Introduction

In Uganda and elsewhere globally, gender norms are a significant driver of the HIV epidemic, which is primarily driven by heterosexual transmission between cis gender women and men [1]. For women, gender norms reinforcing gender inequity and women’s lower status can limit women’s autonomy in HIV prevention and treatment decision-making [2,3], create reliance on men for economic survival [4], and drive gender-based violence (GBV), a risk factor for HIV [5]. Consequently, women are at elevated risk for HIV throughout sub-Saharan Africa, particularly adolescent girls and young women [6]. In Uganda, HIV prevalence is 6.5% vs. 3.6% in adult women vs. adult men [7]. For men, masculine norms, or cultural expectations of what it means to be a man, similarly drive men’s HIV risk behavior [8], while norms of masculine strength, self-reliance, and respect intersect with HIV stigma to undermine men’s engagement across the HIV care continuum [9,10]. Uganda has made progress to close the gap between men and women’s engagement in HIV care in recent years; however, continued efforts focused on men’s HIV testing and care support are still needed [11]. As such, it is critical that HIV services are designed to be responsive to issues related to gender.

The spectrum of gender responsive health programs range to include services that are gender-sensitive (recognize the different needs of women and men but are not focused on gender inequalities), gender-specific (acknowledge gender norms and consider women’s and men’s specific needs), and gender-transformative (address the cause of gender-based health inequities and work to transform gender roles, norms, and relations) [12,13]. Strengthening the gender responsiveness of HIV care aligns with the President’s Emergency Plan for AIDS Relief (PEPFAR)’s priority to ensure client-centered HIV care [14]. PEPFAR defines client-centered care as services that meet the needs of individuals by increasing convenience, providing welcoming services to diverse populations, making services supportive and responsive, and engaging communities and stakeholders [14]. Considerable work has been done to build the gender-responsiveness of health and community systems in the region, such as the strengthening of gender-sensitive data monitoring systems [15] and the use of gender-specific peer educators and support groups [16,17]. Gender-transformative programs have also gained considerable support in their ability to improve HIV outcomes [18–22].

However, there is a gap in research that focuses on building health care providers’ capacity to provide gender-responsive, client-centered HIV care, even though providers are the first line of support in the delivery of antiretroviral treatment (ART) services. Providers are also central to the integration of gender-transformative programs into health systems, which can be complex [23]. For HIV care providers to be responsive to their clients’ gender-related issues, they must be aware of and motivated to respond to them. In Uganda, limited training on GBV and low awareness of the prevalence of intimate partner violence among health workers were identified as barriers to the implementation of GBV screening [24,25]. In Malawi and Mozambique, HIV care providers reported low awareness and empathy for men’s HIV care engagement barriers [26].

Once aware and motivated to respond, HIV care providers must have acquired the skills needed to respond to clients’ gender-related issues. Client-centered communication approaches include tactics to elicit and understand a clients’ perspectives and needs within their own unique psychosocial and cultural contexts [27], which could help providers assess and understand gender norm-related barriers to HIV care. Client-centered care delivered through a gender-sensitive lens might also help providers recognize and mitigate power divides and communication issues that are shaped by gender norms. Research shows masculine norms related to respect and autonomy make men especially sensitive to feeling reprimanded or in the position of “learner” relative to health workers [28,29]. Kenyan HIV care providers reported challenges establishing clear roles and sharing power with clients due to conflicting gender versus patient/provider identities [30]. Interventions to strengthen client-centered communication could help mitigate these issues, as it emphasizes the importance of respect, shared decision-making, active listening, and displays of empathy. Research strongly links client-centered care to improved provider communication, linkage and retention in care, adherence, and patient satisfaction [31,32]. Thus, promoting client-centered care within a gender-norms lens may be an effective approach to strengthening gender-sensitive HIV care and improve client outcomes in sub-Saharan Africa.

