Study overview

This study was conducted within the APS scale-up study, a hybrid type II implementation science study [26] in Homa Bay and Kisumu counties, between May 2018 to March 2021. The study team consisted of a collaboration between the University of Washington, PATH (a non-governmental organization), and the Ministry of Health in Kenya. The study aimed to determine the effectiveness of APS when integrated into existing routine HIV testing, prevention, and care programs, and to evaluate the implementation of APS in these settings, including the integration, implementation fidelity, acceptability, demand, and costs of the intervention. This analysis investigated the acceptability of APS from the perspective of FICs and MPs.

Study sites, participants

From January to December 2019, we conducted in-depth interviews (IDIs) in 8 facilities participating in the APS scale-up project. Facilities were chosen to maximally represent a range of facility types, including both APS high-performing and low-performing facilities (performance defined as the number of MPs enrolled per FIC) and geographic location/size including a range of facility levels from small rural outposts to high-volume urban clinics.

From each facility, we purposively selected one FIC who named ≤2 MPs, and one FIC who named >2 MPs, for a total of 16 FIC IDIs. Eligibility for FICs in this analysis included being enrolled in the APS scale-up project. Two MPs were recruited from each facility (three from one facility), for a total of 17 MP IDIs. Inclusion criteria for FICs included testing positive for HIV, having been enrolled in APS for six weeks, and not having reported intimate partner violence after receiving APS. Inclusion criteria for MPs included testing positive for HIV and eliciting at least one female partner. There were no exclusion criteria.

Study procedures and data collection

FICs and MPs at these facilities who enrolled in APS were invited to participate in phone IDIs with a qualified qualitative interviewer (MO). Those interested were screened for eligibility and provided oral and written informed consent. The IDIs were conducted in Kiswahili, Luo, or English, based on the language preference of the client. The recorded interviews were transcribed verbatim and translated into English for analysis by our study team. The interviews took place at a study facility or another safe and private location chosen by the participant. The interviewer used a semi-structured interview guide, which addressed APS satisfaction, perceived benefits of the intervention, and challenges that may affect APS delivery or uptake. The interview guides for FICs and MPs were similar; the male guide included perspectives on how males felt being contacted by a routine HTS provider and being informed that they were exposed to HIV. Interviews were audio-recorded and transcribed verbatim.

Data analysis

An inductive thematic analysis [27] was conducted to identify emerging themes. After reading the transcripts multiple times, two researchers (BN and MO) independently openly-coded the data line by line, and then reached a consensus on the codebook through discussion with the study team. After the codebook was tested with two transcripts, the two researchers coded all the transcripts independently and discussed findings with each other and the study team. Through group discussion and iteration, similar codes were aggregated to form categories based on similarity/overlap, and BN and MO evaluated the categories to ensure the codes were appropriately aggregated with additional reviews of the interview texts to ensure codes and categories did not deviate from the participants’ meaning. The categories were then further classified as themes and subthemes based on group discussion and alignment. Themes that emerged from the data were identified and discussed. The entire interview for each participant was the unit of analysis.

We then applied the theoretical framework of acceptability (TFA) proposed by Sekhon et al. [25] to guide our presentation of findings. The TFA was developed as a multi-construct theoretical framework of acceptability of healthcare interventions that can be applied to assess acceptability from the perspective of both intervention deliverers and recipients [25]. This framework has been utilized in a number of complex health intervention studies [28–30], and leveraged regionally in SSA [31–33]. Final themes were mapped onto six of the dimensions of the TFA: affective attitude, burden, effectiveness, self-efficacy, ethicality, and intervention coherence. In the analysis, it emerged that participants did not distinguish opportunity costs from burdens or perceived effectiveness; therefore, we did not retain the construct in our analysis. Data were analyzed in ATLAS.ti version 8.4.4, and the findings of this study are reported according to the consolidated criteria for reporting qualitative studies (COREQ) [34].

Ethical consideration

The study was approved by the Kenyatta National Hospital Ethics and Research Committee (P465/052017) and the University of Washington Institutional Review Board (STUDY00002420). All participants provided written informed consent. Girls who are ≥15 years old are considered emancipated minors in Kenya if they meet one of the following criteria: married, pregnant or have had a sexually transmitted infection, including HIV. It was under this condition that those ≥15 years were enrolled. Each recorded interview and corresponding transcript was assigned an ID number and stored in an encrypted place with no names or identifying information included. The recordings will be destroyed at the latest 5 years after the study completed.

Affective attitude

Affective attitude is defined as how an individual feels about an intervention [25]. Participants expressed a wide range of feelings regarding the APS intervention, from comfort with to wariness of APS. For participants conveying comfort with the intervention, the supportive roles of the health care workers (HCWs) in both counseling and notifying sexual partners were highlighted as important facilitators.

