PARTNER NOTIFICATION (PN) is a public health intervention in which sex partners of infected individuals are informed that they may have been exposed to a sexually transmitted disease (STD), are offered healthcare services including testing and treatment, and are counseled about risk reduction. 1,2 PN can be conducted by healthcare providers or the index case (through a verbal contract with the provider) or by specially trained city, state, and federal employees called disease intervention specialists (DISs). PN programs were implemented in the United States in the 1930s as a method to control the spread of syphilis and began to be used for HIV infection prevention in 1985. 3
Most research on HIV PN to date has been process evaluation. Several studies have evaluated HIV PN in terms of the number of partners named, located, and found to be HIV-positive. 4–17 Other studies have examined the cost of PN. 18,19 Investigators in these studies concluded that it is a beneficial effort because infected persons were newly detected as a result of HIV PN and the cost per new infection identified did not outweigh the perceived benefit of detecting a case and counseling the person. These studies assumed that detecting a case and counseling would have a positive influence on behavior.
To date, few studies have examined the influence of partner notification on partnership transmission behaviors and mixing patterns. For example, other than anecdotal descriptions, little is known about the effect of HIV PN on the stability of relationships and on positive and negative outcomes that could occur as a result of HIV PN. PN has the potential for several negative outcomes. The discovery that one's sex partner is infected and that one has been exposed may cause the partner to blame the index case and lead to emotional abuse and/or physical violence. Given the incurable and potentially fatal nature of HIV infection, the potential for negative outcomes (such as violence and abuse) may be even more severe than for syphilis. In addition, it is possible that after HIV PN, a concordant partnership could end. This may cause the infected person to form a new partnership with a person who may be uninfected. If this is the case, then the spread of HIV could be facilitated by PN. An increase in any of these negative behaviors could indicate that PN may not be an effective method of disease prevention.
In addition to negative outcomes, there is also the possibility of several positive outcomes. One of the goals of PN is to provide counseling to individuals in the partnerships to promote risk reduction. Unlike HIV counseling and testing, PN provides the opportunity to offer counseling to members of the whole sexual network. If the counseling is successful, there will be an increase in condom use, an increase in sexual abstinence, and/or a decrease in the number of new sex partners acquired. If any of these situations is found to be true, it will provide more evidence that PN is a beneficial public health intervention.
Evaluation of the effectiveness of HIV PN is problematic. As in many areas of the United States, PN is the standard of care in New Orleans, and therefore a randomized trial is not possible. In addition, because people elect to participate in PN or not, no perfect control group exists for observational studies. Comparisons of partners who complete PN (i.e., the partner is actually found and notified) with those who do not are biased because index cases select which partner(s)’ names they give to the DIS, and the degree of accuracy of contact information depends on their willingness to have that person notified or their actual knowledge of that information. The result is that the PN-completed group is very different from the PN-noncompleted group.
Persons who undergo PN for syphilis may serve as a reasonable control. PN is routinely conducted for syphilis and is a generally accepted public health practice. HIV-infected persons and persons with syphilis and their partners are likely to come from the same pool of high-risk individuals. 19 But the ramifications of PN for the two diseases may be quite different because the nature of the diseases are so different. The effect of PN for syphilis might have fewer negative effects because, unlike HIV infection, the disease is curable and rarely leads to death. Partners’ reactions to PN may vary greatly because of the severity of HIV infection in comparison with syphilis.
The purpose of this study was to compare PN for HIV infection and syphilis in terms of dissolution of partnerships, acquisition of new sex partners, and the incidence of negative outcomes (physical and emotional violence) and positive outcomes (abstinence and condom use) following PN.
This was a prospective cohort study of partnerships in which the index case received PN services for HIV infection or syphilis via a DIS between April 1998 and July 2000 in the New Orleans metropolitan area. Information about the partnerships was provided by the index case through interviews administered at baseline and at 3 and 6 months.
Individuals whose antibody test for HIV was positive (i.e., by ELISA, with confirmation by Western blotting) or with early, primary, secondary, or early latent stage syphilis (as determined by rapid plasma reagin [RPR] or Venereal Disease Research Laboratory [VDRL] testing) diagnosed at public and some private clinics in metropolitan New Orleans were offered PN services during posttest counseling. In addition to receiving standard counseling about reducing high-risk transmission behaviors and initiating proper treatment, patients were asked to name their recent sex partners. For HIV infection PN, patients were asked to provide names and locating information for their sex partners during the previous 12 months, and for syphilis PN, patients were asked to provide similar information relating to the previous 3 months.