In addition to being aware, empathetic towards, and able to respond to clients’ gendered barriers to HIV care, providers need to be conscious of how their own biases and attitudes might affect their provision of HIV care to not perpetuate gender inequities in client outcomes. One study found health workers held negative bias towards men as HIV care clients; they viewed them as “bad clients” based on assumptions about men being prideful and selfish and expressed little sympathy towards their healthcare engagement challenges [26]. Gender biases held by HIV care providers towards women have also been identified; in a study in Kenya, HIV care providers viewed women as more deceitful about ART adherence, due to HIV status disclosure barriers, and judged their reasons for missing appointments as less legitimate than men’s reasons (home vs. work responsibilities) [30]. Culturally-grounded attitudes that unmarried women should not be sexually active have been linked to the stigmatizing treatment of adolescent girls and young women and as a barrier to their receipt of pre-exposure prophylaxis (PrEP) and other HIV services in sub-Saharan Africa [33,34]. Gender biases also intersect and interact with other biases towards people living with HIV, key populations, and other marginalized groups. For example, sex worker stigma is similarly tied to gender norms about women’s sexuality [34]. Health workers in Uganda have reported negative attitudes towards transgender and sexual minority populations [35,36]. As such, interventions focused on bias reduction, toward both women and men, should be a core element of gender-sensitive capacity building for HIV care providers.

To increase the quality of HIV care for both women and men, we developed and pilot tested a training program in gender-sensitivity for HIV care providers and clinic staff in central Uganda. The goal of the training was to increase HIV care providers’ awareness of how gender norms affect client’s HIV care engagement, reduce their gender and related biases, build their client-centered communication skills and their ability to implement gender-specific and gender-transformative approaches to counseling men and women. These intervention goals were focused on cis-gender women and men in rural and peri-urban areas – the main population served by the study clinics. In this paper, we report on the mixed methods findings from a quasi-experimental pilot trial that examined the intervention’s preliminary effects on HIV care providers attending the training, as well as on HIV clients receiving care at the clinics where the training was conducted, compared to those at clinics not receiving the training.

Following recommendations for a stage 1b pilot trial (pilot of a new behavioral intervention), our aim was to gather preliminary evidence of trends towards change [37]. We anticipated providers in the training clinics would bolster their skills to provide gender-sensitive care compared to providers in clinics that did not receive the training (standard-of-care). For ART clients, we anticipated those in the training clinics would have increased ART and HIV clinic appointment adherence and explored the intervention’s preliminary effects on related indicators of perceived quality of/satisfaction with HIV care and experience of HIV stigma.

Methods

Description of study arms

Control.

In the control clinics, the providers received no intervention; thus, the intervention was compared to the standard-of-care implementation of MOH ART guidelines [43]. After the study ended, we provided the training to the control clinics to ensure equal benefit to participating clinics.

Intervention.

The intervention was initially informed by existing strategies developed by Dovidio and colleagues [44], based on research that has shown that increasing provider motivation, awareness, skills, empathy, and emotional regulation can prevent implicit racial and gender biases from affecting clinical judgment and behavior. For the current intervention, these strategies were adapted to increase providers’ knowledge, motivation, skills, and empathy to equitably deliver Ugandan MOH ART program guidelines [43] to men and women clients. In our adaption of these guidelines, we sought to increase awareness of HIV gender disparities, increase empathy/skills to counsel men and women’s gendered barriers to care, promote collaborative decision-making with clients, and promote client-centered and gender-sensitive/specific communication in ART and other HIV-related counseling. The development of content based on these strategies was informed by: 1) the existing literature from sub-Saharan Africa on gender norms’ influence on client engagement in HIV care and on the provision of quality HIV care (focused on the provider-client relationship); and 2) a local needs assessment that was carried out in ART clinics and local communities as a pre-cursor to the trial (inclusive of qualitative data collected from HIV care providers, clients, community members, and HIV care stakeholders); 3) several rounds of direct feedback on the approach and materials from HIV care stakeholders from Mildmay Uganda, the Uganda MOH, and local District Health Teams.

The final intervention package (outlined in Table 1) included four training sessions facilitated by two national trainers experienced in the capacity building of health workers in topics related to HIV and gender. There were a total of 3 training cohorts in the intervention arm (participants’ cohort aligned with their clinic). Sessions 1–3 included directive education, facilitated group discussions, interactive role plays, and vignettes. Session 4 was a practicum-based session that reviewed the content delivered in sessions 1–3. Session content was delivered in Luganda and English (tailored to the needs of the group) and participants received a training workbook to follow along and take notes in.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Overview of content per session in a gender-sensitivity training for HIV care providers and clinic staff that interact with HIV clients, Uganda 2021-2023.

