Failure to identify human immunodeficiency virus-seropositive newborns: epidemiology and enrollment patterns in a predominantly white, nonurban setting.

OBJECTIVE: To describe the epidemiology of newborn seroprevalence for human immunodeficiency virus (HIV) in a predominantly white, nonurban population, and to determine the factors associated with enrollment at a regional pediatric acquired immunodeficiency syndrome (AIDS) center serving that population. DESIGN: Retrospective case series of children enrolled at a regional pediatric AIDS center during a 6-year period and comparison with universal blind newborn screening data collected by the state of New York during the same time interval. SETTING: The Pediatric AIDS Center at State University of New York-Health Science Center at Syracuse, which serves as the only source of HIV-related pediatric care for children in a 16-country region of upstate New York totaling 1.8 million population. RESULTS: One hundred thirty-nine HIV-seropositive infants were born in the region during the 6-year study period; complete blind screening data were available for 138. Sixty-five (47%) of these infants were white. Thirty-nine (28%) of 138 had been enrolled at the Pediatric AIDS Center within the first 90 days of life. An additional 22 (16%) were enrolled at older than 90 days of life. The remaining 77 (56%) have never been seen at the center and are presumed to be unidentified. County enrollment rates varied from 0% to 100% and correlated with percent nonwhite births (r = .58; 95% confidence interval, 0.04-0.86). Children in outlying counties were at greater risk for nonenrollment than children from Onondaga County (site of the Pediatric AIDS Center) (adjusted relative risk, 1.38; 95% confidence interval, 1.05-1.85). White infants residing outside of Onondaga County were at the greatest risk of nonenrollment; of 50 seropositive white infants residing outside of Onondaga County, only 7 (14%) were enrolled at the center within the first 90 days of life. CONCLUSIONS: Local demographic factors can skew the racial distribution of HIV-seropositive infants dramatically compared with the national experience. White race and residence in counties away from the medical center each constituted risk factors for nonenrollment at the Pediatric AIDS Center. The epidemiology of HIV in this predominantly white, rural population, coupled with physician practices, probably contributed to low identification and enrollment rates. As the AIDS epidemic spreads into similar populations elsewhere, HIV infection in pregnant women or newborn infants is likely to become progressively harder to detect, unless universal screening is adopted.