Patients were given the option of notifying their partner(s) themselves. If the patients chose not to notify their own partner(s) or chose to notify their partner(s) but the partner(s) did not come in for testing within the month, the DIS then attempted to contact each of the partners, notify them of their potential exposure, provide counseling about their risk behaviors, and offer them testing for HIV infection or syphilis.
For confidentiality, at the time of PN, the DIS described the study to the index patient. If the index patient agreed to participate, his/her name was given to the study personnel, who then subsequently contacted, obtained informed consent from, and enrolled the patient.
HIV infection or syphilis index cases, within 1 month of having received PN services, were eligible for this study unless they met any of the exclusion criteria: being less than 18 years of age, residing in jail or prison, living outside the New Orleans metropolitan area, or not having or refusing to name a sex partner from the previous 3 months. Because the purpose of this study was to assess changes in sex partnerships and risk behaviors, these outcomes could not be ascertained for individuals who reported that they had no present sex partner. Institutional review boards from all participating institutions approved the study.
The primary source of data collection for this study was an interviewer-administered questionnaire. Questions were drawn from other surveys used in local HIV research projects, and other questions were developed and added as necessary to address the specific hypotheses of this particular study. The questionnaires were then pilot-tested for content validity.
The baseline and follow-up questionnaires consisted of two parts: a section specific to the respondent and a section specific to each of the respondent's partners. At baseline, respondents were asked about their sex partners and behaviors during the 3 months before partner notification. At the months 3 and 6 visits, respondents were asked about their behaviors during the past 3 months with each of the partners they previously reported, and they were also asked about their behaviors with any new sex partners they acquired during that 3-month period. Participants were given a $15 incentive at each of the follow-up visits.
Definitions and Classifications
There were six main outcome variables: partnership dissolution, acquisition of a new sex partner, sexual abstinence, condom use, emotional abuse, and physical violence. To ascertain if the partnership dissolved, index cases were asked, “During the last 3 months, did you have sex or continue having a relationship with this person?” If the person said no, they were probed, “Do you still do anything with this person?” If the index case responded no to both questions, the partnership was considered dissolved. Emotional abuse was identified by the question “During the last 3 months, how often were you criticized, put down, or yelled at by this partner?” If the respondent mentioned “sometimes” or “a lot” of times, then the partnership was classified as including emotional abuse. Physical abuse was determined by the question “During the last 3 months, how often were you slapped, punched, shoved, kicked, or otherwise physically hurt by this partner?” If the respondent mentioned “sometimes” or “a lot” of times, then the partnership was classified as including physical violence. Partnerships were considered abstinent if the index case reported that their partnership had not dissolved and that they had engaged in no vaginal or anal sex in the last 3 months. Consistent condom use was defined as use 100% of the time within a partnership in the last 3 months. Acquisition of a new sex partner was determined if the index case said he/she was having sex with a new partner during follow-up that was not mentioned at baseline.
The main predictor was the disease status (HIV infection or syphilis) for which PN was performed. If a participant was both an index case and a partner (e.g., the individual was named as a partner and then tested positive), she or he was classified as an index case. This occurred for nine of the HIV-infected persons and 10 persons with syphilis. If a participant fell into both the HIV infection and syphilis categories (i.e., the individual tested positive for both HIV infection and syphilis), then she or he was classified in the HIV infection group. HIV concordance or discordance was determined by the index person's perception of the disease status of partners. A main partner was defined as “someone who is special to you, like a husband/wife/boyfriend/girlfriend or lover.” An occasional partner was defined as “an occasional sex partner who is not your main sex partner, whether you had sex only once or many times.” PN was considered completed if the DIS recorded that the partner was contacted and notified.
For evaluation of behavioral outcomes of interest (i.e., physical violence, emotional abuse, abstinence, and condom use), the partnership was the unit of analysis. Because each person could have more than one observation per partnership, generalized estimating equations (GEEs) were used to adjust for multiple observations per partnership. GEE is used for regression analysis when there is dependence among observations. 20–22 Exchangeable correlation matrices and robust estimators of variance were used; parameter estimates were exponentiated to provide odds ratios (ORs) and 95% confidence intervals (CIs). Unadjusted ORs were calculated with use of GEEs to evaluate the unadjusted association between independent variables and outcomes (Table 4). Associations between the outcomes of interest and disease status (i.e., HIV infection or syphilis) were then adjusted for time of observation and stability of partnership (i.e., main or casual partner).