https://doi.org/10.1371/journal.pgph.0004247.t001

Sessions 1 and 2 were delivered over two consecutive days in an off-site setting (outside of the health facility). As outlined in Table 1, Session 1 content was comprised of an introduction to gender (e.g., sex vs. gender, gender norms/roles), and gender’s relationship to: HIV acquisition, barriers to HIV care engagement, and HIV gender disparities—content was grounded in examples specific to the local context. We introduced participants to the spectrum of gender transformative health programs [12,13], with examples applied to HIV services. We engaged participants in an interactive, role play-based workshop that explored how HIV can be viewed as threatening to core elements of masculinity/femininity, leading to HIV stigma, which we referred to as the “HIV threat framework” and created based on prior work from Naugle et al. [45] and the broader literature on gender norms and HIV [9,10]. Through interactive roleplays, providers were taught gender-sensitive, gender-specific, and gender-transformative ways to identify and reduce the “HIV threat” for clients, and thus, mitigate HIV stigma. For example, facilitators encouraged providing referrals to services to reduce HIV stigma (e.g., gender-sensitive/specific support groups) and modeled ways to counsel clients on ART’s ability to restore/maintain their gender roles (gender-specific) and to reject gender norms/roles that conflict with HIV care engagement (gender-transformative).

Session 2 included an introduction to bias and stereotypes, with discussion-based explorations of locally relevant stereotypes and biases related to gender, HIV and other marginalized groups that may negatively affect the provision of quality, equitable HIV care. Through an adaption Bower et al.’s [46] “three-perspective framework,” facilitators taught providers to assess clients’ needs through the client perspective (e.g., personal barriers to care, perceptions of illness), their own perspective (including reflection on one’s own biases), and the societal perspective (incorporating the broader social/cultural context, focusing on gender norms). Facilitators led individual and group exercises that were adapted from Bower et al. [46], applying this framework to a “critical incident” or a real, past experience with a “difficult” client (including both a man and woman client), with the goal of helping providers recognize how their own biases and emotions may negatively affect their provision of HIV care, and to increase empathy for client challenges, including those related to gender. Facilitators presented evidence on how external stressors and stressful work environments can increase the likelihood that bias enters clinical interactions, and shared how to deploy evidence-based emotional and stress regulation techniques (e.g., breathing exercises) to mitigate bias and stress from affecting client relationships [44].

The other modules in session 2 covered how gender intersects with power dynamics in provider-client relationships, and shapes men and women’s communication styles, with guided discussions on how gendered power in provider-client dyads might impact providers’ relationships with clients. Client-centered care/communication was introduced as an approach to providing gender-sensitive care to clients, helping to overcome power imbalances and improve verbal/non-verbal communication with clients, and providing a framework to elicit and respond to gender preferences for care and gendered barriers to HIV care. Through participatory role playing specific to gendered barriers to HIV care including “HIV threats” (i.e., HIV stigma viewed through a gender lens), participants practiced the application of client-centered techniques for building rapport, eliciting the clients’ agenda, verbal and non-verbal communication skills, expressing empathy, and engaging in collaborative decision-making, using content adapted from Hofert et al. [47].

Session 3 consisted of a 2-hour on-site group training session conducted at the clinics. The content focused on gender-based violence (GBV), developed to reinforce the MOH’s existing GBV training, highlighting GBV’s effects on health and the health worker’s role in responding to GBV [48]. We built on this content to reinforce and not duplicate existing MOH GBV content but enhanced it by more explicitly linking between GBV to HIV (the MOH training is for all health workers) and by linking GBV to session 1 and 2’s content on broader gender norms and inequity. Facilitators emphasized the importance of responding to GBV as part of providing gender-sensitive, client-centered care.

The final session was a “refresher” or review session of sessions 1–3, with a focus on the practical application of the content into care. The session began with one of the facilitators directly observing providers’ interaction with a client in the ART clinic. The facilitator made notes on the interaction and completed a checklist that included a list of client-centered communication and gender-sensitive counseling techniques; the checklist was adapted from an existing client-centered communication assessment for providers [47]. The checklist was piloted before use and adapted with feedback from the facilitators. One facilitator observed up to four providers on a given day. After the observation, the facilitator provided individual feedback to the providers based on the observed interaction, and then met with the small group of providers. In the group discussion, the facilitator provided a review of core session content following a structured guide, and then facilitated a group discussion following semi-structured prompts on successes and challenges in translating the training’s content into practice. The facilitator provided suggestions for overcoming the identified challenges, and encouraged group-problem solving, eliciting other providers’ suggestions based on their own experiences.

Data collection procedures and measures

Quantitative.

