Public health partner services (PS) seek to identify and notify the sex partners of persons with sexually transmitted diseases (STDs), including human immunodeficiency virus (HIV), with a goal of identifying undiagnosed persons, linking them to care, and assuring their treatment. Public health staff providing PS have traditionally conducted in-person interviews with infected persons, also known as original patients (OPs), to elicit the names of potentially exposed partners. The reason for using in-person meetings for PS is multifactorial, including confidentially concerns,1 facilitating collection of specimens for testing,2 providing treatment and linkage to medical care as appropriate,2 and the commonly held belief that in-person interviews are more effective.
The most recent Center for Disease Control guidelines for STD/HIV PS recommend that interviews should be conducted in person, but if no reasonable alternative exists, telephone interviews may be conducted.3 In King County, Washington, disease intervention specialists (DIS) routinely interview persons with HIV and syphilis over the telephone with safeguards in place to ensure privacy and confidentiality. For the most part, DIS only perform in-person interviews if a patient is in the Public Health-Seattle and King County (PHSKC) STD clinic or at the nearby affiliated hospital, if efforts to interview the OP over the telephone are unsuccessful, or if a person with newly diagnosed HIV does not successfully link to medical care. In-person interviews are commonly performed for patients at the STD clinic or the nearby affiliated hospital because DIS work onsite at the STD clinic. Other health departments have reported using telephone interviews for PS interviews, and some use email and text messaging for PS communications.4–8
Given resource constraints, it is important to maximize efficiencies in STD/HIV PS.9 Telephone interviews may be more efficient than in-person interviews; however, their relative effectiveness is uncertain. A survey of attendees of STD and community clinics as well as a community based organization about hypothetical PS found that there was no significant difference in favorability of in-clinic notification compared with telephone notification from the perspective of OPs or partners.10 Another study conducted in New York City found that a change in program policy that allowed DIS to conduct telephone interviews of partners exposed to HIV resulted in an increase in number of partners notified and tested, though the proportion of notified partners known to have tested declined.11 To our knowledge, no previous studies have compared the outcomes of in-person versus telephone PS interviews for OPs.
In this study, we compared in-person versus telephone PS interviews performed for OPs with early syphilis (ES) or new HIV diagnoses from 2010 to 2014 in King County, Washington to determine if there were differences in numbers of partners named, notified, tested, and newly diagnosed with infection. For ES diagnoses, we also evaluated the number of partners empirically treated and number of infected partners treated.
MATERIALS AND METHODS
Study Design and Data Sources
We conducted a population-based retrospective cohort study of OPs who underwent PS interviews in King County, Washington from 2010 to 2014. Data were extracted and deidentified from PS records completed by DIS at PHSKC, the state's PS data management system, and the enhanced HIV/AIDS Reporting System. The University of Washington Institutional Review Board defined the project as a program evaluation that did not require human subjects research review and approval.
The study population included all King County, Washington, residents reported to PHSKC with ES or newly diagnosed HIV infection who were interviewed by DIS for PS between 2010 and 2014. We excluded records from OPs with incomplete interviews. We defined cases as ES (primary, secondary, or early latent syphilis) based on the staging evaluation of a single DIS who reviews all syphilis cases for the county. We considered HIV diagnoses to be new if they were never previously reported in Washington State, were not defined to have been previously reported in another state based on Center for Disease Control deduplication procedures, and the patient denied a previous HIV diagnosis during their PS interview.
Study Data and Outcomes
Study data included information obtained through interviews and subsequent investigations with the OP and their partners. The primary exposure of interest for this analysis was OP interview method (telephone or in-person). DIS recorded this information using a checkbox on the PS interview. Covariates related to the OP included age, sex, race/ethnicity, sexual orientation, diagnosing clinic type, time from diagnosis to interview, HIV status (for ES only), and the DIS who performed the interview.