Interviewer: Did you have fears or you were just comfortable to give their contacts? Participant: I did not have any fear and that is why I was giving out their contact. Interviewer: You were just comfortable, why were you comfortable? Participant: Because I had been counselled and taught and so I also wanted to help others as well. (IDI 036, FIC)

Because of the approach they [HCWs] gave me, they were open with me and I was also open with them and didn’t feel uncomfortable about anything. (IDI 035, MP)

Notable discomfort and wariness of naming and notifying partners was expressed as well. Some participants explained that “there are some private matters that feel awkward when someone asks you about and is not easy to speak up,” and reflected they were thinking to themselves “how do I start explaining such a thing to someone else.” The role that the healthcare workers played was seen as mitigating these fears and wariness, but this challenge was reflected consistently.

Although when she [HCW] first mentioned it to me, I was worried how I was going to approach my sexual partners but then she told me I should leave that to her because that is now her job. (KII 038, MP)

Burden

Burden refers to the perceived amount of effort that is required to participate in an intervention [25]. The sub-themes identified included the technical ease of engaging with the intervention, but also the risk to the relationship of disclosing one’s status, and the risk of a woman being exposed to interpersonal violence.

The low burden of engaging with the intervention was noted routinely as “easy” and “without any fear,” particularly with the benefit of “helping me and my partners,” and one’s children. However, the consistent detracting burden noted for APS acceptability was the risk and fear of repercussions of having one’s status disclosed. Participants explained that partners “might leave them” if partners find out about their HIV status during the APS process. MPs relayed concern about the risk of disclosing multiple sexual partners, particularly in exposing those non-primary partners with whom they are in “secret” relationships.

The difficulty was the part where I was expected to expose my sexual partners. This is someone we have been having sex with in secret and now I was about to expose her and tell this provider who they are and the things we have done together. I was scared that she could find out that I was the one who shared their details with the provider. However, because of the kind of teachings I got from the provider, I opened up because I wanted all of us to get help. (IDI 055, MP)

FICs expressed similar sentiments, but often with the elevated risk of their MPs “leaving them,” dissolution of their relationships, and often the corresponding loss of access to financial resources and housing. These burdens of APS escalated to the risks of intimate partner violence (IPV) against the FICs if the sexual partners could somehow find out about having been named to healthcare staff.

Sometimes they think they can be tracked and the husband will find out that they are the ones who gave out the contact. And if they find out then sometimes they can be beaten thoroughly. You can even be sent away from the home. (IDI 061, FIC)

MPs also highlighted IPV as a concern that their female partners, or any FIC, would be considering in the acceptability of APS, in that any resulting disclosure of HIV status would be dangerous.

The reason why the women will fear is that–you know some of us men are very arrogant and some of the men when they will hear about this then they will not take it lightly. They will be very harsh and they can even become violent and the woman can be beaten (IDI 064, MP)

Ethicality

According to the TFA, ethicality is the extent to which the intervention is a good fit with an individual’s value system [25]. Interestingly, though the interview questions did not explicitly ask about value systems or ethics of this intervention, two sub-themes were repeated in the IDIs: value for the partner, and the opportunity for this intervention to save lives. Participants expressed that only those who “didn’t value their relationships” or “didn’t love each other” would not accept APS, and that if they truly “loved someone you wouldn’t want to harm them” by not naming partners to be notified.

Participants suggested that their perspectives on the acceptability of APS were influenced by strong motivations to “do good” and “save a life,” and this was noted across male and female interviews.

I feel good because I have saved their lives somehow by helping them get tested and enrolled into care. I have saved their lives and now each one of us can now take good care of their lives and enroll into care. (IDI 048, FIC)

This facilitating factor i.e., “to help their partners and family,” was consistently identified as playing a motivating role for the intervention’s acceptability.

I didn’t find it bad because I wanted to help my life and that of the other partners as well. I wanted them to be reached and talked to so that if they also find it good they can also enroll into care. (IDI 067, FIC)

Intervention coherence

The next theme in Sekhon et al.’s framework is intervention coherence, which denotes the extent to which the participant understands the intervention and how it works [25]. In our study, we found that aspects regarding the acceptability of APS were not only expressed in terms of ‘understanding’ the intervention and how it works, but also in awareness of and understanding in the role of trust in the intervention design. Participants consistently indicated that the acceptability of APS was predicated on trusting “the provider not to reveal their identity to their sexual partners,” as there was a prevailing risk that “the provider can easily mention their names,” but noting that “they are not allowed to do that”.

The reason some people would refuse to give the contact details, let me start from there, is because she/he knows that at the end of the day the partner would want to know who gave out his/her contacts and will go back to her/him and ask why they gave out their contacts. That is why some people don’t like doing that. (IDI 035, MP)

Finally, FICs and MPs explained that an important component of APS was in understanding the design of the intervention, with a HCW playing a critical role in counseling and reaching out to partners instead of the participant having to do it themselves. FICs explained how their MPs were more willing to listen to a HCW than their female partners, where “if he hears it from someone else then he might take it seriously,” and “it is better if the provider just contacts them and notifies them.” Both female and male participants “found it hard to go and talk to them [sexual partners] myself,” and that APS was effective in doing so instead.