Experienced interviewers conducted a computerized, structured interview following enrollment and again at approximately 6- and 12-months follow-up for participants in both the HIV care provider/staff and client cohorts. Interviews took place in a private setting, including the clinic or another agreed upon location, or over the phone, if necessary. Participants in both cohorts received 10,000 Ugandan Shillings (UGX) (~$3) per baseline and 6-month assessment and 18,000 UGX (~5 USD) for the 12-month assessment, aligned with the ethics review board standards. At baseline for providers, socio-demographics, occupational role, and years working at the current facility and with HIV clients were collected. For HIV clients, socio-demographics were also collected at baseline, along with health and HIV history. As detailed in Table 2, measures to examine change in the intervention’s hypothesized primary and secondary outcomes for the HIV provider and client cohorts were collected, along with additional validated measures for potential covariates relevant for HIV care engagement outcomes.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Quantitative measures, gender-sensitivity provider training intervention pilot trial, Uganda 2021–2023.

https://doi.org/10.1371/journal.pgph.0004247.t002

Qualitative.

All participants who attended the training across the three intervention sites were invited to participate in an exit focus group discussion that took place after the 12-month quantitative assessment. Focus groups were designed to be homogenous by type of health care role and each group included 4–8 people, aligned with best practices in qualitative research in focus groups [49]. A total of 46 providers attended 1 of 7 focus groups, which were conducted at each clinic and segregated by cadre (4 health worker groups and 3 lay health worker groups); 7 additional participants who were not available to attend a focus group discussion participated instead in an individual interview. Taken together, 87% of intervention arm providers completed a qualitative follow-up. A trained and experienced qualitative interviewer led the focus groups and interviews following an interview guide with questions that elicited participant feedback on their experience and perceptions of the training sessions, including questions on what they learned and what changes they made (if any) to their practice. An experienced qualitative interviewer, independent of the study team, moderated and transcribed the focus group discussions. The interviewer was trained on the strengths (capitalizing on group dynamics) and weaknesses (views produced at the poles of debates) of focus groups and were trained to probe neutrally to ensure that a range of views were represented. Focus groups lasted approximately 90 minutes, and individual interviews lasted approximately 30 minutes. Participants received 25,000 UGX (~7 USD) for their participation. All sessions were conducted in Luganda and audio recorded by a trained research assistant; transcripts were translated into English.

Results

Fig 1 displays the Consolidated Standards of Reporting Trials (CONSORT) study diagram. Of the HIV care providers/staff screened for participation, the final enrolled samples in each study arm were 61 in the intervention clinics and 83 in the control clinics; only 2 eligible providers (3.2%) declined participation in the intervention clinics and no eligible providers declined in the control clinics. At 6- and 12-month follow-up, 98% (n = 60/61) of providers completed the 6- and 12-month questionnaires in the intervention clinics, compared to 94% (n = 78/83) and 96% (n = 80/83) in the control clinics. Provider attendance of the intervention sessions was 86%. Of all planned content, 100% of intervention steps were delivered; 78% were delivered with fidelity or as intended by the study investigators, measured through review and assessment of a random selection of 20% of recorded session transcripts.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. CONSORT diagram.

https://doi.org/10.1371/journal.pgph.0004247.g001

For the client cohort, a total of 119 individuals per study arm were recruited at baseline. In the intervention clinics, 81% (97/119) completed the 6-month assessment and 88% (105/119) completed the 12-month assessment. In the control clinics, 76% (n = 91/119) completed the 6-month assessment and 90% (108/119) completed the 12-month assessment.

Table 3 details the sample characteristics of the HIV care provider/staff cohort at baseline (N = 144). There were more women (72.2%) than men (27.8%)— an expected imbalance, as there are more women health workers than men in Ugandan health facilities. The full range of provider occupations are detailed in Table 3; 62.5% were considered lay health workers or other non-clinical staff with regular interaction with HIV clients vs. 37.5% certified health workers or counselors. Providers had been working at their specific health facility an average of 4.7 years (standard deviation [SD]=5.3) and practicing in their current role an average of 7.1 years (SD = 7.5). No statistically significant differences were identified between arms on these or other baseline characteristics, but providers in the control clinics were in their current clinic and role slightly longer than those in the intervention clinics and had a slightly higher percentage of lay staff (see Table 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Sample characteristics at enrollment by study arm for the HIV care provider/staff and client cohorts, with comparisons between study arms, gender-sensitivity training intervention pilot trial, Uganda 2021–2023.