Data on partner outcomes were extracted from partner PS records and linked to the corresponding OP by case number. The outcome variables created for ES were number of partners named, notified, tested for syphilis, tested for HIV, empirically treated for syphilis, diagnosed with syphilis, diagnosed with HIV, and diagnosed with and treated for syphilis. The outcome variables created for newly diagnosed HIV were number of partners named, notified, tested for HIV, and diagnosed with HIV. To be counted as a named partner, the partner record had to include at least one of the following: first and last name, usable contact information of any kind (e.g. phone number, address, email, internet profile, etc.), or any known final disposition. Similar to previous PS program evaluations, we assessed each outcome as an index of total numbers of partners divided by total number of OP interviews.12,13 Each index therefore represents the mean number of partners with the outcome per OP interviewed. The six indices evaluated in this study were (1) contact index: number of partners named per OP interviewed, (2) notification index: number of partners notified per OP interviewed, (3) test index: number of partners tested per OP interviewed, (4) epidemiologic index: number of partners empirically treated per OP interviewed, (5) case-finding index: number of partners newly diagnosed with infection per OP interviewed, (6) brought-to-treatment index: number of infected partners treated per OP interviewed.
We used Stata 14.1 software for all statistical analyses.14 We first compared key covariates between OPs that underwent in-person interviews and OPs that underwent telephone interviews. Age and time from diagnosis to interview were evaluated as continuous variables; Wilcoxon rank sum tests were used to evaluate between-group differences in medians. All other variables were categorical and we used a Pearson χ2 test to evaluate between-group differences. For our notification, test, epidemiologic, case-finding and brought-to-treatment indices, we included outcomes based on either OP self-report or DIS confirmed outcomes. Although this approach may overestimate outcomes due to misreporting by OPs, the alternative, assuming all OP-reported outcomes are false, clearly underestimates PS outcomes. To address uncertainty related to this issue, for the syphilis brought-to-treatment index and the HIV case-finding index, we also calculated verified indices that only included outcomes that were verified by the health department through direct communication with the partner, direct communication with the provider, or medical record review. We used unadjusted Poisson regression models to calculate indices and relative risks (RR) with 95% confidence intervals (CI). We then used multivariable Poisson regression models, which adjusted for OP age, sex, race/ethnicity, sexual orientation, time from OP diagnosis to interview, diagnosing clinic type, and the DIS who performed the interview, to calculate adjusted RRs with 95% CIs. For ES outcomes, we also adjusted for HIV status. Next, we evaluated whether in-person interviews were conducted in the field or at the STD clinic and repeated the multivariable Poisson regression models to compare field-based interviews versus telephone interviews and STD clinic interviews versus telephone interviews. Lastly, we evaluated the association between OP age, sex, race/ethnicity, sexual orientation, and time from OP diagnosis to interview and the syphilis brought-to-treatment and HIV case-finding indices.
Clinicians and laboratories reported a total of 1,630 OPs with ES, of whom 1328 underwent a complete PS interview and were included in the analysis. A total of 682 OPs underwent in-person interviews and 646 OPs underwent telephone interviews (Table 1). Compared with persons interviewed by telephone, OPs interviewed in person were younger, interviewed sooner after syphilis diagnosis, and were more likely to be non-white, HIV negative, and diagnosed in the PHSKC STD Clinic or an affiliated community-based organization (CBO) (P < 0.001). The percentage of cases receiving in-person and telephone interviews varied by DIS (P < 0.001).
Compared with telephone interviews, in-person interviews resulted in significantly more partners per OP being named, notified, tested for syphilis, tested for HIV, and empirically treated for syphilis, with adjusted relative risks (aRRs) ranging from 1.19 to 1.68 (Table 2). Despite these differences, there was no significant difference between in-person and telephone OP interviews in syphilis case-finding or brought-to-treatment indices. Similarly, the HIV case-finding index among persons with ES did not vary significantly between OPs interviewed in-person and by telephone.
Of the 682 in-person interviews, 26 were conducted as part of field investigations and 624 took place in the STD clinic. We found fewer differences between field-based interviews and telephone interviews with no significant difference in the contact, syphilis test, HIV test, epidemiologic, syphilis case-finding, HIV case-finding, and syphilis brought-to-treatment indices. Field-based interviews were associated with a lower notification index than telephone interviews (aRR, 0.60; 95% CI, 0.38–0.98). The evaluation of STD clinic interviews versus telephone interviews resulted in the same findings as our primary analysis. Overall, white race (RR, 1.28; 95% CI, 1.01–1.62) and a time from diagnosis to interview of less than 14 days (RR, 1.44; 95% CI, 1.16–1.80) were associated with higher brought-to-treatment indices.
