J. Pitt et al., HIV VERTICAL TRANSMISSION RATE DETERMINATIONS ARE SUBJECT TO DIFFERING DEFINITIONS AND THEREFORE DIFFERENT RATES, Journal of clinical epidemiology, 51(2), 1998, pp. 159-164
The HIV infection status of a cohort of 600 prospectively followed chi
ldren horn to HIV infected mothers was determined using HIV peripheral
blood culture tests at 0, 3, and 6 months of age, HIV serology at gre
ater than or equal to 15 months, and CDC AIDS criteria. We estimated t
ransmission rates using five methods which differed in how HIV indeter
minates are handled. These methods were applied at two points in time
to illustrate effects of length of follow-up of the cohort on results.
In January 1997, 30 months after the last birth, transmission rate es
timates ranged from 15.5% (known positives/known positives x known neg
atives) to 18.1% (known positives x those with one positive culture x
deaths/entire cohort minus those lacking negative cultures at age grea
ter than or equal to 5 months). Estimates ranged from 14.8% to 20.7% u
sing the subcohort of 284 children followed greater than or equal to 1
2 months as of May 1993. These results indicate that methods for assig
ning HIV infection status and for handling HIV indeterminates should b
e carefully defined when estimating transmission rates. (C) 1998 Elsev
ier Science Inc.