https://doi.org/10.1371/journal.pgph.0004247.t003

The client cohort (Table 3) was balanced by sex in both arms, based on purposive, equal recruitment of men and women. The average age of the overall sample was 39.2 (SD = 11.3), and the majority were from the Buganda tribe (66.0%) with primary education or less (69.7%). Most participants were already on ART at baseline (92.4%), among which, the average number of years on ART was 4.4 (SD = 4.9). The analysis for baseline equivalence identified several characteristics where the two arms differed at baseline (p < 0.05). Details are in Table 3, but in summary, clients from intervention clinics, on average, knew their status for fewer years (B = -2.70, 95% CI = -4.83, -0.56, p = 0.01) and were on ART for fewer years than those from control clinics (B = -1.88, 95% CI = -3.77, 0.02, p = 0.05). Intervention clients reported more hazardous alcohol use (OR = 2.63, 95% CI = 0.74, 4.51, p = 0.006), food insecurity (B = 0.07, 95% CI = 0.22, 0.36, p = 0.64), lifetime experience of intimate partner violence (OR = 2.81, 95% CI = 1.28, 6.16, p = 0.01), higher monthly income in units of 40,000 UGX (~10 USD) (B = 2.63, 95% CI = 0.74, 4.51, p = 0.006), and less travel time to the HIV clinic than the control in units of 30 minutes (B = -0.91, 95% CI = -1.32, 0.51, p < 0.001).

Table 4 displays the results of the models testing the intervention’s effect on each outcome, as well as the descriptive statistics for each outcome at all time points. Table 5 provides an outline of the main qualitative findings with illustrative quotations. In the narrative below, these findings are presented together, organized by overarching thematic areas on the explored intervention effects.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Exploratory outcomes at baseline, 6-month, and 12-month follow-up by condition (unadjusted) and the arm-by-intervention effect (adjusted) for the HIV provider/staff cohort and client cohort, gender-sensitivity provider training intervention pilot trial, Uganda 2021–2023.

https://doi.org/10.1371/journal.pgph.0004247.t004

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Qualitative themes with select, illustrative quotations from exit focus groups/interviews with HIV care providers/staff who participated in the training, gender-sensitivity provider training intervention pilot trial, Uganda 2021–2023.

https://doi.org/10.1371/journal.pgph.0004247.t005

Thematic Area 1: Overall Competence in and Provision of Gender-Sensitive HIV Care (Knowledge, Attitudes, Self-Efficacy, Skills)

Providers who received the training reported a significant increase in the overall gender-sensitive competence score compared to those that did not; the overall effect was statistically significant (Wald χ² = 20.94, p < 0.001) with the difference observed at 6-months (B = 0.24, 95% CI = 0.12, 0.37, p < 0.001), but not maintained at 12-months (B = -0.02, 95% CI = -0.14, 0.10, p = 0.80) (see Fig 2a). As depicted in Figs 2b and 2c, this effect was stronger for men than women at 6-months, and for lay health workers/non-clinical staff compared to certified health workers and counselors, with the effect attenuated but still maintained for lay health workers/staff at 12-months (statistics for interactions are reported in Tables A and B in S1 Text).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. a-c. Graphic depictions of the intervention by time effect comparing HIV care providers and staff in the gender-sensitivity training intervention and control clinics in competence in gender-sensitive care (primary exploratory outcome) with gender and cadre interactions.

Notes: Models control for clinic, gender, and cadre. Error bars represent 95% Confidence Intervals.

https://doi.org/10.1371/journal.pgph.0004247.g002

The qualitative exit focus group/interviews corroborate the quantitative data’s findings on providers’ gender-sensitive care competence, and provides insight into specific improvements, as well as gaps, in knowledge, attitudes, self-efficacy, and related skills (presented next and in Table 5).

Knowledge.

In qualitative focus groups/interviews with training attendees, participants shared that the training had improved their understanding of gender and HIV. Specifically, providers described learning about the difference between gender and sex, how HIV gender disparities tie to broader gender norms, roles, and inequities, and how gendered barriers to HIV care affect women and men. Discussions in the focus groups on gender vs. sex demonstrated that the training helped some providers understand that people’s sex may not match their gender identity or roles, with examples of women doing men’s work or men that do not conform to masculine roles, and statements that clients should be treated based on their gender identity or non-conformism, as opposed to their sex.

