
On Monday, March 24 the CDC released the 2006 HIV/AIDS Surveillance Report: Cases of HIV infection and AIDS in the United States and Dependent Areas. In this report, CDC states that the number of reported cases of HIV infection in 2006 based on 45 states and 5 U.S. dependent areas with confidential name-based HIV infection reported was 52,878.
The five states not included in this number are Hawaii, Maryland, Massachusetts, Montana, and Vermont. In 2005, the comparable number of reported HIV cases was 35,537 based on 38 states. For the first time the 2006 data includes the following seven newly reporting states: California, Delaware, Illinois, Maine, Oregon, Rhode Island, and Washington.
CDC has released this new surveillance report with seemingly no explanation of the dramatically increased figure or the differences in their reporting regarding the number of states from which the figure is based.
2006 Reported HIV/AIDS cases can be found on page 38. Thanks to AIDS Action for this. They plan on doing a further analysis and inquiry and will keep us - and you - informed.
What is the rate of new infections?...
ReplyDeleteHi Zak -- The new data principally shows that more states reported HIV cases by name than ever before. The 52,000 figure is "reported cases."
ReplyDeleteThe data breaks down "reported cases" by year of report (strictly, when public health entities received the name-based report forms) and also by year of diagnosis (when the individual learned his/her HIV+ status).
Both ways of looking at the data are limiting. Looking at the data by year of report strictly tells you when case reports were turned in. Looking at the data by diagnosis strictly tells you how many HIV+ people received testing and a diagnosis.
All this is to say, we don't yet know the rate of new infections because many transmissions are not diagnoses or reported.
In addition, the 2006 data is hard to compare against 2005 because of the new states. One can make some comparisons strictly for states with name-based data in 2005 and in 2006.
Is it true that if we had universal testing of the general population we would learn the rate of new infections?...
ReplyDeleteHello LifeLubers ---
ReplyDeleteI want to share this comment from Walt Senterfitt of the Community HIV/AIDS Mobilization Project - further discussing the CDC release and whether it shows a "spike" or not... Thanks - Jim de LifeLube
-------------
[From Walt]
Whoever put out the original headline about "Report Shows Large Spike" may have been honestly confused given the report's lack of clarity, but misread the data and did the community a disservice by feeding inaccurate sound-bite info sharing.
One bottom line of the report is that, among the 33 states which have had name-based HIV reporting at least since 2003 (and thus long enough for the data reporting to be consistent and reliable), the epidemic was stable or level. For instance the annual total for these states was 36,100 in 2003, dropped under 35,000 in 2004 and 2005 and was 36,800 in 2006. Among these cases, the proportion among men increased slightly (and amounts to 74% of the total) and the proportion among women has declined slightly. By race/ethnicity, the proportions among blacks and Latinos remained steady (49% and 18% respectively), while it has risen slightly among whites and Asians and Pacific Islanders and fluctuated among American Indians and Alaskan Natives. By transmission category, the proportion attributed to male-male sex has risen, whereas the proportion due to high-risk heterosexual sex has remained stable and the proportions due to injection drug use and to mother-to-child transmission have declined. Etc.
These are the only apples-to-apples and oranges-to-oranges comparisons of this report, and are actually rather interesting and useful.
The added tables reporting data from 7 additional states (including two with very large populations and large proportions of the overall epidemic) cannot be accurately compared with the other data for a variety of reasons, including that the data have not been adjusted for delays in reporting and year-of-report vs. year-of-diagnosis, but also are likely to include a one-time or short-term bulge in newly reported cases as always happens when an expanded surveillance system is put into place (remember the effects of the changes in the AIDS definition in 1993 to include low CD4 counts). So, it is natural to expect that there will be an apparent-but-misleading "spike" in terms of case reports as batches of "older" (prevalent) cases of HIV-but-not-AIDS are entered into the case report systems of the newly name-based states.
We badly need a complete and mature national system to monitor and track changes in the national epidemic as well as the new national incidence estimating system to be in place. Unfortunately it is at least 2 to 3 years away. In the meantime, it serves us best if we try to make our comparisons as precise, accurate and clear as possible -- even in our necessary agitational rhetoric. :-)
Walt Senterfitt, PhD
CHAMP
If we had universal testing of the general population could we learn who are the people exposed to human immunodeficiency virus who don't get infected?... the genetic combinations of the people who have resistance to HIV could be studied, that could help other people not get infected when exposed.
ReplyDelete"If we had universal testing of the general population could we learn who are the people exposed to human immunodeficiency virus who don't get infected?"
ReplyDeleteI don't think so because even the best "universal" screening programs would miss people and under-estimate the rate of new infections. What we really need are more sentinel studies. Basically population-based surveys to estimate the rate of infection in communities and populations. This is how HIV trends are monitored in most parts of the world. It's efficient and gives a better more representative picture of the epidemic than trying (and failing) to find everyone with HIV.
If we had universal testing of the
ReplyDeletegeneral population could we learn who
are the people exposed to human
immunodeficiency virus who don't get
infected?"
> I don't think so because even the
> best "universal" screening
> programs would miss people and
> under-estimate the rate of new
> infections.
I proposed a thought experiment and you
start in with something else. It's a
thought experiment. Assume that it's
possible.
> What we really need are more
> sentinel studies. Basically
> population-based surveys to
> estimate the rate of infection in
> communities and populations. This
> is how HIV trends are monitored
> in most parts of the world. It's
> efficient and gives a better more
> representative picture of the
> epidemic than trying (and
> failing) to find everyone with
> HIV.
But you still have to do testing and how
do select that population and how do you
get those people to cooperate? There are
groups like that. Anybody who goes into
the military gets tested.
Are you confounding speculation about
what you believe are the intentions of
this thought experiment with finding
everybody with HIV.