Perceived barriers to partner notification among sexually transmitted infection clinic patients, Cape Town, South Africa

Abstract

Background

Public health approaches to sexually transmitted infection (STI) prevention in resource-limited countries rely on patients to self-notify sex partners of their STI. However, a majority of partners go uninformed and remain untreated and infectious.

Methods

Anonymous surveys collected from 776 men and women receiving STI clinic services in Cape Town, South Africa.

Results

Half of patients surveyed intended to inform their partners, while half did not intend to notify partners. Women were more likely than men to intend to notify their partners. Patients who completed formal education were also more likely to indicate intentions to notify partners. There were no associations between numbers of partners patients had or partner types with intentions to notify partners. Among both men and women, concerns about adverse partner reactions were associated with intentions not to notify partners. Multivariable analyses stratified by gender and controlling for confounds showed that intentions to notify partners were significantly related to men's concerns that their partner could react violently against them and women's concerns that their partner may leave them and refuse to see them again.

Conclusions

Interventions that assist patients to develop strategies to safely inform their partners are needed to increase patient-initiated partner notification.

Introduction

Identifying, notifying and treating sex partners of sexually transmitted infection (STI) patients is a cornerstone of public health approaches aimed to stop the spread of STI, including HIV.^1–3 Notifying and treating partners prevents reinfection of index patients, decreases the disease burden in sexual networks and prevents forward STI transmission. Failure to inform partners of exposure to STI leaves sexual network members infectious. Furthermore, a significant number of infected individuals who are uninformed of their risks remain asymptomatic and untreated.⁴ Increasing the capacity to detect and treat STI in sex partners of index patients can therefore impact the health of entire sexual networks and significantly reduce the rate of new infections.⁵

Studies show that only a fraction of sex partners of STI patients are notified, highlighting that relatively few partners are ultimately tested and treated.⁶ The STI reproductive rate is a function of contact frequency with infected sex partners making efforts to reduce new infections futile without removing source infections.⁷ Fear of negative partner reactions, including ending relationships and violent responses, is among the most common perceived barriers to partner notification.^8–10 In addition, men and women may vary in their perceived partner reactions in response to partner notification. For example, women in Dar es Salaam, Tanzania identified fear of partners' reactions, decision-making and communication patterns between partners and perceived partners' attitudes towards testing as the greatest barriers to informing partners of patient HIV status.¹¹ For men in Uganda, however, beliefs that notifying partners would reveal unfaithfulness and lead to the ending of a relationship were identified as barriers to notifying partners.¹² As is common in resource-limited settings, partner notification in South Africa depends almost entirely on index patients notifying their own sex partners. This study sought to identify and perceive social and relationship barriers to patient-initiated partner notification.

Methods

Participants and setting

Participants were 324 men and 452 women receiving STI diagnostic and treatment services from a community-based clinic in Cape Town, South Africa. The public health clinic that served as the site for the current research treats about 300 STI patients per month.

Measures

Measures were collected in an anonymous self-administered survey format in English and Xhosa, the two languages spoken by nearly all clinic patients. Participants were provided a survey and clipboard and were instructed to read each question and provide their responses. Participants requiring assistance (<5%) were administered the survey in an interview by research staff. Measures included demographic and health characteristics, sexual and substance use behaviors and partner notification experiences, intentions and outcome expectancies. All measures were conceptually derived from previous studies on HIV and STI partner notification barriers^8–11 and were pilot-tested with native speakers for clarity and translation/back translation accuracy.

Demographic and health characteristics

Participants reported their age, gender, race, level of formal education, sources of income and marital status. We asked participants about their HIV testing history and HIV status. Participants also indicated whether they had ever exchanged sex for money or materials, either as the recipient or provider of sex. We also asked whether they were at the clinic for STI services, had been told to come to the clinic that day to be checked for an STI and whether they were currently experiencing genital pain, unexplained genital discharge or genital ulcer.

Sexual behaviors

Participants responded to a series of questions assessing their number of male and female sex partners and frequency of sexual acts in the previous month, specifically vaginal intercourse with and without condoms. We also asked participants the number of sex partners they had in the previous month and whether they currently had main and casual sex partners. Participants were instructed to think back over the past 30 days (1 month) and estimate the number of partners and occasions in which they practiced vaginal intercourse with and without condoms. For condom use, participants were specifically asked the number of times they or their sex partner had used condoms in the past month during vaginal intercourse. A single form was used for men and women, so all items were gender neutral. Participants were also asked whether they had ever used a condom and whether a condom was used during their most recent sexual occasion.