In addition, providers often gave applied examples of the link between gender norms and clients’ challenges; common examples included masculinity-related barriers to engaging men (e.g., work responsibility, stigma, difficulty communicating with them), HIV status disclosure issues for women tied to dependence on male partners, and GBV. Quotation 1 (Q1) in Table 5 also showcases the salient theme that providers gained deeper insight into gender beyond GBV; at the start of the program, participants often thought “gender” referred only to “GBV,” and used the terms interchangeably, due to a strong emphasis of GBV in HIV programming in Uganda. Many participants shared how the training helped them to understand gender’s effect not only on GBV, but also on HIV outcomes.

While gains in knowledge were salient throughout all focus groups and interviews, some discussions more heavily focused on men’s gendered barriers to care than women’s – this could suggest the intervention affected knowledge about men more than about women. Further, a subset of discussions displayed an overgeneralization or oversimplification of the content learned through statements, such as all men are too busy to come to the clinic, without deeper discussion of the role of gender norms provided or recognition that such statements may not be true about all men.

Attitudes towards providing gender-sensitive care.

Taken together, providers’ narratives demonstrated increased motivation to provide gender-sensitive care to clients. In addition, across all focus groups, providers explicitly expressed increased motivation to better address GBV among HIV clients by following existing MOH GBV protocols (for screening and referral, which the training reinforced) and to better counsel GBV-affected clients. Many providers disclosed that addressing GBV among clients was not previously a priority to them, and a few providers made clear that they were not previously connecting the importance of addressing GBV to their goals in improving HIV care engagement (see Q2 in Table 5).

Self-efficacy and skills to provide gender-sensitive care.

In focus groups/interviews, providers displayed increased self-efficacy for providing gender-sensitive care and described specific skills that they acquired relevant to gender-sensitive counseling and service provision. A dominant theme, particularly among women lay workers, was that they perceived gains in their confidence and communication skills to respond to men, based on their understanding of how masculine norms affect men’s communication style and presentation to care (see Q3 in Table 5). Related to the reported change in attitudes towards responding to GBV, there was also reporting of increased self-efficacy to follow GBV protocols and skills to counsel GBV-affected clients (see Q4 in Table 5).

Providers gave examples that showcased their application of specific skills acquired from the training in their client interactions. Examples mostly fell within gender-sensitive or gender-specific approaches, such as working on men faster who have work conflicts, linking clients to appropriate gender-specific services (e.g., support groups, home visits, male champions), and the use of the “HIV threat framework” to identify and counsel clients on gender-specific stigma (see Q5 and Q6 in Table 5). While the training also discussed gender-transformative counseling approaches aimed to move clients towards gender equitable norms, there were fewer examples of this discussed in practice. Further, there were a few instances of providers sharing examples of counseling patients in ways that maintain traditional gender norms and equity. In examples provided in Table 5 (Q7), one health worker explains how they now coach women on how to not evoke violence from their spouse and another (Q8) explains how they now counsel women on faithfulness as a strategy to reduce HIV transmission, but not men.

Discussion

The findings of this pilot trial provide preliminary support for a training intervention’s ability to improve providers’ competence for gender-sensitive HIV care in health settings in Uganda. Through a mixed methods evaluation, the findings specifically support the training’s potential to improve providers’ knowledge, attitudes, and skills to respond to clients’ gendered barriers to HIV care engagement. The findings also support the promise of using of a gender norms lens as an appropriate framework to increase client-centered care and reduce gender and other related biases towards clients. In this pilot, the observed provider effects largely did not translate into clients’ quantitatively captured outcomes related to HIV care engagement, perceived quality of provider communication, or satisfaction with care. However, providers reported improvements in their relationships with clients, and in turn, perceived a positive effect of the intervention on client outcomes. Further, the client cohort data demonstrated an increase in men’s perception of their level of participatory decision-making in their HIV care and a reduction in HIV stigma among clients in the intervention clinics. As a pilot trial, these effects are preliminary and need to be replicated in a larger, fully powered trial.