Factors associated with dissolution of the partnership were also explored, with use of logistic regression, with the partnership again the unit of analysis. Predictors of acquiring a new sex partner were also examined. For this analysis, the index case was the unit of analysis. Logistic regression was used to evaluate factors associated with acquisition of a new sex partner. Characteristics of HIV infection and syphilis index cases were compared with the likelihood ratio chi-square test and Fisher's exact test as necessary. All statistical analyses were conducted with STATA version 7.0 (STATA, College Station, TX).
All HIV infection and syphilis index cases newly reported to the Office of Public Health for the study period were screened. Of 318 HIV infection index subjects screened, 208 were eligible. Of those HIV infection index subjects who were eligible, 93 refused and 39 were lost to follow-up, leaving 76 for analysis. Of 277 syphilis index subjects screened, 221 were eligible. Of those syphilis index subjects who were eligible, 81 refused and 59 were lost to follow-up, leaving 81 for analysis (Figure 1). Of the ineligibles, 76% of HIV infection and 41% of syphilis index cases were classified as such because they did not name a partner. Syphilis index subjects were slightly more likely than HIV infection index subjects to be excluded from the analysis because they were lost to follow-up (27% versus 19%: P < 0.01).
Subjects who were lost to follow-up (n = 98) were more likely to be non–African American than African American (60% versus 23.1%;P < 0.01), although the percentage of non–African Americans was only 6% of the population. They did not, however, differ in disease status, age, gender, sexual orientation (for males), marital status, education, employment status, pregnancy status (for females), or baseline behaviors (i.e., condom use, injection drug use, number of sex partners).
Persons included in the analysis (n = 157) were more likely than those who were excluded (n = 272) to have syphilis rather than HIV infection (38.5% versus 25.2%;P < 0.03), to be drug-users (i.e., users of marijuana, crack, or any injection drug) versus non-drug-users (36.7% versus 19.9%;P < 0.02), and to report only one partner rather than multiple partners (70.0% versus 28.0%;P < 0.006). The groups did not differ significantly in race, age, marital status, pregnancy (for females), sexual orientation (for males), or employment status.
Description of the Index Cases
A total of 76 HIV infection and 81 syphilis index cases were included in the analysis. Index cases were mostly African American (93.6%), >30 years of age (60.5%), and had a monthly income of <$500 per month (64.9%). The majority of index cases reported no injection drug use in the last 3 months (99.7%) and named only one sex partner (71.9%) in the last 3 months. Condom use at last sex act was reported for 36.9%, and 15.3% reported they had received PN at least once before this PN event. The HIV infection and syphilis index cases were similar in age, race, education, current employment status, self-reported monthly income, self-reported injection drug use in the past 3 months, and condom use. HIV infection index cases were more likely to be male, be homosexual or bisexual, be married, and report having only one sex partner in the past 3 months than were syphilis index cases (Table 1).
Description of HIV Infection and Syphilis Partnerships at Baseline
A total of 220 partnerships (94 HIV infection and 126 syphilis cases) fit the eligibility criteria, had follow-up data available, and were included in the analyses. Partnerships mostly were heterosexual (85.9%), defined as a main partnership (64.1%), of ≥1 year's duration (73.3%), and non-drug-sharing (92.7%). At baseline, participants’ reports indicated that 27.3% lived with the partner, 14.6% had children with the partner, 63.6% depended on the partner for money or emotional support, 69.6% were satisfied with the relationship, and 60.0% thought the relationship was important. According to enrollment criteria, all were involved in at least one sexually active relationship at baseline, 35.9% used a condom at last sex act, 42.3% experienced emotional abuse, and 23.6% reported physical violence. HIV infection partnerships were more likely than syphilis partnerships to be same-sex partnerships (24.5% versus 6.3%;P < 0.05). Stability of relationship, duration of relationship, economic dependency, baseline emotional abuse, physical violence, and condom use were not different for HIV infection versus syphilis partnerships.
The PN process was completed (i.e., the DIS documented contacting and notifying the partner) for 32.7% of partnerships (31.9% of the HIV infection and 33.3% of the syphilis partnerships [P > 0.80]). For both HIV infection and syphilis, partnerships for which PN was completed were more likely to be main and cohabiting partnerships. For HIV infection, the index cases of partnerships for which PN was completed were also more likely to report that they depended on the partner financially or emotionally than those for whom PN was not completed. For syphilis, index cases of partnerships for which PN was completed were also more likely to report that they were satisfied with the relationship and that they did not break up after PN, in comparison with those for whom PN was not completed. (Table 2).
