Background

The prevention of new human immunodeficiency virus (HIV) infections and treatment of infected persons emphasizes the importance of HIV status disclosure among HIV-infected clients, particularly to their sexual partners [1]. In sub-Saharan Africa (SSA), more than 90% of adults acquire HIV infection through unprotected sexual intercourse with infected partners in discordant relationships [2]. Partners in HIV sero-discordant relationships are at most risk of getting infected [3]. Therefore, HIV status disclosure among partners has overarching implications for safer sex practices, treatment adherence and consequently, has the potential to limit new infections [4]. In SSA, almost 70% of males were HIV negative at the time their partners were tested positive. During their marriage, about one-third of the negative partners become infected [5]. In Ghana, more women know their HIV status than men due to HIV testing and counselling (HTC) during antenatal care with the aim of prevention-mother-to-child-transmission [6]. To prevent the continuous spread of the infection, by reducing the incidence of the infection in a locality, status disclosure is essential for safe sex practicing. Some antiretroviral therapy (ART) clients disclose their status for various reasons. For instance, some have indicated that they disclose because they want to help others adopt preventive behaviors to remain uninfected, whereas others disclose because they want support from their partners [7]. However, disclosure has been reported to be associated with some risks. Others did not disclose their status for fear of the partner’s withdrawal/divorce, physical violence, fear of abandonment, blame, disruption of relationships, physical and emotional abuse [8]. Although disclosure has some negative consequences, it may motivate partners to seek testing, cause behavioral change, provide support for treatment, and ultimately decrease transmission of HIV [7].

The Upper East Region is one of the sixteen administrative regions in Ghana. Over the last 10 years (2010–2019), the average prevalence of HIV in the region was 1.7%. The prevalence ranged from 1.3 to 2.4%; the lowest prevalence (1.3%) was recorded in 2017. From 2017 to 2019, there was an increase in the prevalence to 2.1%, a marginally higher prevalence than the national average of 2.0%. The nature of the prevalence in the region over the 10-year period was undulating [9]. In the region, polygamy is common [10]. This practice has been reported to increase the risk for HIV infections [11, 12]. Considering the fact that the HIV prevalence in the region was marginally higher than the national average, together with the fact that polygyny is common in the region, it is imperative to understand some of the dynamics of the infection in the region while Gazimbi et al. [13], in their review, did not associate polygyny with HIV transmission. Therefore, this study was designed to identify factors associated with HIV disclosure in the Upper East Region of Ghana [13].

Methods

Study Design and Study Setting

Case–control study approach was used for this study. This approach was deemed appropriate because the study sought to determine the demographic characteristics, socio-economic factors, and ART service-related factors associated with cases (respondents that had not disclosed their HIV status to their partners) using individuals that disclosed their HIV status to their partners as controls. Equal numbers of cases and controls were obtained from the respective ART sites. The cases and controls were matched on sex and age in years (± 5) to control the confounding effects of those two variables. It shares borders with the Republic of Burkina Faso to the north, the Republic of Togo to the east, the Northern Region to the south and the Upper West Region to the west. The region has sixteen (16) administrative Districts/Municipalities [14].

Study Population

The study population consisted of all people living with HIV (PLHIV) aged 18 years or above, with sexual partners and are receiving routine ART services in selected ART sites in the Upper East Region.

Inclusion and Exclusion Criteria

For both cases and controls, respondents 18 years or more were included in this study, with prior consent. Respondents that were severely sick to provide accurate responses as well as respondents that could not provide historical events were excluded from the study.

Sample Size Determination

The sample size was calculated using a desired power of 80%, precision at 95% confidence level, and a ratio of cases to control as 1:1. The ratio of 1:1 for cases to controls is recommended for a matched case control study with more exposure variables, a single non-dichotomous or continuous variable [15]. Using the difference in proportion formula [16], a minimum sample size of 196 consisting of 98 cases and 98 controls was required for the study. However, a sample size of 200 consisting of 100 cases and 100 controls consented and participated in the study. The sample size for each ART site was determined proportionately by calculating the number of ART clients at that site out of the total number of ART clients from all selected sites in a zone by the overall sample size for the study.