Substance use

Alcohol use was assessed with a single item taken from the Alcohol Use Disorders Identification Test (AUDIT) that is commonly used to index current frequency of drinking. The item asked how often participants drink alcohol beverages, including various beers, wines and spirits, and was responded on a time frequency scale that included ‘never’. For other drug use, participants were asked whether they had used cannabis (dagga), methamphetamine (Tik), inhalants, cocaine or any other recreational drug in the previous month.

Prior partner notification experiences

We asked participants whether they had ever informed a sex partner that they had been diagnosed or treated for an STI. We also asked whether a sex partner had ever informed them of an STI. Specifically, ‘Have you ever told a sex partner that THEY should go to a clinic because you had an STI?’ and ‘Have you ever been told by a sex partner that YOU should go to a clinic because he or she had an STI?’ For both questions, participants responded dichotomously, as to whether or not they had the experience.

Intentions to notify partners

To assess participant intentions to notify sex partners of their current STI, we asked whether they would tell partners or have the clinic inform their partners of their need to be checked. The specific question stated, ‘If you had an STI, which of the following would you do to let your sex partner know that they should visit a clinic to be checked for STIs?’ Note that participants in the current study were all being seen at the clinic for a current STI. The options for informing partners included the participants themselves or the clinic staff performing the notification. The options also included not notifying partners at all with the item, ‘I would not tell my partner and I would not want a nurse or counselor to tell them either.’ Options for informing partners were presented for both main and casual partners. Responses were used to define intentions to inform sex partners of an STI.

Partner notification outcome expectancies

We assessed whether participants perceived the potential for adverse outcomes should they inform sex partners that they were currently being treated for an STI (specific items shown in Results section). Participants indicated whether they thought the partner response would or would not occur as ‘yes’, ‘maybe’ or ‘no’. We collapsed yes and maybe responses to form a single affirmative response category and responses were summed to create an index of adverse outcome expectancies.

Procedures

Individuals aged 18 years and older who were receiving clinic services were approached to complete an anonymous survey while waiting to see a clinic nurse. Surveys used a paper–pencil format, required ∼10 min, and were anonymous. Participants were offered a nominal incentive, a chocolate candy bar, for completing the survey. A total of 1105 clinic patients, 376 men and 729 women, completed surveys, representing ∼80% acceptance rate. In total, 329 (30%) did not indicate being seen for an STI visit or experiencing current STI symptoms. Thus, 776 persons reporting indicators of an STI were included in the analyses. South African and the US institutional ethical review boards approved all procedures.

Data analyses

We examined factors associated with whether participants indicated that they would or would not inform current sex partners of their STI. Participants who indicated that they would not inform their main or casual sex partners themselves or have the clinic inform them were defined as not intending to inform partners. We compared this group to the 372 (48%) participants who did intend to inform their partners on demographic and heath characteristics. We then partitioned the sample by gender for further analyses of intentions to inform partners. For these analyses, we examined sexual and substance use behaviors as well as partner notification experiences and expected outcomes stratified by gender. Finally, we examined factors associated with intentions to inform partners in multivariable models stratified by gender that included relevant participant characteristics and outcome expectancies to test the independent factors associated with intentions to inform partners. All analyses were performed using logistic regressions, reporting odds ratios and 95% confidence intervals.

Results

Results indicated that 217 (67%) men and 275 (61%) women had told a sex partner about a previous STI, χ² (1776) = 3.06, P = .08, and that 178 (55%) men and 217 (48%) women had been told by a partner about a past STI, χ² (1776) = 3.18, P = .07. Table 1 summarizes the characteristics of STI clinic patients who did not intend to notify their partners (n = 404) and those who stated they intended to inform partners (n = 372).

Table 1.

Demographic and health characteristics of STI clinic patients who do not and do intend to notify partners.