Stability of Partnerships After PN
Of the 220 original relationships, 46.8% dissolved. Dissolution was similar for HIV infection and syphilis partnerships (45.7% for HIV infection and 47.6% for syphilis;P > 0.72). However, partnerships for which PN was completed were less likely to dissolve during follow-up than those for which PN was not completed (24.3% versus 75.7%;P = 0.012). Partnerships were more likely to dissolve if they had incomplete PN (OR, 2.08; 95% CI, 1.16–3.70), were occasional partnerships (OR, 2.86; 95% CI, 1.62–5.07), or were partnerships with a duration of less than 1 year (OR, 1.89; 95% CI, 1.03–3.45). Age, gender, and education of index case, sexual orientation (for males), concordance (perceived or actual for HIV infection partnerships only), and baseline violence were not associated with dissolution. Disease status was not associated with dissolution of partnership.
Of the 157 index cases, 15.9% overall acquired a new partner (14.5% HIV infection cases and 17.3% syphilis cases;P < 0.64). Logistic regression was conducted to examine factors associated with acquiring a new partner. Men were 2.88 times more likely than women (95% CI, 1.16–7.19) to acquire a new partner during follow-up. Race, age, marital status, employment, sexual orientation, education level, and disease status were not statistically associated with acquisition of a new partner in unadjusted analyses.
Behavior After PN
Emotional abuse and physical violence were experienced at least once over follow-up in 23.5% and 8.8% of the partnerships, respectively. There was no difference, with regard to disease status, in emotional abuse (24.2% HIV infection versus 20.5% syphilis) and physical violence (9.1% HIV infection versus 7.7% syphilis) over follow-up. Baseline partnerships in which emotional abuse was reported were no more likely to dissolve than baseline partnerships in which emotional abuse was not reported (4.2% versus 15.7%;P > 0.14), and baseline partnerships who reported physical violence were no more likely to break up than partnerships that reported no physical violence (0.0% versus 14.4%;P > 0.20). These findings were similar for both HIV infection and syphilis baseline partnerships. In overall prevalence, both emotional abuse and physical violence decreased over time (Table 3).
Of the partnerships reporting sex at least once during follow-up, 42.7% reported using condoms at the last sex act. HIV infection partnerships were more likely than syphilis partnerships to report condom use at last sex act (61.7% versus 30.4%;P < 0.001). Of partnerships present at baseline with at least one follow-up, 20.6% reported abstaining from sex at the last available follow-up visit (28.6% HIV infection, 13.8% syphilis;P < 0.06). Abstention from sex increased over time for both HIV infection and syphilis (Table 3).
Unadjusted associations of various predictors with the five outcomes of interest (i.e., dissolution, abstinence, condom use, physical violence, and emotional abuse) were determined (Table 4). Partnerships that dissolved were less likely to have PN completed, more likely to be occasional and not married, and more likely to dissolve soon after PN. Abstention from sex was more frequent among partnerships for which PN was completed and was more likely to happen immediately after PN and then diminish over time. The only factor associated with condom use at last sex act was if the partnership involved HIV infection rather than syphilis. Physical violence and emotional abuse decreased over time but were not associated with any of the other factors examined. After adjustment for time of observation, disease status was not associated with any of the five outcomes. Main partnerships were less likely to dissolve and to use condoms than were occasional partnerships.
To ensure that the observed results were not excessively influenced by the subjects who did not have both follow-up visits, a sub-analysis was conducted including only partnerships that had both follow-up visits. Unadjusted ORs for characteristics associated with dissolution, abstaining from sex, condom use, and violence (physical or emotional) echoed findings for the whole sample, with confidence intervals somewhat tighter overall. Findings were similar for the adjusted analyses, except for the finding that HIV infection status became significantly associated with condom use at last sex act (OR, 1.90; 95% CI, 1.09–3.32;P < 0.05). The results of this sub-analysis confirm the overall results and suggest that subject loss did not bias the results.
The purpose of this study was to determine the effect of PN on relationship stability and risk behavior for HIV infection versus syphilis. We were particularly interested in examining whether HIV PN caused excess partnership dissolution. If this were the case, then HIV infection index cases who underwent PN might acquire new, potentially uninfected partners and therefore amplify the spread of the infection. We did not find more dissolution of partnerships post-PN for HIV infection than for syphilis.
