Study design and setting
The AMBIT Project’s SENTINEL study (16) is a repeated, cross-sectional, mixed-methods survey of differentiated service delivery for HIV treatment at public sector clinics in Malawi, South Africa, and Zambia. It includes surveys of ART patients, HIV testers, and treatment providers. Here we report results from the AMBIT round 1 provider survey conducted between April 2021 and January 2022.
SENTINEL study sites include 12 public sector clinics in Malawi, 21 in South Africa, and 12 in Zambia. Study sites were purposively selected to provide sufficient ART client volumes, variation in settings (rural or urban), and experience with DSD models for HIV treatment. Additional information on study sites is provided in Supplementary Table 1 and in the published protocol (16). DSD models were incorporated into national HIV treatment guidelines in 2018 in Malawi, 2016 in South Africa, and 2017 in Zambia. There were initially a large number of models implemented, each designed slightly differently and often with the support of nongovernmental partners and external funders (7). As countries, implementing partners, and facilities gained experience with DSD, a smaller number of national models were scaled up in each country, with a few bespoke, population-specific models remaining alongside them.
At the time of data collection for this study, the most commonly implemented models were six-month medication dispensing (6MMD) and mother-infant pair in Malawi; facility-based pickup and external pickup points in South Africa; and 6MMD and community ART access points (CAAPs) in Zambia (Table 1).
Table 1. Models of care commonly offered at study sites
Model
|
Description
|
# of expected facility visits and other provider interactions per year in DSD guidelines at time of data collection
|
Malawi
|
|
|
Conventional care
|
Clients receive a 3-month supply of medications at each full clinic visit
|
4 facility visits; 0 other interactions
|
Six-month medication dispensing (6MMD)
|
Clients receive a 6-month supply of antiretroviral medications at each full clinic visit
|
3 facility visits*; 0 other interactions
|
Mother-infant pair
|
The post-partum visits for the mother are aligned to the infant visit schedule. Infant’s schedule is based on the vaccination milestones. Thereafter the mother receives a 3-month supply of antiretroviral medications at each visit
|
4 facility visits**; 0 other interactions
|
South Africa
|
|
|
Conventional care
|
Clients receive a 2-month supply of medications at each full clinic visit
|
6 facility visits; 0 other interactions
|
Facility-based pickup points
|
Between full clinic visits, clients pick up medications (usually a 2-month supply) at specified pickup points in facilities
|
2 facility visits; 4 medication pickups
|
External pickup points
|
Between full clinic visits, clients pick up medications (usually a 2-month supply) at specified pickup points in the community (e.g. commercial pharmacy)
|
2 facility visits; 4 medication pickups
|
Zambia
|
|
|
Conventional care
|
Clients receive a 3-month supply of medications at each full clinic visit
|
4 facility visits; 0 other interactions
|
Six-month medication dispensing (6MMD)
|
Clients receive a 6-month supply of antiretroviral medications at each full clinic visit
|
3 facility visits; 0 other interactions
|
Community ART access points (CAAPs)
|
A lay worker collects 3-month supply of medication for 8 clients and distributes it at a designated CAAP
|
2 facility visits; 2 other interactions
|
*The third visit is for viral load review
** Visits are not uniform; timing varies depending on the infant’s age
Theoretical framework
Herzberg’s two-factor theory, also known as the motivator-hygiene theory, suggests that two distinct and parallel sets of factors contribute to job satisfaction (17,18). Hygiene factors are factors that, if lacking, can cause dissatisfaction among employees, but their presence does not necessarily lead to satisfaction. Examples include job security, salary, and working conditions. For clarity, in this paper, we refer to hygiene factors as maintenance factors.
Motivating factors are those that, when present, can lead to improved job satisfaction and motivate employees to perform at a higher level. Examples include achievement, recognition, responsibility, advancement, and the work itself.
Fundamentally, Herzberg’s theory suggests that satisfaction and dissatisfaction are on different continuums and influenced by different factors, with maintenance factors preventing dissatisfaction and motivating factors driving satisfaction and motivation. Herzberg’s model presumes that individuals often experience motivators and maintenance factors simultaneously, and that these factors cannot influence each other (i.e., motivators cannot increase or decrease dissatisfaction; they can only influence the degree of satisfaction). Understanding that an employee may at once be satisfied and dissatisfied is important to improving the work environment and is highly applicable to our study. Herzberg’s theory has been widely applied in the healthcare, hospitality and tourism(19), utilities, services, retail, manufacturing industries(20), among others, and within varying cultural contexts, including Jordan(21), Sweden(19,22), Saudi Arabia(23), and the United Kingdom(20). A recent systematic review has utilized Herzberg’s factors to frame findings of job satisfaction among primary healthcare providers across the world (24).
