South Africa has the largest HIV epidemic in the world, with 7.9 million people infected and prevalence rates of 22.6% among females and 11.5% among males aged 15 to 49 years [1]. South Africa’s epidemic accounts for 19% of the global population of people living with HIV, 15% of new infections, and 11% of AIDS-related deaths worldwide [2]. KwaZulu-Natal Province is particularly affected; 40.8% of adults over age 15 years are living with HIV, and people living in this province have a 46% higher risk of infection than those outside the province [3]. South Africa has made considerable progress toward achieving the global 95-95-95 targets, but in 2023 only 83% of people living with HIV were on antiretroviral therapy (ART) and just 78% had suppressed HIV RNA [4]. Available interventions to support patient care engagement have not closed enough of the treatment gap, and innovative interventions to promote patient engagement are urgently needed, particularly those that support the needs of frontline providers, such as nurses.

Despite bearing nearly one-fifth of the global HIV burden, South Africa has only 3% of the world’s health care workers, a shortage caused by uneven geographical distribution of personnel, migration, and burnout [5,6]. Reported reasons for provider migration out of South Africa include dissatisfaction with pay, lack of advancement opportunities, and work-related stress [7-9]. Nurses constitute the largest group of health care providers in South Africa—81% of public sector health professionals in 2015 [10]. Yet, in 2019, there was an estimated shortage ranging from 26,000 to 62,000 nurses, and the shortage is expected to increase to 131,000 to 166,000 by 2030 [11]. Limited resources lead to overwhelming caseloads, resulting in burnout, low retention, chronic stress, hopelessness, and cynicism among nurses [9,12]. South African nurses are faced with chronic work overload and occupational stress [13-15], in part due to inadequate staffing, resources, and wages [9,14,16,17]. Since 2010, South Africa has increasingly adopted task-shifting as part of an effort to scale up its ART program; for nurses, this has meant an ever-greater task burden [18,19]. Household challenges can exacerbate work stressors, as the demands of nursing combined with traditional gender norms and family economic dependency contribute to chronic stress [20]. In a sample of South African hospital nurses, both personal and work-related stress were significantly associated with burnout [21]. Burnout—characterized by exhaustion, cynicism, and inefficiency—is a prolonged response to chronic emotional and interpersonal stressors [22] and is linked to job dissatisfaction and intention to leave [7,23].

Resource limitations and associated stress and burnout in nurses lead to negative patient outcomes that disrupt the HIV treatment cascade and compromise HIV outcomes. Patient engagement with HIV care (eg, attending appointments, adhering to ART, refilling prescriptions, managing side effects) is facilitated principally through nurse-patient relationships. Stress and burnout among nurses can negatively affect people living with HIV in a variety of ways, including (1) self-reported suboptimal care of and negative attitudes toward patients [24]; (2) decreased patient satisfaction [25,26]; (3) increased posthospitalization recovery time [26]; and (4) reduced ART adherence, HIV testing uptake, and prevention of mother-to-child transmission care due to experienced or anticipated negative provider interactions [27-29]. Patient perceptions of the patient-provider relationship are a core element of engagement and retention in HIV care [30,31]. Among first-time ART users, a strong patient-provider relationship has been found to be particularly important for maintaining adherence [32,33]; yet, patients have reported discrimination, stigmatization, and provider breaches of patient confidentiality at HIV clinics [34,35]. In general, patients prefer providers who have good communication skills and are empathetic, engaging, and validating [36]. These qualities build trust, which is positively associated with HIV care adherence [37]. When nurses who care for people living with HIV experience stress and burnout, patient-nurse relationships suffer and trust may be impacted; providers experiencing burnout report making treatment mistakes, yelling at patients, and rushing through diagnostic tests [38]. In South Africa, stress due to increased workload, staff shortages, and poor working conditions has been associated with nurse absenteeism [39], which further damages these relationships and cripples an underresourced health care system.