Newly Diagnosed HIV
Clinicians and laboratories reported 1167 new HIV diagnoses, of whom 959 underwent a complete PS interview. Of the 959 interviewed persons, 847 had method of interview (telephone or in-person) recorded and were included in the analysis. A total of 358 OPs underwent in-person interviews and 489 OPs underwent telephone interviews (Table 3). Compared with persons interviewed by telephone, those interviewed in-person were younger, interviewed sooner after HIV diagnosis, and were more often male, non-white, men who have sex with men, and diagnosed at the PHSKC STD Clinic or an affiliated CBO (P < 0.001). Compared with telephone interviews, in-person interviews resulted in significantly more partners per interviewed OP being named, notified, and newly diagnosed with HIV (Table 4). Adjusting for OP age, sex, race/ethnicity, sexual orientation, time from diagnosis to interview, type of diagnosing clinic, and DIS who performed the interview, in-person interviews were associated with a more than two-fold increase in HIV case finding. There was no significant difference between in-person and telephone interviews in number of partners tested for HIV per OP interviewed.
Of the 358 in-person interviews, 27 occurred as part of field investigations and 323 took place in the STD clinic. We found fewer differences between field-based interviews and telephone interviews with no significant difference in the notification, HIV test, and HIV case-finding indices. Field-based interviews were associated with a lower contact index than telephone interviews (aRR, 0.63; 95% CI, 0.40–0.97). The evaluation of STD clinic interviews versus telephone interviews resulted in the same findings as our primary analysis. There were no associations found between OP age, sex, race/ethnicity, sexual orientation, and time from OP diagnosis to interview and the HIV case-finding index.
There were no substantial differences between verified and unverified outcomes.
For in-person interviews, the verified syphilis brought-to-treatment index was 0.26 (94% of outcomes were verified) and the unverified brought-to-treatment index was 0.27. For telephone interviews, the verified syphilis brought-to-treatment index was 0.22 (91% of outcomes were verified) and the unverified brought-to-treatment index was 0.24. We found that all new HIV cases were verified, so the verified and unverified HIV case-finding indices were the same.
In this retrospective cohort study of OPs who underwent PS interviews in King County, Washington from 2010 to 2014, we found that in-person PS interviews for syphilis were associated with identifying, notifying, testing, and empirically treating more partners compared to telephone PS interviews, but were not associated with identifying and treating more partners diagnosed with syphilis. In contrast, in-person HIV PS interviews were associated with better outcomes for almost all metrics, including HIV case finding. We found fewer differences between field-based in-person interviews and telephone interviews; however, the small number of field-based interviews conducted by our PS program limited the power and utility of that analysis.
The overall case-finding indices and brought-to-treatment indices we report are consistent with those reported in previous studies. A study that compiled PS outcomes across the United States from 1976 to 2004 found that the median case-finding index for syphilis was 0.22 (range, 0.05–0.46) and for HIV was 0.13 (range, 0.03–0.75).12 Our overall syphilis brought-to-treatment index (0.26) was almost identical to this historical estimate, while our HIV case-finding index (0.07) was lower, though still within the range previously reported. More recent HIV case-finding indices from Texas (index = 0.06), Florida (index = 0.11) and North Carolina (index = 0.06) in 2015 are similar to those we report (personal communication with Emily Rowlinson, Dan George, and Erika Samoff).
Because in-person and telephone interviews for ES had similar brought-to-treatment indices, we interpreted our findings to indicate that in-person interviews were not clearly better for syphilis. Some persons looking at these data would emphasize that contact indices and epidemiologic indices were higher for in-person interviews and conclude in-person interviews were better. The issue is really whether they were superior enough to justify their cost. Different programs will likely come to different conclusions on this point based on available resources, the competing demands on their programs, the value they place on treating a small number of persons who may have incubating syphilis, and the extent to which their staff are effective in linking partners identified through syphilis interviews to other services (eg, HIV care or HIV preexposure prophylaxis).
