concurrently with HIV infection, a small overestimate of HIV/AIDS cases also occurred (0.4% for 2001--2005; 2% in 2005).

The corrected estimates for 2005 are 40,608 estimated AIDS cases for the 50 states and the District of Columbia and 37,331 estimated HIV/AIDS cases for the 33 states with longstanding HIV reporting.

Click here for the full announcement from CDC. This is also a good time to remind people about the recent estimated 33% fall in new HIV infections in San Francisco by the San Francisco health department:

While there was a ten percent decrease in total estimated new [HIV] cases, this seemingly modest decrease is actually a much greater prevention success than it appears.

From 2001 to 2006, the estimated number of gay men living in San Francisco increased from 46,800 to 58,343.

The increase was likely due to real growth in the gay community and, potentially, in part the result of an underestimation of the population size in 2001.

When the effect of the increase in the population size of MSM is taken into account, new infections have decreased by an estimated 33 percent. [Emphasis added]