Perceived effectiveness

Sekhon et al. defines the perceived effectiveness construct as the extent to which an intervention is expected to achieve its purpose from the participant’s perspective [25]. Participants distinguished their framing of effectiveness in the context of acceptability across several sub-themes, including effectiveness of improving health, of impact on their relationship with their partner, and the importance to the success of the intervention which was the role of the HCW eliciting and notifying partners.

Participants expressed favorable benefits of participating in APS in that they understood their own HIV status, and that of their partner. Participants felt that they had more information about themselves due to APS, and one FIC explained that she is “aware of many things which I didn’t know before,” in regard to her own status. Though the APS intervention is designed for confidentiality, participants felt that engaging in the intervention supported their male sexual partners in finding out their own statuses, which ultimately could protect the health of the FIC. One FIC explained why the effectiveness of APS was important, in that “I wanted [my partners] to be reached and talked to so that if they also find it good they can also enroll into care.”

Though FICs expressed the motivating role of protecting partners and family more than did MPs, some MPs expressed similar urgency to engage with the intervention from the lens of perceived effectiveness.

You see I was considering that if I kept quiet then I would have died and my children would remain alone without a father. So I got courage and decided that let me just be open up so that both of us could be helped. (IDI 050, MP)

Participants also explained how APS supported their relationships by facilitating more openness and communication by leveraging provider-initiated contact for HIV testing.

So if I give out his contact information in secret and then he is traced and found that he is also on care then that would make us to be open to each other. (IDI 061, FIC)

This experience was a good one because at least it saved us from the chaos that would have erupted if I was the one who went home and told my wife about my status. When you go home and tell your wife about it, she will be shocked and it will create problems in the house and cause a lot of blame games on who brought it to the other. However, because the results came back when we were both together, we went back and talked about it and sorted everything out amicably without causing chaos or problems in our marriage. (IDI 055, MP)

Trust in the confidentiality of the intervention was thus important, but there were also reflections that participants, particularly male participants, trusted the HCWs to be more successful in ensuring linkage to care of their partners than they could on their own.

The reason I chose her to do it on my behalf is because she is more knowledgeable on how to handle this issues than myself. She really challenged me and I trusted her to do a better job getting my wife and sexual partners on my behalf, counsel, test, and enroll them into care. So this is the reason why I decided to share their contacts with the provider because I don’t have the experience of counselling anyone…That is why I chose her to go and approach my sexual partners to handle the difficult parts. (IDI 055, MP)

MPs additionally critically noted that APS was effective in reaching them for HIV testing, and that they otherwise would not have known their status.

I did not know my HIV status but being that I was called and got tested made me know my status. I can say that it has really helped me up to now because I am still alive and I am continuing with my business. If it were not for the program, I would have died a longtime ago. (IDI 049, MP)

Finally, some FICs also explained how this risk of IPV motivated the engagement with APS and framed its perceived effectiveness, in that the role of the assisted notification served as a potential protective measure for themselves.

Interviewer: So it is fear that makes most women chose provider referral. Participant: Yes, it is mostly because of fear. Women would think about how their spouse or partner usually reacts to things and say they cannot dare face him. Some of them change when things happen and they become very wild like a lion or burn like fire and are very mad. So they will just think that if this guy reacts very badly to simple things, will he not kill me when he hears this? (IDI 067, FIC)

Self-efficacy

The self-efficacy construct is defined as the participant’s confidence that they can perform the behaviors required to participate in the intervention [25]. This links closely with the perceived effectiveness expressed above, with respondents framing their ability to engage with the intervention often in the context of how well the intervention would work in preserving confidentiality and in helping oneself and one’s partner. Participants noted that their initial hesitation and lack of confidence in the intervention’s requirement to name partners affected their views of the acceptability of APS.

It was not easy for me to make that decision because I felt she would breach the confidentiality and disclose my status to my sexual partners. But with time, after I saw her experience and the way she managed to counsel me and convince me to get tested and enroll into care, I then said let me just let her reach out to my sexual partners because I felt she was the only one who could help me with it successfully. (IDI 055, MP)

While lack of confidence was noted throughout, the primary sub-theme for self-efficacy was the more specific fear of partner elicitation, due to barriers including stigma and the burden and risks of disclosure, including relationship dissolution and violence. Some participants felt a real risk in engaging with APS across these components, and this impacted their confidence in engaging with the intervention.

When you have someone you want to get married to, you will start thinking, ‘What if he is invited to the hospital and he finds out that I am the one who gave out his contacts. Will he leave me?’ So this is where the fear comes in. What will happen to me when he leaves me? (IDI 056, FIC)