Summary of CDC incidence estimates

1
Summary of CDC incidence estimates

• Presentation to San Francisco HIV Prevention
  Planning Council (HPPC)
• August 14, 2008
• Grant Colfax, MD
• Director of HIV Prevention
• San Francisco Department of Public Health

2
Some definitions

• HIV incidence The number of people who were
  newly infected with HIV in a given period.
• HIV diagnoses The number of HIV diagnoses during
  a given period, regardless of when the infections
  occurred.
• HIV prevalence The proportion of a given
  population infected with HIV, regardless of when
  the infections occurred.

3
What changed

• CDC previously estimated there were 40,000 new
  infections in the US each year
• The new estimate is that there were approximately
  56,300 new infections in 2006
• The range of this estimate is 48,200-64,500
• Remember this is an estimate!
• It does not represent increasing incidence,
  rather it is a new estimate of incidence
• Published in August 6 2008 issue of JAMA

4
How did they do it?

• Used HIV test data from 22 states with mature
  named reporting
• Newly diagnosed 2006 HIV cases (33,802)
• Represent 73 of all AIDS cases diagnosed in the
  US in 2006
• California not represented
• Tested samples of persons newly diagnosed with
  HIV for recent infection
• Serologic testing algorithm to test for recent
  HIV seroconversion (STARHS)
• 30 of the sample without an AIDS diagnosis were
  detuned for recent infection
• Statistically adjusted for missing data
• After getting results for the 22 states, then
  extrapolated to other states and D.C. for
  national estimate

5
Number of Recent HIV Infections
incidence estimation
w
AIDS HAART
w
w
Incidence S weights x stat adjustment
w
w
w
w
w
w
w
w
Not STARHS tested
w
w
w
STARHS TESTED
w
w
w
R
NR
New HIV detected by testing
6
Results
7
(No Transcript)
8
New infections by race/ethnicity
9
Results by ethnicityHIV incidence per 100,000
10
What are next steps?

• Remember these data are national estimates
• The Epidemiology section, led by Dr. Susan
  Scheer, will be using the methodology of the CDC
  to calculate SF-specific incidence
• This will be based on SF data, however
• Like the CDC data, the data will not be complete
• The incidence number will be an estimate
• It will be another data point for us to consider
  along with all our other data in our planning
  efforts
• It is not a magic number that should supersede
  all our other data
• Unlike many jurisdictions, SF has many data
  sources to help guide our planning efforts