Characteristic	Not intending to notify partners n = 404		Intending to notify partners n = 372		OR	95% CI
	n	%	n	%
Men	186	46	138	37
Women	218	48	234	63	1.44**	1.08–1.92
Black African	391	98	352	96	1.20	0.91–1.58
Unmarried/single	367	91	342	92	1.18	0.71–1.97
Education
Not completed schooling	182	45	225	61	REF
Completed schooling	132	33	97	26	2.39**	1.58–3.60
University	87	21	45	12	1.42	0.91–2.21
Age (M, SD)	28.7	8.1	28.0	7.9	0.98	0.97–1.00
Sources of income
Family	109	27	143	38	0.60**	0.44–0.81
Employed	151	38	112	30	1.40*	1.04–1.90
Self-business	30	8	20	5	1.42	0.79–2.56
Sex work	1	1	4	1	0.23	0.02–2.07
Pension/grant	43	11	51	14	0.75	0.49–1.16
Illegal activities	5	1	5	1	0.92	0.26–3.23
No sources of income	64	16	48	13	1.28	0.85–1.92
HIV testing history
Has been tested for HIV	384	95	337	91	1.88*	1.06–3.33
Been tested with a partner	208	52	162	44	1.33*	1.01–1.77
HIV status
Positive	70	17	82	22	REF
Negative	294	73	241	66	1.01	0.59–1.73
Unknown	39	10	45	12	0.71	0.44–1.12

*P < 0.05, **P < 0.01.

Sexual behaviors and substance use

Among men, 71% indicated having a main sex partner and 44% had casual sex partners. More than half of men reported two or more sex partners in the previous month. Condom use during most recent sex was reported by less than half of men, and condoms were used during <50% of intercourse occasions in the previous month. In terms of substance use, 77% of men reported current alcohol use and 50% used other drugs in the past month. None of men's sexual or substance use behaviors were associated with intentions to inform partners (see Table 2).

Table 2.

Sexual and substance use behaviors among STI clinic patients who do not and do intend to notify partners.

Behavior	Men				OR	95% CI	Women				OR	95% CI
	Not intending to notify partners n = 186		Intending to notify partners n = 138				Not intending to notify partners n = 218		Intending to notify partners n = 234
	n	%	n	%			n	%	n	%
Has a main sex partner	140	75	91	67	1.53	0.94–2.51	162	74	162	70	1.26	0.83–1.91
Has casual sex partners	77	41	64	47	0.80	0.51–1.25	39	18	56	24	0.68	0.43–1.08
Number of sex partners
1	84	45	67	49	REF		165	76	174	74	REF
2	50	27	39	28	0.77	0.44–1.33	34	16	36	15	0.79	0.41–1.51
3+	52	28	32	23	0.78	0.43–1.44	18	8	24	10	0.79	0.36–1.71
Has used a condom	163	89	115	83	1.55	0.82–2.93	184	85	195	84	1.09	0.65–1.84
Used a condom at last sex	77	42	68	50	0.72	0.46–1.12	79	36	125	54	0.49**	0.33–0.71
Has received materials for sex	12	7	16	11	0.52	0.24–1.15	5	2	21	9	0.23**	0.08–0.64
Has given materials for sex	25	13	23	16	0.77	0.42–1.43	4	2	17	7	0.23**	0.07–0.72
Alcohol use
None	36	20	35	25	REF		81	37	94	40	REF
Weekly	96	52	70	51	1.58	0.83–2.99	114	52	91	39	0.53*	0.29–0.95
Monthly	52	28	32	23	1.18	0.69–2.02	22	10	48	20	0.36**	0.20–0.65
Drug use	92	50	69	50	1.01	0.65–1.57	47	22	47	20	1.08	0.68–1.70
Vaginal sex without condoms	3.44	8.08	2.86	4.40	0.98	0.94–1.02	3.30	5.65	2.68	4.93	1.02	0.98–1.06
Vaginal sex with condoms	3.57	4.92	4.10	6.25	1.01	0.97–1.05	3.73	8.55	3.34	7.63	1.01	0.98–1.03
% intercourse with condoms	50.13	35.63	45.32	34.48	0.67	0.35–1.29	51.92	36.66	47.21	35.01	1.44	0.81–2.57

*P < 0.05, **P < 0.01.

For women, 72% had a main sex partner, 21% had a casual sex partners and 25% of women reported two or more sex partners in the previous month. Frequencies of sexual behaviors reported by women were similar to those reported by men, and women also indicated using condoms during about half of intercourse occasions. Women who intended to inform their partners were significantly less likely to have used a condom at last sexual occasion and were less likely to have exchanged sex for money or other materials. In addition, women who drank weekly were significantly less likely to intend to inform partners, whereas women who drank less frequently (monthly) were more likely to inform partners. Other drug use was less frequent than alcohol use among women, 21% reported drug use in the past month and drug use was not associated with intentions to inform partners.