Even though HIV care nurses in South Africa experience particularly high levels of stress and burnout [9,13,14,16,40,41], there have been no systematic efforts to develop skills-based programs to help nurses manage and cope with the occupational-related and personal stressors that negatively affect patient outcomes. Stress reduction interventions have been used in a variety of medical settings across North America and have been found to decrease burnout and stress among nurses in oncology [42], pediatric intensive care units [43], and large health care organizations [44], as well as among nursing students [45]. In South Africa, an intervention that increases nurses’ resiliency may decrease stress, prevent burnout, and lead to improved patient engagement in HIV care and may be important to support nurse well-being. Resiliency is a “multidimensional construct that refers to the ability to maintain adaptation and effective functioning under significant adversity or challenging life conditions” [46], or stress. Skills that foster resilience include increasing social support, mastery (a sense of control over one’s life), optimism, acceptance, will to live, effective self-management, and generativity [47,48]. Resiliency interventions go beyond traditional stress management interventions by adding elements of positive psychology, which may teach individuals not only how to manage immediate stressors but also how to live purposefully, optimize health, and build “immunity” for better long-term stress management—key in this setting where widespread structural changes to decrease nurses’ stress are unlikely in the near future.

The stress response, or the “fight-or-flight” response, is a set of physical changes in the body in response to demands ranging from major negative life events to small daily challenges. The goal of the Stress Management and Resiliency Training-Relaxation Response Resiliency Program (SMART-3RP) is to decrease the physiological, emotional, cognitive, and behavioral effects of the stress response by promoting the physiological, emotional, cognitive, and behavioral effects of the relaxation response (RR) [46]. To achieve these goals, the SMART-3RP focuses on three major areas: (1) eliciting the RR, (2) increasing stress awareness, and (3) promoting adaptive strategies. Eliciting the RR involves learning strategies to reduce muscle tension, breathing rate, heart rate, and blood pressure, with the aim of helping to regulate the cellular responses to stress. Decreasing stress reactivity involves increasing one’s awareness of being in the stress response (negative thoughts, emotions, physical reactions, and behaviors) and learning skills to change or alter these components (eg, cognitive restructuring). Adaptive strategies focus on learning and using different techniques to promote positive growth and self-efficacy in response to stress. Given high levels of community stigma toward people living with HIV [35] and patients’ desire for providers who have good communication skills and are empathetic, engaging, and validating [36], adaptations to the SMART-3RP were made by adding stigma reduction and communication skills as adaptive strategies. The intervention was further adapted based on phase 1 focus group discussion (FGD) feedback [49]. It is the learning and integration of these 3 core elements that enables individuals to adaptively cope with stressful situations and thus increase resiliency. The SMART-3RP model has been adapted and refined to reflect additional clinician-patient dynamics and interdependence (Figure 1; novel components added to the model are indicated with an asterisk).

We aim to conduct a small proof-of-concept study (N=8) and a pilot randomized controlled trial (RCT; N=60) of the SMART-3RP, a skills-based intervention that has been tailored to enhance resiliency among professional nurses providing HIV care in the public sector in South Africa. We will test the feasibility and acceptability of all study procedures in preparation for a larger trial testing the efficacy and implementation strategies of the intervention. To assess feasibility, we tracked the number of participants screened, eligible, and enrolled; the number of treatment and assessment sessions completed; assessment duration; and reasons for declining enrollment and for prematurely leaving the trial. To assess acceptability, we used a measure of participant satisfaction and qualitative exit interview data. We hypothesize that the intervention and study procedures will be feasible and acceptable.

The sample size (N=60) was chosen based on studies of similar scope (eg, other pilot behavioral trials funded via the same mechanism) and established guidelines for pilot behavioral studies to establish feasibility [62-64].

To evaluate feasibility as a coprimary outcome, we will assess the number of nurse participants screened, eligible, and enrolled (ie, whether >70% of those screened enroll/agree to participate); the number of treatment sessions (ie, whether >70% of participants randomized to the intervention complete at least 2 out of 3 sessions) and assessment visits completed (ie, whether >70% of all participants complete at least 2 out of 3 assessments) in the intervention arm; assessment duration; and reasons for declining enrollment and prematurely leaving the trial. Acceptability (second primary outcome) will be evaluated using the Client Satisfaction Questionnaire (CSQ-8) [61] and qualitative exit interview data among intervention arm participants. Scores on the CSQ-8 will be categorized into 4 levels of satisfaction: ‘‘poor’’ (score 8‐13), ‘‘fair’’ (score 14‐19), ‘‘good’’ (score 20‐25), and ‘‘excellent’’ (score 26‐32). All qualitative exit interview data will be explored for categories and themes around intervention acceptability. Qualitative data will be analyzed using content analysis, facilitated by the Dedoose software [65], to organize data and facilitate analyses. Two study team members will independently review the transcripts to generate an overarching thematic framework for data interpretation, in which major and minor themes are identified. Thematic frameworks will then be compared for consistency, and discrepancies will be discussed until agreement is reached. Data will be reexamined, messages will be extracted and highlighted, and ongoing discussion between coders will allow for further theorizing and making interconnections between research questions, coding categories, and raw data [66].