Selection of ART Sites and Sampling of Respondents

A multistage sampling was used for the study (Fig. 1). The Upper East Region was stratified into Western (4 districts with 9 ART sites), Central (5 districts with 8 ART sites), and Eastern (6 districts with 9 ART sites) zones. In each stratum, the Districts/Municipals (subsequently referred to as districts) were ranked based on the number of active ART clients from the highest to lowest and the first or top three selected. In the districts/municipalities, the health facilities with ART sites were ranked based on the number of active ART clients from the highest to lowest and the first ART site selected. Clients who met the inclusion criteria for controls, were then selected based on the matching variables sex and age. Whenever a selected client dissented, the process was repeated to replace until the required sample size was obtained for that ART site. The ART client was classified as a “Case” if he/she was sexually active and did not disclose his/her status to the sexual partner. The number of active ART clients in a selected ART site was used to proportionately determine the sample size for that ART site in each stratum [17]. In each selected ART site, systematic sampling was used to select cases [18]. Any consented ART client selected was surveyed. Within each ART site, respondents were systematically selected by selecting participants that were given even numbers in the order they visited the clinic.

Fig. 1
figure 1

The flow chart for selection of respondents. Cases were respondents that had not disclosed HIV status, control were respondents that had disclosed HIV status

Full size image

Data Collection Procedure and Measurement of ART Adherence

Trained health care workers, namely, general nurses, public health, and community health nurses, administered the questionnaires. The instrument for data collection was adapted from similar studies in some parts of Ghana [19, 20]. The study respondents (PLHIV) were sampled at each ART site where they were receiving ART services. Data were collected on ART clinic days across the study sites for a period of 10 weeks. The data were collected using a pretested structured questionnaire and administered in a local language the respondents understood. Adherence to ART was assessed by timely attendance at appointments for delivery of antiretroviral drugs and counting the remaining tablets.

Study Variables

The structured questionnaire used for this study had three sections, namely, socio-demographic characteristics of respondents, knowledge of sexual partner’s HIV status, and other factors associated with HIV status disclosure.

Socio-demographic factors surveyed were gender of respondents, religious affiliation, marital status, and highest educational level of respondents.

Under knowledge of the sexual partner’s HIV status, the following questions were used to elicit responses: how old (in years) were you when you were diagnosed with HIV? Why did you get tested for HIV? Do you know your sexual partner’s HIV status? What is your sexual partner’s HIV status? How did you get to know your sexual partner’s HIV status?

Under other factors associated with HIV status disclosure, responses were elicited using the following questions: Have you received any counselling with respect to HIV status disclosure? Were you ever abused by your partner before HIV diagnosis? Were you aware of your partner’s involvement in other sexual relationships prior to diagnosis? How many other sexual partners did your partner have prior to diagnosis? What is the highest educational level of your partner? What is the occupation of your partner? Which of the following best describes the average monthly income of your partner? How long did it take you to start taking ART after the diagnosis? How long have you been into ART? Why did you choose him/her as your treatment monitor?

Data Management and Analysis

The questionnaire was designed with validation rules in EpiData manager version 4.0.2.101 and the data collected was entered using EpiData entry client version 4.0.2.49. The data was first exported to STATA v 14.1 for entry accuracy and completeness checks and then exported to MS Excel 2016 for further cleaning. Study variables were assessed using Stata to determine the Cronbach’s Alpha (CA) value (range 0–1.0) of the instrument. A CA value in the range of 0.60–10 was considered desirable. The cleaned data was then imported into STATA v 14.1 for analysis. Descriptive analysis was done for some variables and presented as frequencies and percentages using cross tabulations and presented in tables. Data on the independent variables were exported to excel for concordant and discordant pair analysis to draw the triplet’s pair two-by-two and calculate the respective odds ratios [Mantel Haenszel matched odds ratio (ORMH)] for each variable studied. The cases and controls were discordant or not with respect to the exposure variable. The ORMH is the ratio of the sum of the discordant pairs in which the case is exposed over the sum of the discordant pairs in which the case is not exposed. Concordant pairs (in which the case and the control are either both exposed or both unexposed) do not contribute to the numerator nor to the denominator of the ORMH. Bivariate analysis using conditional logistic regression was done to determine the crude association between the independent variables and HIV status disclosure to a partner. Variables with p value of less than 0.05 were included in the multiple regression analysis to adjust for confounding.

Ethical Consideration

This study was reviewed and approved by the University of Health and Allied Sciences Research Ethical Committee (UHAS-REC) with the reference number UHAS-REC A. l [40] 19–20. A written informed consent was obtained from the selected individual respondents before the questionnaire was administrated.

