Partner notification for HIV: running out of excuses.

RELUCTANCE TO CONDUCT partner notification (PN) for HIV infection may be ebbing after nearly 20 years at high tide. Two studies appearing in this issue of the journal ought to accelerate the ebb.1,2 They provide evidence that HIV PN neither promotes risky sexual configurations nor partnership violence or dissolution. These imagined fears ought to be the last in a long line of untested scenarios advanced by opponents of HIV PN since the beginning of the epidemic. In the early 1980s, opponents argued that there was no need to notify partners of AIDS cases because there was no test for infection. By mid-1985, availability of reliable HIV antibody tests helped shift grounds for objection to fear that PN would contribute to compilation by government agencies of lists of stigmatized individuals, especially gay men.3 Moreover, they argued that, with treatment not available, PN would needlessly anguish notified partners and perhaps cause an increase in suicide. After approval of the first antiretroviral drug (AZT [zidovudine]) in 1987, detractors opined that PN would be inordinately expensive4 and that, moreover, it would not be welcomed by high-risk populations and would be ineffectual because of partner anonymity.5,6 HIV control monies would in any case be more effectively spent on other (read: less personally intrusive) interventions. None of these objections was supported by experience and none of these fears materialized. By the mid-1990s, and with the advent of efficacious anti-HIV medications,7 this litany of concerns faded.8 Nevertheless, some detractors continued to imagine new scenarios to discourage PN. Notifying partners would “promote the breakup of old partnerships and increase the acquisition of new partners, thereby spreading HIV infections”1 and could stimulate domestic violence.9 With publication of the present studies, these final (?) imagined concerns, like their predecessors, bite the dust. Researchers in New Orleans and Denver coevally sought to determine the influence of PN on partnership stability and risky sexual pairing. In addition, the New Orleans study1 attempted to gauge the potential contribution of PN to partnership abuse and violence. That these separate investigations should exhibit commonalities may in part be due to researcher concurrency: at least two (T. A. P. and J. E. M. from the Centers for Disease Control and Prevention) of the combined 13 authors contributed to both. Notably, neither site sought to examine the impact of PN on patterns of injecting drug use, which I surmise is related to a dearth of eligible subjects. Similarity of results in both cities and heterogeneity of participants bode well for fans of generalizability. The stereotypic participant in New Orleans is black, heterosexual, and in a steady long-term relationship, while the stereotypic one in Denver varies ethnically, tends to be homosexual or bisexual, and reports more diverse relationships. Although index subjects and their comparisons differ in each study and although reported data are not strictly comparable, similarities emerge. In brief, both studies show that PN increases neither dissolution nor formation of partnerships, and both show its association with increases in reported condom use at follow-up. Although both sites report high partnership dissolution rates (about half to nearly two thirds), these likely represent background noise. Both report similar incident partnership formation rates (about 16%). Last, substantial decreases in reported emotional and physical abuse at 6-month follow-ups are associated with PN where measured (New Orleans). Both studies reveal weaknesses commonly experienced with PN populations: low enrollment interest and high follow-up loss. Only about two fifths of eligible subjects participated, and follow-up yielded populations largely representing presently nonpromiscuous people in main relationships. Although not noted by the authors, such bias may be a strength, because the outcomes of interest—partnership dissolution and abuse/violence in particular—may be of greater relevance within main partnerships. After all, who expects less than high dissolution rates with occasional partners? Or high abuse/violence rates with casual partners, since the usual association is with domestic partners? This is precisely what was observed in New Orleans: the likelihood of partnership dissolution is strongly related to casual and short-lived relationships. In essence, not only are results of these two studies consonant with anecdotal experience but also substantial participant loss at follow-up seems unlikely to have seriously biased the results and hence the conclusions. Reluctance to conduct PN has been part of the broader reluctance to implement standard public health interventions in response to the HIV/AIDS epidemic.8,10 Activism occurred early for public health–exempt status, an exemption “granted” by many public health authorities.11 Until recently, name-reporting, contactThe author thanks Richard B. Rothenberg, MD, MPH, for insightful discussion of the role of PN in risk-network intervention. Correspondence: John J. Potterat, 301 South Union Blvd, Colorado Springs, CO 80910. E-mail: jjpotterat@ earthlink.net Received February 28, 2002 and accepted July 10, 2002. Independent consultant, Colorado Springs, Colorado