Partner notification experiences and outcome expectancies

The majority of participants had previously informed a sex partner of a past STI and nearly half had themselves been informed of an STI by a partner. Results did not show any associations between past experiences with partner notification and intention to notify partners among men or women. For both men and women, adverse outcome expectancies were associated with intentions to not inform partners of their STI exposure (see Table 3). Analyses showed that men who did not intend to inform their partners of an STI were significantly more likely to expect their partner to leave them, end their relationship and act out violently toward them should they inform them of the STI.

Table 3.

Partner notification experiences and outcome expectancies among STI clinic patients who do not and do intend to notify partners.

Experience	Men				OR	95% CI	Women				OR	95% CI
	Not intending to notify partners n = 186		Intending to notify partners n = 138				Not intending to notify partners n = 218		Intending to notify partners n = 234
	n	%	n	%			n	%	n	%
Has informed partner of an STI	87	63	129	70	0.74	0.46–1.18	144	62	131	61	1.02	0.70–1.50
Has been informed by a partner of an STI	74	54	104	56	0.90	0.57–1.40	114	48	103	47	1.04	0.72–1.51
Expected outcomes
Would be angry	73	54	90	49	1.24	0.79–1.93	99	42	80	36	1.29	0.88–1.89
Would tell you to leave and never see you again	58	42	46	25	2.20**	1.37–3.54	74	31	47	22	1.69*	1.10–2.59
Would end relationship	63	46	53	28	2.19**	1.38–3.49	78	33	65	30	1.16	0.78–1.74
Would act out violently	63	46	42	22	2.91**	1.80–4.72	65	27	38	17	1.82**	1.15–2.86
Number of adverse outcomes
0	36	26	75	40	REF		93	40	102	47	REF
1	27	20	49	26	2.77**	1.56–4.93	52	22	46	21	1.91*	1.14–3.22
2	23	16	23	12	2.42**	1.29–4.52	33	14	38	17	1.54	0.86–2.78
3+	52	37	39	21	1.33	0.65–2.71	56	24	32	15	2.01*	1.06–3.81

*P < 0.05, **P < 0.01.

Among women, 45% expected at least one negative outcome if they were to inform their partners. Women who were not intending to inform their partners were significantly more likely to expect adverse outcomes, including expecting their partner to leave them and act out violently toward them.

Multivariable models

Among men, completing formal education was significantly associated with intentions to inform partners. Men not intending to inform partners were also more likely to have casual sex partners and expect informed partners to become violent. For women, completing education was also related to informing partners and expecting their partner to leave and not see them again was associated with intention to not inform partners (see Table 4).

Table 4.

Multivariable models predicting intentions to notify sex partners among men and women STI clinic patients.

Factor	Men		Women
	Adj OR	95% CI	Adj OR	95% CI
Education	1.75*	1.25–2.45	1.41**	1.07–1.87
Has been tested for HIV	0.45	0.17–1.16	1.75	0.64–4.82
Been tested with a partner	1.05	0.62–1.78	0.88	0.59–1.33
Alcohol use	1.34	0.92–1.96	0.87	0.64–1.68
Number of sex partners	1.20	0.85–1.70	0.99	0.71–1.38
Has a main sex partner	1.19	0.68–2.10	1.21	0.76–1.91
Has casual sex partners	0.52*	0.29–0.92	0.72	0.43–1.21
Has informed past partner of an STI	0.75	0.42–1.34	1.11	0.70–1.76
Has been informed by a past partner of an STI	1.05	0.60–1.83	0.91	0.58–1.43
Would be angry	0.92	0.52–1.62	1.03	0.65–1.63
Would tell you to leave and never see you again	1.32	0.65–2.68	1.77*	1.00–3.14
Would end relationship	1.64	0.86–3.13	0.70	0.41–1.19
Would act out violently	1.98**	1.05–3.72	1.37	0.78–2.38

*P < 0.05, **P < 0.01.

Discussion

Main findings

Our sample of South African STI clinic patients demonstrated high rates of sexual risk behaviors with multiple partners in a 1-month timeframe. More than half of men and one in four women receiving STI clinic services reported more than one sex partner in the previous month, with more than 70% of men and women reporting that they have a main sex partner. For both men and women, condoms were used less than half of sexual occasions. Substance use was also common, with 75 and 50% of men and 62 and 20% of women reporting alcohol and other drug use, respectively.