Results

Distribution of the Study Participants

Samples were collected from nine out of the fifteen districts and municipalities (subsequently referred to as districts) in the region. The study districts were Bawku municipal, Bawku West district, Bolgatanga municipal, Bingo district, Builsa North district, Garu district, Kassena Nankana municipal, Kassena Nankana West district and Nabdam district. From each district, an equal number of cases and controls were collected. Majority of the cases and controls were contributed by the Upper East regional hospital located in Bolgatanga municipal (33%) while Paga hospital in the Kassena Nankana West District contributed the least cases and controls (4%) (Table 1).

Table 1 Distribution of the ART clients that participated in the study
Full size table

Demographic Characteristics of Study Respondents

The mean age of the cases was 35.4 ± 7.5 years (range: 19–50 years) while that of the controls was 36.1 ± 8.1 years (range: 21–55 years). The age distribution between cases and controls was comparable (t = 0.65, p = 0.5112). Similarly, cases and controls were comparable in terms of gender, type of settlements, and family type. Cases and controls only differed in terms of educational level (Table 2).

Table 2 Background characteristics of respondents
Full size table

Mantel–Haenszel Matched Odds Ratio (ORMH) Test of the Association Between Demographic Characteristics and HIV Status, Disclosure to Partners

None of the demographic variables associated with HIV status disclosure except educational status where ART clients which no formal education were 2.4 times more likely to disclose their HIV status to their partners compared to those who had formal education [ORMH = 2.4; 95% CI (1.15–5.02) p = 0.020] (Table 3).

Table 3 Association between demographic characteristics and HIV status disclosure to partner
Full size table

Mantel–Haenszel Matched Odds Ratio (ORMH) Test of the Association Between Socio-Economic Factors and HIV Status Disclosure to Partners

ART clients who were not staying with their partners were 70% less likely to disclose their HIV status to their partners compared to those who were staying together with their partners [ORMH = 0.3; 95% CI (0.14–0.61) p < 0.001], Similarly, ART clients who had no children with their partners were also less likely to disclose their status to their partners compared to those who had children [ORMH = 0.3; 95% CI (0.11–0.88) p = 0.022]. The other variables did not associate with partner HIV disclosure (Table 4).

Table 4 Association between socio-economic factors and HIV status disclosure to partner
Full size table

Association Between ART Service-Related Factors and HIV Status Disclosure to Partners

ART clients who were on ART for more than 3 years were 2.2 times more likely to disclose their HIV status to their partners compared to those who were on ART for less than 3 years or 3 years [ORMH = 2.2; 95% CI (1.07–4.53) p = 0.028]. The ART clients who were not adhering to ART were less likely to disclose their HIV status to their partners compared to those who adhered to ART [ORMH = 0.1; 95% CI (0.01–0.53) p < 0.001]. Moreover, respondents who did not have treatment monitors were less likely to disclose their HIV status to their partners compared to those who had treatment monitors [ORMH = 0.1; 95% CI (0.02–0.19) p < 0.001]. Additionally, ART clients who started ART in less than 1 month after diagnosis were 3.7 times more likely to disclose their HIV status to their partners as compared to those who started ART in 1 month or more [ORMH = 3.7; 95% CI (1.44–11.05) p < 0.001] (Table 5).

Table 5 Association between ART service-related factors and HIV status disclosure to partner
Full size table

Discussion

HIV sero-positivity disclosure has been previously shown in several studies to result in better adherence to ART treatment, good treatment outcomes, and reduction in the risk of HIV transmission among couples [21, 22]. However, previous studies have reported between 7 and 79% of partner disclosure rates [23,24,25]. HIV status disclosure to partners is essential to reduce the spread of the virus to an uninfected partner, therefore, identifying factors that will encourage status disclosure will inform post testing counselors the areas to emphasize during posttest counselling. Although factors affecting HIV status disclosure have been published from other parts of Ghana [26,27,28], no such study has been done in multicenter in the Upper East Region of Ghana, a region with high prevalence of polygamy [10].

In this study, ART clients who adhered to treatment were more likely to disclose their HIV status to their partners compared to those who did not. The higher adherence among the disclosed group could be because of a strong believe that there will be acceptance and support by the partner. Furthermore, to be able to adhere to ART services, the infected partner may have accepted his or her status and therefore could manage issues that may come out after disclosure. This support in turn has had an appositive impact by ensuring the partner adheres to the ART. This is consistent with findings from other studies that found significant associations between HIV status disclosure to partners and ART adherence [29,30,31]. However, Mabunda et al. [32] did not find an association between HIV status disclosure to partners and ART adherence. This finding suggests that reduced adherence to ART could cause nondisclosure of status as previously reported by Tsega and others [33]. The reverse has been previously reported where lack of disclosure has been associated with reduced ART adherence. This is because lack of disclosure is often associated with reduced social support, increased anxiety and depression, reduced emotional support as well as reduced financial support. These limit adherence to ART services participation [34, 35].