Onsite DIS and decreased time from diagnosis to interview have been associated with improved PS outcomes.15,16 A study evaluating HIV PS outcomes in San Francisco found a case-finding index of 0.13 for OPs interviewed within 2 weeks of diagnosis compared with 0.05 for those conducted more than 2 weeks after diagnosis.17 Interestingly, our study found that in-person HIV PS interviews, which are typically performed by onsite DIS, occur a median of 10 days after diagnosis with a case-finding index of 0.10 and our telephone HIV PS interviews occur a median of 42 days after diagnosis with a case-finding index of 0.05. Although time to interview may have influenced our results, the differences we observed in case finding between in-person and telephone HIV PS interviews remained significant after adjusting for time to interview, suggesting that the timing of interviews did not entirely explain the differences we observed.
It is also possible that the significant psychosocial impact of receiving a new HIV diagnosis may be a reason why telephone interviews are not as effective as in-person interviews for HIV PS. A previous study found that HIV peer counselors perceive in-person interventions to be more effective than telephone interventions because face-to-face counseling allows for personal contact and facilitates trusting relationships.18 Another study revealed that patients notified of their HIV status by telephone delayed establishing primary care considerably longer than those notified in person,19 and it may be that HIV patients contacted by phone are reluctant to participate in detailed PS interviews.
The strengths of our study include the large, population-based sample size, the long-standing use of both in-person and telephone interviews for PS in our HIV/STD control program, and our use of multivariable models to adjust for baseline differences in characteristics between persons interviewed in-person and by telephone. However, our analysis has significant limitations. First and foremost, this is an observational study, and we cannot eliminate the possibility that our findings are influenced by residual confounding. In particular, the results for in-person interviews may be confounded by the fact that in-person interviews were used when telephone attempts were unsuccessful. However, this occurred infrequently, and we believe it is unlikely to have significantly influenced our results. Second, our results may have been influenced by ascertainment bias. Many persons diagnosed with STDs, including HIV, notify their partners without assistance from DIS. It is possible that PS interviews allow health departments to identify and link more partners to index patients, but that those same partners would have been notified and tested in the absence of any intervention. Insofar as in-person interviews identified more partners, it may be that this ascertainment bias is greater for in-person interviews than for telephone interviews. Indeed, we previously found that most partners identified through PS with newly diagnosed HIV infection were actually notified and tested before being contacted by DIS.20 A randomized trial of different approaches to PS interviews would overcome the problem of residual confounding but might still be affected by ascertainment bias. To date, only one very small randomized controlled trial has evaluated HIV PS in the United States, and we are not aware of any recent randomized trials related to syphilis PS.21 Given the significant programmatic costs associated with PS and the many competing demands on DIS time, it may be worth conducting trials to better assess the true benefit of the more expensive traditional approach that focuses on in-person interviews. Furthermore, our results may not be generalizable to other health jurisdictions. Predictors of PS success have been shown to vary by location.22 Perhaps even more importantly, very few of the in-person interviews undertaken by King County DIS involved field investigations; most persons were interviewed in our STD clinic, either because they were diagnosed there or came to the clinic after being diagnosed. As a result, our procedures were dissimilar from many other areas where most interviews are conducted in the field and are initiated within days of a program learning of a new HIV or syphilis diagnosis.
Many US health departments are trying to modernize their PS programs to evolve toward a broader model of field services. This effort requires reorganizing programs to confront a growing syphilis epidemic while simultaneously taking on new work providing HIV PS to all persons with newly diagnosed HIV infection, linking and relinking patients to HIV care, and promoting HIV pre-exposure prophylaxis. Success with this expanded portfolio of work will require willingness to reconsider traditional procedures. Our findings suggest that in-person interviews may not be superior to telephone interviews for syphilis, at least in terms of syphilis case finding and treatment of infected partners. Although our results suggest that in-person interviews are better for HIV, insofar as they involve field investigations, they are almost certainly substantially more costly. If the resources required for in-person interviews lead to fewer persons with newly diagnosed HIV being interviewed, the cost of the intervention could offset its greater efficacy leading to lower public health impact. Ultimately, how to allocate scarce DIS resources requires defining the costs and outcomes associated with different bodies of work and then setting priorities. We believe that our findings support health departments adopting a more flexible approach to PS, one that allows for more routine use of telephone PS for syphilis, and perhaps for some use of telephone interviews for HIV, particularly when such interviews facilitate future in-person contact, ensure linkage to care, or are the only viable option given resource constraints or patient's willingness to participate in PS. More rigorous, prospective studies of procedures are needed to define the most cost-effective methods for accomplishing the many tasks assigned to PS teams in the modern era of high-impact HIV prevention.
