99/12/30

Vaccine Preparedness Study (VPS) Data:

  These data are taken from the VPS which was conducted to ascertain whether
it would be feasible to conduct large scale (phase III) HIV vaccine trials
in the United States.  A "high-risk" cohort was assembled and followed
prospectively for 18 months.  At baseline 3,257 seronegative gay men (or
MSM = men who have sex with men) were enrolled.  A baseline questionnaire
was administered and ascertained demographic information and recent sexual 
behavior information.  Behavioral measures are used to determine the 
potential HIV exposure risk profiles (since HIV is sexually transmitted).

  A subset of the data has been assembled to investigate whether drug and
alcohol behavior is associated with risk taking behavior.  The primary
outcome measures for this analysis include:

     * total number of HIV- male partners in the last 6 months
     * total number of HIV+ male partners in the last 6 months
     * total number of HIV:unknown male partners in the last 6 months

The drug and alcohol measures include:

     * any poppers use in the last 6 months
     * any amphetamine use in the last 6 months
     * current "heavy" alcohol use

Additional variables include:

     * site (integers 1-6) since recruitment was done in six cities
     * age
     * cohort (either subject was enrolled from previous study or was "new")


The data file "hiv.partner.data" contains the following variables:

   N.POS        = number of HIV+ male partners
   N.NEG        = number of HIV- male partners
   N.UNK        = number of HIV:unknown male partners
   SITE	        = integers 1-6 coding the six recruitment cities
   AGE	        = participant age (in years)
   COHORT       = 0 if from previous cohort; 1 if a new participant
   POPPERS      = 0 if no poppers use; 1 if any poppers use
   AMPHETAMINES = 0 if no amph use; 1 if any amph use
   ALCOHOL      = 0 if current use is < heavy; 1 if current use is heavy

A random subset of 2000 men has been selected for this data analysis.