We use this theory to organize and interpret the results of this study. Throughout, we consider any response that allows providers to spend more time with patients as a Motivating Factor (work itself, time spent with patients) and time for all other tasks as a Maintenance Factor (working conditions, time for other tasks) (Table2).
Table 2. Herzberg’s Two-factor Theory and its application to our analysis
Variable
|
Definition
|
Herzberg theory factor examples (applications to our study)
|
Maintenance factors (Hygiene)
|
Necessary to not be dissatisfied, but do not necessarily lead to satisfaction
|
Job security
Working conditions (e.g. time for tasks, burn out from overwork)
Compensation
|
Motivating factors (Motivators)
|
When present, can lead to improved job satisfaction
|
Achievement Recognition (e.g. meeting DSD targets) Responsibility Advancement (e.g. training opportunities)
Work itself (e.g. intrinsic motivation from relationships with patients, time predictability)
|
Participants and data collection
At each study site, facility operations managers identified and referred to the study staff up to 10 healthcare workers who had been employed at the facility for at least 6 months and directly or indirectly involved in the implementation of ART and DSD models. Potential survey participants included facility operations managers, nurses, lay counsellors, community health workers (CHWs), pharmacists, pharmacy assistants, expert clients (individuals living with HIV who assist others in navigating HIV care) (25), and other cadres involved in DSD for HIV treatment. Study staff introduced the study, conducted the written informed consent process, and then administered the survey to the participants individually in a private area at the facility during a time that was convenient for the respondent.
Data were captured on tablets. Responses to qualitative and open-ended questions were typed verbatim into the tablet by the data collector.
Measurements
The survey instrument (supplementary file 1) included open- and closed-ended questions pertaining to providers’ experiences with DSD model implementation. The survey questions encompassed provider involvement in DSD models offered at the facility, their opinions regarding DSD models, and the challenges they faced in implementing them, as well as how DSD models affected their job responsibilities, time allocation, and job satisfaction. All participants were asked about their current experience delivering care with the DSD models. Participants who had been engaged in service delivery before DSD models were available were also asked about changes they had observed since DSD had been introduced.
To capture the effect of DSD models on job satisfaction, survey participants responded to 7 statements that were developed by the study team on the effect of DSD models on their job satisfaction using 5-point Likert scale responses with options of strongly disagree, mildly disagree, neither agree nor disagree, mildly agree and strongly agree. The statements assessed participants’ job satisfaction, joy, relationships with other colleagues and senior management, and relationships with patients since the facilities began offering DSD models.
Quantitative analysis
We first generated descriptive statistics for participants’ work characteristics, their roles and involvement in DSD models, their views on the effect of DSD models on their job responsibilities and job changes after DSD implementation. We next conducted a principal component analysis (PCA) to create an index for job satisfaction from the 7 Likert scale questions. The final scale's reliability was determined using a Cronbach's alpha coefficient of 0.60, with factors loading above 0.60 were retained for subsequent regression analyses. We categorized the final mean satisfaction scores as "high" satisfaction (score >4) or "low" satisfaction (score ≤4) in order to simplify interpretation. We then used logistic regression to examine univariate and adjusted associations between key predictor variables and low reported job satisfaction. Significance was considered at p<0.05, confidence intervals (CI) of 95%.
Qualitative analysis
For qualitative responses to open-ended questions, a codebook was developed inductively for each question by reading through at least 60% of responses. Codes were developed and refined and concepts collapsed or separated based on the content of the responses as coding proceeded. Once a codebook was finalized, each question was reread and assigned all relevant codes. Additional codes were added to the codebook if needed as they arose from the data. During analysis of each open-ended question, codes were compared by country and provider cadre (clinical vs. non-clinical); those with the highest volumes were identified as major themes in the data. Notable divergent views were identified and reviewed by the study team to reach concurrence. Results were summarized and quotes were identified within each stratum as illustrative examples. Some quotations were edited slightly for grammar and clarity. Results were triangulated with quantitative findings and interpreted within the Herzberg Two Factor Theory.