The educational status of the ART client was significantly associated with HIV status disclosure. In this study, those who had no formal education were more likely to disclose their status to their partners than those who had some level of formal education. A previous study gave the reason why HIV infected with formal education were less likely to disclose their HIV status to anyone. Notably, among the reasons were fear of loss of social standing and fear of stigmatization in the face of their high social status [36].

In another study in the Volta Region of Ghana, a similar finding was reported [19]. In contrast, Gultie et al. (2015) posited that educated persons were more likely to disclose their status than those who had no formal education [37]. Possible reasons for this finding were that, unlike the educated people living with HIV, individuals who had no formal education may be less knowledgeable on HIV/AIDS, they had much trust in health workers and paid attention to them. However, Longinetti et al. (2014), found no significant association between educational level and HIV status disclosure in a study in South Africa [38]. Findings from this study and that of Amavih, 2017 [19] is rather worrying because, globally and in developing countries, there are concerted effort to improve literacy and numeracy rates through formal education. This could be due to the reason that health care givers may assume that the educated are more likely to understand and hence, may not give adequate education. This area needs further research to generate data to contribute to policy and practice. It is therefore important that health care workers are encouraged to provide adequate information, education, and communication literature to everyone, especially those with some level of formal education and offer comprehensive post-testing counselling on the importance of status disclosure to these cohorts of individuals. This will promote partner status disclosure in individuals with formal education living with HIV.

It was interesting to observe that starting ART in less than 1 month after diagnosis and being on ART for more than 3 years associated with HIV disclosure. This could be explained that newly diagnosed persons may still be in the state of denial, hence, likely to disclose to the partner, especially as the partner is polygynous. On the other hand, persons that have been on ART for over three years might have had experience with their colleagues with the advantages of disclosure, hence could easily disclose their status as has been previously reported [39]. Furthermore, continuous counselling to the person on ART may reduce self-stigma, hence encourage disclosure [40].

This study found a significant association between having a treatment monitor and HIV status disclosure to a partner. In Ghana, a treatment monitor is a person an ART client trusts and has full confidence in, that the client voluntarily present to his or her health care providers to monitor compliance to treatment. The person also assists in the detection of adverse reactions to the drugs. In this study, ART clients who had treatment monitors were more likely to disclose their HIV status to their partners relative to those who did not have treatment monitors. This finding is similar to what was found in the Central region of Ghana by Boampong-Konam [20]. Contrary to the findings of this study, Amavih (2017), reported that having a treatment monitor was not associated with HIV status disclosure to a sexual partner. With this finding, possibly ART clients who did not have treatment monitors were not in marital or stable relationships, anticipated negative reactions from their partners, or were not getting enough emotional support from their partners to foster disclosure. It must be emphasized that adequate privacy was not ensured during post-test and in-treatment counselling in the study sites. This was because the counselling session was interrupted on a few occasions by other clients and health workers. Notwithstanding that observation, the counselling room was very conducive for effective counselling.

Conclusion

This study employed an age-gender case–control matched study approach to determine factors that affected HIV status disclosure among sexual partners. In summary, the following factors are associated with increased odds of HIV status and partner disclosure; without formal education, being on ART for more than 3 years, having treatment monitors; living with a sexual partners; having children with sexual partners; adherence to ART and starting ART in less than 1 month after diagnosis. It is recommended to promote HIV status disclosure through adequate post testing counselling by trained counselling and testing personnel. Furthermore, it is important that treatment is initiated as soon as the test result are ready and an assurance is given to clients on the benefit of the treatment. During treatment, counselling should be intensified for clients to come to terms with the need to disclose their status to their partners. Finally, it is imperative to ensure adequate privacy during posttest counseling and in-treatment counselling.

Limitations of the Study

The findings reported in the study cannot be generalized because some of the frequencies of the variables tested were few. Additionally, this study was done in a setting with high HIV prevalence, therefore, the findings may not be applicable to other regions with different HIV prevalence and socio-cultural characteristics. Finally, inadequate privacy during counselling may affect the results reported in this study. This is because the presence of other health care providers in the counselling room may not provide adequate security and trust when sensitivity issues are being discussed.