RT Journal
A1 Jarlais D, Friedman SR
T1 MOdeling the spread of hiv-reply
JF JAMA
JO JAMA
YR 1989
FD September 8
VO 262
IS 10
SP 1330
OP 1331
DO 10.1001/jama.1989.03430100063026
UL http://dx.doi.org/10.1001/jama.1989.03430100063026
AB In Reply.— 
    In our article on human immunodeficiency virus (HIV) exposure among intravenous (IV) drug users in Manhattan, we noted that the stabilization of seroprevalence rates has occurred at a level—approximately 50%— where simple models of epidemics predict the greatest rate of new infections. Salzberg et al propose a model that is consistent with the seroprevalence data we reported. It is not consistent with other data, both contained in the article and published elsewhere. The largest problem with the model of Salzberg et al is the use of a static conceptualization of IV drug use at the individual and population levels. The risk status of individual IV users is assumed to remain constant, despite the known variability in frequency of both heroin and cocaine injection among persons considered to be addicts.1 Any classification of IV drug users into constant levels of risk behavior over long periods must be treated with