We used syphilis cases for comparison because syphilis is treatable and we thought that relationship stability and risk behavior would not be affected much by PN; therefore, this group could serve as a representative control. The perfect controls would have been HIV infection index cases who did not complete PN. However, index cases may be more likely to name stable partners because they can give better locating information, so comparisons would not be very meaningful. Since we did not have a perfect control group, we chose to compare the experience of PN among HIV infection partnerships versus syphilis partnerships. We did not find excess violence, and whereas the rate of dissolution was high, it was no higher for HIV infection than for syphilis, and HIV infection index cases were no more likely to acquire a new partner than were syphilis index cases. Studies of similar populations in the United States suggest that dissolution rates in general are high. 23,24 This suggests that HIV PN may not have much influence on the breakup of partnerships.
The overall rates of dissolution (i.e., 46.8%) and new partner acquisition (i.e., 15.9%) are high and merit attention. Although we do not know the disease status of these newly acquired partners, it is clear that ongoing notification of partners is needed. This may be accomplished by several methods, including continued follow-up by DIS or by providing index cases with disclosure skills.
Since this was an observational study and almost 50% of the eligible persons refused to enroll, the findings may be biased. Because we have very little information on the partnerships for which the index case refused to enroll, quantifying this bias is not possible. Enrolling people in a research study at the moment they discover their disease status is a daunting task. Participation may be lower for HIV infection than for syphilis because of the severity of the infection. We chose to try to recruit as soon after PN as possible (i.e., within 1 month) to reduce recall bias and to ensure that we could find the potential participants again (since this population is highly mobile). This may have resulted in improved recall but low recruitment. The inclusion of the small monetary incentive may have helped in recruiting otherwise hard-to-reach people such as drug users, 25 but this improved recruitment did not translate into improved follow-up.
It is important to note that despite intensive efforts by the DIS to complete the PN process, only about one third of the partners index cases told us about were notified of their disease exposure by DIS. This may be an underestimation, given that index cases do not always reveal the names of all of their partners. When self-reported disclosure of disease status by the index case was added, the percentage of partners who were notified of their exposure grew to two thirds (data not shown). Although the self-reported disclosure could not be verified, some PN is occurring as a result of index self-disclosure. Whereas DIS-delivered PN is effective for some partnerships, it may not work for all partnerships. Much of the difficulty with PN is that index cases often offer only pseudonyms with vague or nonexistent addresses or initials. This makes it impossible for the DIS to locate the partner. Another problem is that generally PN is offered only once (after index notification), and therefore new partnerships do not benefit from this service.
However, HIV infection partnerships were more likely to use condoms and to abstain from sex post-PN than were syphilis index cases. While is it not certain that this is a direct result of PN, these are the counseling messages that the DIS gave, and therefore it is likely that PN has some influence.
This study is among the first to evaluate the effect of PN on partnership stability and transmission behaviors. Published studies thus far have examined process evaluation of HIV PN and have not compared HIV PN to syphilis PN. 26–30 This comparison provides us with important information on post-PN partnership behavior for a curable disease versus a noncurable disease. We found that PN for HIV infection did not cause more dissolution of partnerships or other negative outcomes such as violence and new partner acquisition than did PN for syphilis. Behavior change to reduce HIV transmission (e.g., abstention from sex) was more common among partnerships in which PN was completed. These data add to the growing body of evidence that there is a public health benefit to HIV PN.
1. Centers for Disease Control and Prevention. Revised guidelines for HIV counseling, testing and referral. MMWR Morb Mortal Wkly Rep 2001; 50(RR-19).
2. Parran T. Shadow on the Land. New York: American Social Hygiene Association, 1938.
3. Centers for Disease Control and Prevention. HIV Partner Counseling and Referral Services Guidance. December 30, 1998.
4. Rutherford GW, Woo JM, Neal DP, et al. Partner notification and the control of human immunodeficiency virus infection: two years experience in San Francisco. Sex Transm Dis 1991; 18: 107–110.
5. Pavia AT, Benyo M, Niler L, Risk I. Partner notification for control of HIV: results after 2 years of a statewide program in Utah. Am J Public Health 1993; 83: 1418–1424.
6. Spencer NE, Hoffman RE, Raevsky CA, Wolf FC, Vernon TM. Partner notification for human immunodeficiency virus infection in Colorado: results across index case groups and costs. Int J STD AIDS 1993; 4: 26–32.