1. O'Connor J, Matthews G. Informational privacy, public health, and state laws. Am J Public Health 2011; 101:1845–1850.
2. Hogben M. Partner notification for sexually transmitted diseases. Clin Infect Dis 2007; 44(Suppl 3):S160–S174.
3. Centers for Disease Control and Prevention (CDC). Recommendations for partner services programs for HIV infection, syphilis, gonorrhea, and chlamydial infection. MMWR Recomm Rep 2008; 57(RR-9):1–83; quiz CE1–4.
4. Hightow-Weidman L, Beagle S, Pike E, et al. “No one's at home and they won't pick up the phone”: using the internet and text messaging to enhance partner services in North Carolina. Sex Transm Dis 2014; 41:143–148.
5. Kachur R, Adelson S, Firenze K, et al. Reaching patients and their partners through mobile: text messaging for case management and partner notification. Sex Transm Dis 2011; 38:149–150.
6. Udeagu CC, Bocour A, Shah S, et al. Bringing HIV partner services into the age of social media and mobile connectivity. Sex Transm Dis 2014; 41:631–636.
7. Ehlman DC, Jackson M, Saenz G, et al. Evaluation of an innovative internet-based partner notification program for early syphilis case management, Washington, DC, January 2007-June 2008. Sex Transm Dis 2010; 37:478–485.
8. Mendez J, Maher J. Evidence supporting the use of text messaging for partner services. Sex Transm Dis 2012; 39:238–239.
9. Hogben M, Collins D, Hoots B, et al. Partner services in sexually transmitted disease prevention programs: a review. Sex Transm Dis 2016; 43(2 Suppl 1):S53–S62.
10. Hennessy M, Williams SP, Mercier MM, et al. Designing partner-notification programs to maximize client participation: a factorial survey approach. Sex Transm Dis 2002; 29:92–99.
11. Udeagu CC, Bocour A, Shepard CW. For partner services, do we need “face-time,” or FaceTime?: trends in relative proportion of in-person notifications and HIV testing after introduction of a telephone option for HIV exposure notification by public health. Sex Transm Dis 2014; 41:671–673.
12. Brewer DD. Case-finding effectiveness of partner notification and cluster investigation for sexually transmitted diseases/HIV. Sex Transm Dis 2005; 32:78–83.
13. Holmes K, Sparling P, Stamm W, et al. Public health partner notification practices in high-income nations. In: Sexually Transmitted Diseases. 4th ed. McGraw-Hill: 2008.
14. StataCorp LP. Stata statistical software. 2015; Release 14.
15. Taylor MM, Mickey T, Winscott M, et al. Improving partner services by embedding disease intervention specialists in HIV-clinics. Sex Transm Dis 2010; 37:767–770.
16. Rudy ET, Aynalem G, Cross J, et al. Community-embedded disease intervention specialist program for syphilis partner notification in a clinic serving men who have sex with men. Sex Transm Dis 2012; 39:701–705.
17. Marcus JL, Bernstein KT, Klausner JD. Updated outcomes of partner notification for human immunodeficiency virus, San Francisco, 2004–2008. AIDS 2009; 23:1024–1026.
18. Messias DK, Moneyham L, Murdaugh C, et al. HIV/AIDS peer counselors' perspectives on intervention delivery formats. Clin Nurs Res 2006; 15:177–196.
19. Samet JH, Freedberg KA, Stein MD, et al. Trillion virion delay: time from testing positive for HIV to presentation for primary care. Arch Intern Med 1998; 158:734–740.
20. Golden MR, Stekler J, Kent JB, et al. An evaluation of HIV partner counseling and referral services using new disposition codes. Sex Transm Dis 2009; 36:95–101.
21. 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.
© Copyright 2017 American Sexually Transmitted Diseases Association
22. Hoots BE, Lewis FM, Anschuetz G, et al. Would targeting increase efficiency of syphilis partner services programs?—data from New York City, Philadelphia, Texas, and Virginia. Sex Transm Dis 2014; 41:407–412.