With respect to exploratory evaluations of quantitative data and secondary outcomes, we expect to compare each outcome between groups using mixed model regression analyses, with participant ID as a random effect. Fixed covariates will include assessment time point (baseline, posttreatment assessment, and 3-month follow-up) and years working in HIV care. Other covariates, such as education, age, and baseline levels of stress, anxiety, and depressive symptoms, and clinic characteristics (eg, clinic size, availability of equipment, and urban vs peri-urban) will also be considered. We will consider exploratory mediational and moderational analyses, based on procedures outlined by Hayes [67]. We will examine patterns in missing data to see if missingness is related to measured participant characteristics. We will adjust for missing data using multiple imputation, as appropriate. We will also adjust for multiple comparisons using the method of Hothorn et al [68].

Quantitative data were collected by trained study staff and entered into REDCap (Research Electronic Data Capture) data management system using encrypted tablets and computers. All tablets were encrypted to the highest standard. Qualitative data were collected by trained study staff, and deidentified transcripts will be coded using Dedoose software [65] to facilitate analyses. Data files were sent to the investigators from the site via secure file transfer.

All research was performed in accordance with the Declaration of Helsinki and approved by local and international ethics committees. Ethics approval for the study protocol was obtained from the Human Research Ethics Committee (Medical) of the University of the Witwatersrand (240106, Johannesburg, South Africa Provincial Department of Health, the Massachusetts General Brigham Institutional Review Board (2024P001407, Boston, Massachusetts, United States), and letters of support obtained from the District and supporting organizations. Study updates and final findings will be communicated to study participants. All data shared between study sites are fully deidentified. Consent procedures complied with Institutional Review Board requirements by ensuring that participants had the option to opt out without adverse consequences (ie, affecting their employment). Study updates and deidentified findings will also be shared at Data and Safety Monitoring Board (DSMB) meetings organized by the Principal Investigator and the Community Advisory Board meetings convened quarterly at Wits Match Research Unit. Lastly, data will also be presented at national and international scientific meetings and conferences and published in peer-reviewed journals.

We have convened a DSMB comprised of individuals from South Africa with relevant expertise to this study; meetings are conducted in accordance with a DSMB manual outlining the responsibilities of the DSMB, DSMB meeting procedures, and templates for DSMB reports (developed in collaboration with the DSMB members). The investigators are responsible for compiling and submitting annual reports. The DSMB members have reported no conflicts of interest, and their membership was approved by the funder. The DSMB has met twice to allow for additional, external review of study procedures and participant safety. No significant safety concerns were identified. The charter for this study can be requested from the corresponding author.

Nurses providing HIV care in South Africa manage substantial amounts of stress arising from multiple sources, including the aftereffects of the COVID-19 pandemic and structural challenges associated with the nursing profession. High levels of stress can negatively impact the health of nurses themselves and may lead to negative patient outcomes via reduced patient satisfaction, reduced treatment adherence, and reduced testing uptake. There is a dearth of evidence as to how best to support the mental health of nurses, including the feasibility and acceptability of asking overburdened health care providers to take part in multisession behavioral interventions. We hypothesize that we have developed a feasible and acceptable intervention that has the potential to demonstrate meaningful changes for managing stress among nurses in a resource-limited setting. Next steps will include the RCT data analyses and reporting of findings and may include a larger trial to examine effectiveness and implementation outcomes.

Structural changes are needed to ensure the well-being of health care providers across all contexts. However, given that structural changes take time, money, and political capital to execute, we must develop interventions to support mental health while advocating for systematic change.