This pilot trial found preliminary support that participation in this training increased providers’ competence to provide gender-sensitive care. This was evidenced through modest, greater relative change in intervention vs. control participants in our overall competence measure. The measure was inclusive of providers’ knowledge of HIV gender disparities, perceived importance of providing gender-sensitive care, and self-efficacy to provide gender-sensitive care. The qualitative data validated and expanded on these findings—showcasing increased understanding of gender vs. sex, how gender affects men and women’s communication and care engagement, and the importance of considering gender norms in providing HIV services and counseling to clients. Notably, the intervention’s effect on this outcome was attenuated by 12-months, highlighting the potential need for more follow-up support. In addition, the effect was larger for men vs. women providers and for lay health workers and staff vs. certified health workers and counselors. These groups may have had less previous training on the topic, but there may also be a need to consider tailoring the content by group in the future. The qualitative data also highlighted a potentially positive effect of the intervention on responding to GBV—with narratives to support increased perceived importance of following GBV protocols and skills to handle GBV clients. Thus, this gender-sensitivity training could help overcome known barriers to implementation of existing GBV protocols in HIV clinical settings in Uganda and similar settings [25], but in the future testing of this intervention, quantitative metrics should be added to validate these findings.

The training intervention included content to help providers strengthen their client-centered communication skills and use them to help identify and address clients’ gendered barriers to HIV care. Providers who attended the training qualitatively reported efforts to build rapport with clients, reduce harsh language and judgement, and use perspective taking to put themselves in clients’ shoes. Through quantitative measures, there was also an intervention effect on increased empathy towards clients—which others have linked to better treatment of patients living with HIV [56]. We anticipated improved client-centered care might also address some previously identified barriers to HIV care engagement that intersect with gender in provider-client relationships (inequitable power dynamics, gendered communication styles, masculine norms of respect conflicting with harsh treatment from providers) [9,10]. For men who received the intervention, there was a sustained intervention effect on the perceived level of participatory decision-making between clients and providers. This may have been an outcome of this aspect of the intervention, which may have been most relevant for improving care with men, who providers (mostly women) discussed having difficulty communicating with prior to the intervention.

In our study, however, other quantitative indicators of client-centered care did not have an effect or improved in the control more than in the intervention. As a pilot trial, it may be that the study needed more clusters, a larger sample size, and more time to see these results fully translate into client outcomes. Providers did note that relationship-building takes time. Moreover, the study highlighted considerable barriers to the provision of client-centered care related to the clinic environment (time/space constraints). Others have pointed to these same barriers as limiting the provision of client-centered care throughout sub-Saharan Africa, as well as a dearth of research on how to improve client-centered care in the region [32]. A recent pilot trial in Kenya showed the promise of research with this aim—reporting increased viral suppression among clients living with HIV in clinics that received a patient-centered care intervention [57]. Considering the challenges to scaling up client-centered care in this and similar settings, adding more than one practicum-based follow-up in the future scale up of this intervention may help support the translation of content into practice in facilities. In addition, future research focused on gender-sensitive capacity building could also explore ways to formally change institutional procedures, practices, or environmental contexts in ways to support gender-sensitive, client-centered care provision.

This pilot also supports the potential positive effects of pairing client-centered care promotion with bias reduction training, grounded in a gender norms lens. The qualitative findings highlighted increased empathy towards different groups of clients that providers held gender bias towards. Among the general population, which was the training’s focus, negative attitudes were expressed towards men in general, non-suppressed clients, young women, youth generally, and clients who drink alcohol. Similar to others [26], a salient finding was providers’ expressed frustration and difficulty communicating with men, who were perceived as rude and impatient. Our findings suggest that teaching providers about how gender norms affect men’s care engagement can increase their empathy towards men and build skills to counsel them. Future research and programming aimed to address the “HIV blind spot” in men’s care engagement should consider these findings and the role HIV care providers can play in either insolating or engaging men living with HIV.

Although this study focused on cis-gender heterosexual women and men, introducing gender as a social construct provided an entry point for discussions on gender and sexual minorities, and intersections with other stigmatized groups. As a result, after the training, providers expressed increased empathy towards, and reported intentional efforts to be client-centered with, key populations, including gender and sexual minority individuals and sex workers. Other intervention studies in Uganda have shown that sensitivity trainings with providers can improve their attitudes and empathy for key populations [35,58]. Using a broad gender-norms lens for sensitivity trainings can be explored as a potentially culturally acceptable way to increase provider empathy for and reduce bias against all clients. This has important implications in a climate that may not currently support sensitivity trainings explicitly directed towards these groups.

Increased client-centered communication and reduced bias towards clients may explain this pilot study’s observed reduction of HIV stigma reported among clients in the intervention arm clinics vs. those in the control arm clinics. Providers characterized their communication with clients before the training as harsh and lacking empathy, especially towards those struggling with treatment adherence. Providers also described stigmatizing treatment of some groups, such as isolating men and those who drink, reprimanding young girls and sex workers for the way they dress, and, in some cases, victim-blaming rape victims. Such treatment is likely to increase a clients’ experience of HIV stigma, which is associated with poor care engagement [59,60].