For men, we did not observe any significant associations between intentions to inform partners and sexual and substance use behaviors. In contrast to men, women who intended to notify partners were significantly more likely to use condoms in their most recent sexual encounter and were more likely to have traded sex for money or materials. In addition, women intending to inform partners were less likely to use alcohol on at least a weekly basis. Having informed partners as well as having previously been informed of an STI by partners were not related to current intentions to notify partners.

For both men and women, concerns about the potential adverse partner reactions were related to intentions to inform partners. A majority of men and women indicated at least one concern for potential adverse reactions. It was expected that partners would be angry, but this factor was not associated with notification intentions. Rather, men and women who did not intend to notify partners endorsed significantly more concerns that partners would leave, end their relationship and act out violently. In multivariable models, men's concern about violent reactions and women's concern about their partner leaving were independently associated with intentions to notify partners. That men expressed concerns about their partners reacting violently was unexpected, and we are not aware of men reporting these concerns in previous research. One possible reason for this unique finding is that our survey asked the same questions of men and women, whereas past research may simply have not asked men about concerns over experiencing violence based on preconceived assumptions about gender. While women also expressed concerns about violent reactions, this factor was not significant in the multivariable model. We speculate that men's concern about violence ensuing from notifying their partner is significant because of the social and legal constraints placed on men in situations of domestic violence. Women reacting violently in relationships may therefore put men at risk for arrest should they respond with further violence. Thus, for men a partner's violent reaction to STI notification is contextually different than women and should not be disregarded in developing strategies for partner notification.

What is already known

Partner services are essential to preventing the further spread of STI and yet relatively few partners of STI clinic patients are ultimately tested and treated.⁶ Past studies demonstrate that patients fear their partners' adverse reactions to being informed of potential STI exposure even when their partner is the source of their infection.⁹ Most of what is known about barriers to partner notification is from studies of HIV infection and the degree to which these same factors contribute to curable and non-life threatening STI has been unclear. In addition, gender differences in fears of partner reaction to notification are often inferred without examining the same potential barriers facing men.¹¹

What this study adds

The results of this study confirm that concerns and fears about partner reactions, particularly anger and violent responses, create significant barriers to partner notification. A contribution of this study is that these concerns and fears were observed in men not just in women. In contrast to previous research, men and women's fears of adverse partner reactions differ in type but not along lines that are stereotypically inferred. Differences between our study and past research, such as the social contexts that vary by country as well as examining patients diagnosed with a broad array of STI and not just HIV may account for some of the observed differences. In addition, we examined intentions to notify partners among STI patients rather than naming partners for public health service intervention. All of these factors likely play a role in the study findings.

Given the concerns endorsed by patients regarding potential partner reactions to notification, interventions are needed to assist patients who are requested to initiate partner notification. A large body of literature supports the use of brief communication skills building to reduce sexual risks and these interventions include components geared toward talking with sex partners about risk reduction strategies.¹³ Behavioral skills interventions offer the opportunity to problem solve potential adverse partner reactions, develop strategies for managing threatening situations and build confidence in being able to inform partners. Interventions designed to increase partner notification should avoid making gender-based assumptions of concerns about adverse partner reactions given the unexpected findings in this study. Our findings clearly indicate the importance of asking direct questions regarding barriers to notifying partners and offering solutions to overcome those barriers regardless of patient gender. In addition, the technologies to assist patients in informing their partners, including the use of anonymous e-cards and text messaging, are now commonly available.^14,15 With increased access to mobile communications and other communication technologies, methods of partner notification using text message services and the Internet will become viable options for patients in resource-limited countries. Given that most settings will continue to rely on patients to initiate partner notification, public health interventions are needed to remove barriers to partner notification.

Limitations

The sample was one of convenience and cannot be considered representative of STI clinic patients in South Africa. In addition, our measures were self-administered and may have resulted in under-reporting of sensitive behaviors, particularly sexual behaviors, substance use and intentions to notify sex partners. Although participants were assured of anonymity, collecting data in the context of the clinic may have suppressed reporting. Thus, rates observed for intentions to notify partners may underrepresent actual behaviors. Another factor that may have limited our study is the potential impact of reading literacy on survey responses. While few participants required assistance to complete the survey, the degree to which misinterpretation of items and poor item comprehension interfered with survey responses is unknown.

Funding

This project was supported by National Institute of Child Health and Development grant R01-HD74560.