7. Landis SE, Schoenbach VJ, Weber DJ, et al. Results of a randomized trial of partner notification in cases of HIV infection in North Carolina. N Engl J Med 1992; 326: 101–106.
8. Hoffman RE, Spencer NE, Miller LA. Comparison of partner notification at anonymous and confidential HIV test sites in Colorado. J Acquir Immune Defic Syndr Hum Retrovirol 1995; 8: 406–410.
9. Giesecke J, Ramstedt K, Granath F, Ripa T, Rado G, Westrell M. Efficacy of partner notification for HIV infection. Lancet 1991; 338: 1096–1100.
10. Pattman RS, Gould EM. Partner notification for HIV infection in the United Kingdom: a look back on seven years experience in Newcastle upon Tyne. Genitourin Med 1993; 69: 94–97.
11. Wells KD, Hoff GL. Human immunodeficiency virus partner notification in a low incidence urban community. Sex Transm Dis 1995; 22: 377–379.
12. Fenton KA, French R, Giesecke J, et al. An evaluation of partner notification for HIV infection in genitourinary medicine clinics in England. AIDS 1998; 12: 95–102.
13. Daniell FD, Skelly RR. Partner notification in cases of HIV infection. N Engl J Med 1992; 327: 435–436.
14. Ramstedt K, Hallhagen G, Lundin B, et al. Contact tracing for human immunodeficiency virus (HIV) infection. Sex Transm Dis 1990; 17: 37–41.
15. Worm A-M, Smith E, Sorensen H, Haxholdt H. Partner notification for HIV infection in Denmark: attitudes and preliminary results. Genitourin Med 1996; 72: 283–285.
16. Wykoff RF, Jones JL, Longshore ST, et. al. Notification of the sex and needle-sharing partners of individuals with human immunodeficiency virus in rural South Carolina: 30 month experience. Sex Transm Dis 1991; 18: 217–222.
17. Louisiana Office of Public Health. AIDS update: counseling of sex partners. Louisiana Morbidity Report 1997; 8: 5.
18. Holtgrave DR, Valdiserri RO, Gerber R, Hinman AR. Human immunodeficiency virus counseling, testing, referral and partner notification services: a cost-benefit analysis. Arch Intern Med 1993; 153: 1225–1230.
19. Varghese B, Peterman TA, Holtgrave DR. Cost-effectiveness of counseling and testing and partner notification: a decision analysis. AIDS 1999; 13: 1745–1751.
20. Liang K-Y, Zeger SL. Regression analysis for correlated data. Ann Rev Pub Health 1993; 14: 43–68.
21. Stokes ME, Davis CS, Koch GG. Categorical Data Analysis Using the SAS System. 2nd ed. Cary, North Carolina: SAS Institute, 1998.
22. Zeger SL, Liang K-Y. Longitudinal data analysis for discrete and continuous outcomes. Biometrics 1986; 42: 121–130.
23. Finer LB, Darroch JE, Singh S. Sexual partnership patterns as a behavioral risk factor for sexually transmitted diseases. Fam Plan Perspect 1999; 31: 228–236.
24. World Health Organization. Consensus statement from consultation on partner notification for preventing HIV transmission. J Acquir Immun Defic Syndr 1989; 2: 248–255.
25. Kissinger P, Kopicko J, Myers L, et al. The effect of modest monetary incentives on follow-up rates in sexually transmitted disease studies. Int J STD AIDS 2000; 11: 27–30.
26. Centers for Disease Control and Prevention. Partner notification for preventing human immunodeficiency virus (HIV) infection: Colorado, Idaho, South Carolina, Virginiam. MMWR Morb Mortal Wkly Rep 1988; 37: 393–402.
27. Fenton KA, Peterman TA. HIV partner notification: taking a new look. AIDS 1997; 11: 1535–1546.
28. West GR, Stark KA. Partner notification for HIV prevention: a critical reexamination. AIDS Educ Prev 1997; 9 (suppl B): 68–78.
29. Oxman AD, Scott EAF, Sellors JW, et al. Partner notification for sexually transmitted diseases: an overview of the evidence. Can J Public Health 1994; 85 (suppl 1): S41–S47.
30. Gorbach PM, Aral SO, Celum C, et al. To notify or not to notify: STD patients’ perspectives on partner notification in Seattle. Sex Transm Dis 2000; 27: 193–200.