The training may have also reduced HIV stigma among clients through the content that explained clients’ experience of HIV stigma through a gender norms lens. Facilitators discussed men and women’s gender norms and roles, and illustrated how HIV can be viewed as threatening to those ideals. Facilitators taught providers counseling techniques to elicit clients’ “HIV threats” and mitigate HIV stigma by: 1) counseling them away from norms that do not align with HIV care engagement (gender-transformative), 2) highlighting ART’s ability to help them restore/maintain their roles (gender-specific), and 3) linking them to services that could reduce HIV stigma (e.g., male champions, support groups) (gender-sensitive/specific). While providers’ qualitative narratives included applied examples of the latter two, a potential gap was highlighted in the focus groups through relatively less discussion of the use of gender-transformative counseling. Moreover, there were a few examples of providers citing the training on gender norms but using it to uphold traditional norms that underpin gender inequity between men and women. Others have warned about similar unintentional consequences of gender-focused public health work [61]. Taken together, these findings support the continued exploration of interventions that address HIV stigma reduction through a gender norms lens, but underscore the need for intensive training and oversight in the implementation of gender-sensitive and transformative interventions to ensure all messaging is received as intended.

This pilot trial is not powered to fully evaluate this intervention’s effects. While these data support this intervention’s promise, the findings should be replicated and expanded on in a future, fully powered trial. In addition to the findings presented here, this trial collected mixed methods data to examine acceptability, feasibility, and factors affecting implementation [39], the findings of which will inform the refinement of this intervention before future testing. Beyond the sample size limitations of a pilot trial, the pilot had few clusters, limiting our ability to fully account for within-cluster correlation in our analysis. Cluster randomization was necessary to reduce the risk of contamination between arms and to establish procedures for a future multisite trial. However, having only two clusters introduces the risk of confounding variables, and we did identify differences between clients in the two arms at baseline. Clients in the intervention clinic had higher reporting of several factors known to negatively affect HIV care engagement, including intimate partner violence history, alcohol use, and food insecurity. While we controlled for baseline differences in analyses, these differences may have reduced the likelihood of seeing an intervention effect in this group relative to the control. We also saw changes in the control arm on different outcomes, which could be explained by measurement reactivity or external factors, such as other programs or trainings, which cannot be ruled out with this pilot design. Given the strong covariate balance in the provider cohort and the small sample size, we did not apply inverse probability of treatment weighting, though we acknowledge it as an alternative approach in larger studies with greater imbalance. Another limitation of the study is the use of mainly self-reported measures. In a future hybrid effectiveness-implementation trial, the study will collect viral load directly from clients. Consequently, participants’ responses may have been influenced by recall and social desirability bias.

The mixed methods design of this study is a strength. The two data sources served to validate each other, and the qualitative data provided insight into providers’ experience with the intervention and into nuanced findings that are difficult to capture quantitatively, such as the findings on provider bias. However, the qualitative data are also subject to social desirability bias. While the exit focus groups were conducted by a moderator independent of study team, providers may have still felt the need to provide positive feedback on the intervention. A future trial could strengthen the mixed methods approach to further validate or contest these findings through observational methods, supervisor interviews to corroboration or challenge reported effects, and the collection of quantitative process measures that could capture some of the changes described qualitatively (such total number of clients screened for GBV).

Conclusions

This training intervention responds to calls to modify health systems, and HIV care specifically, to be more gender-responsive. Providers and staff with regular interaction with clients can play a critical role in reducing gender norm-related barriers to HIV care for men and women. This pilot trial provides preliminary support for a training intervention that improves gender-sensitive care by increasing providers’ awareness of gender norms and HIV gender disparities, building empathy and reducing gender and other bias towards clients, and strengthening client-centered communication skills to respond to clients’ gender-based preferences and barriers to HIV care engagement. This intervention should be fully tested in a future, hybrid effectiveness-implementation trial to establish its effects on provider and client outcomes. If found efficacious, this intervention would have important public health implications for Uganda and potentially for other settings where gender norms influence HIV outcomes. Future research could also explore the adaption of this approach for improving the provision of health services beyond HIV where providers’ delivery of services and clients’ outcomes are similarly affected by gender norms.

Supporting information