Perspectives of people living with HIV and health workers about a point-of-care adherence assay: a qualitative study on acceptability

ABSTRACT

Current antiretroviral therapy (ART) adherence monitoring is premised on patients’ self-reported adherence behaviour (prone to recall error) and verified by blood viral load measurement (which can delay results). A newly developed Urine Tenofovir Rapid Assay (UTRA) assesses tenofovir in urine at point-of-care and is a novel tool to test and immediately respond to adherence levels of people living with HIV (PLHIV). We explored PLHIV and health workers’ initial perceptions about integrating the UTRA into routine medical care for adherence support. We conducted a series of once-off in-depth qualitative interviews with PLHIV (n = 25) and health workers (n = 5) at a primary care health facility in Cape Town, South Africa. Data analysis involved descriptive summaries of key emergent themes with illustrative case examples. We applied a deductive, outcomes-driven analytic approach to the summaries using the Implementation Outcomes Framework proffered by Proctor et al. (2011). The three relevant concepts from this framework that guided our evaluation were: acceptability, appropriateness, and feasibility. We found positive perceptions about the UTRA from many PLHIV and health worker participants. Many PLHIV reported that the immediate results offered by the UTRA could enable them to have constructive discussions with health workers on how to resolve adherence challenges in real-time. Few PLHIV reported concerns that drinking alcohol could affect their UTRA results. Many health workers reported that the UTRA could help them identify patients at risk of treatment failure and immediately intervene through counselling, though some relayed that they would support the UTRA’s implementation if more staff members could be added in their busy facility. Overall, these findings show that the UTRA was widely perceived to be acceptable and actionable by many PLHIV and health workers in the study.

KEYWORDS:

• ART adherence monitoring
• HIV
• PLHIV
• Point-of-care
• Tenofovir
• UTRA‌

Introduction

South Africa has the largest HIV epidemic in the world, with an estimated 7.8 million people living with HIV (PLHIV) in 2020, all requiring life-long treatment (UNAIDS, 2021). In 2016, the South African National Department of Health adopted a universal test-and-treat (UTT) approach to HIV (Meyer-Rath et al., 2017; UNAIDS, 2021). These efforts aimed to reduce new HIV infections by achieving viral load suppression among PLHIV (Meyer-Rath et al., 2017). However, South Africa’s HIV incidence rate remains high with approximately 200,000 new infections in 2019 (UNAIDS, 2020).

Optimal adherence to antiretroviral therapy (ART) is essential to both improve individual health and reduce the likelihood of onward transmission (Moosa et al., 2019). Optimal long-term adherence and viral load suppression are achievable in structured ART programmes, with close adherence monitoring (Gandhi et al., 2018; Spinelli et al., 2020). However, conventional ART adherence monitoring procedures used by HIV programmes have notable gaps (Haberer et al., 2017).

In South Africa, ART adherence monitoring is premised on patients’ self-reported adherence which is prone to recall error and social desirability bias (Huerga et al., 2017; National Department of Health, 2019). Patients’ self-reported adherence is then retroactively verified by infrequent viral load blood tests (Huerga et al., 2017). Most HIV viral load testing is centralised at specialised virology laboratories and requires substantial financial and human resources to execute and is infeasible to implement in daily routine clinical care (Gandhi et al., 2019; Phillips et al., 2019).

New, affordable, patient-centred technologies (including point-of-care viral loads, real-time urine-based tenofovir (TFV) assays, and electronic monitoring systems) are being trialled in low-income countries to help monitor and support ART adherence (Campbell et al., 2019; Drain et al., 2020; Gandhi et al., 2020; Herzer et al., 2012; Nathan Georgette & Siedner, 2017).

The Urine Tenofovir Rapid Assay (UTRA) was recently developed by our group to assess whether TFV is present in the urine and is a potential point-of-care adherence intervention to avert subsequent virologic failure (Gandhi et al., 2020). Evaluation of the effectiveness and accuracy of the UTRA is underway; this formative research assesses the acceptability, appropriateness, and feasibility of the UTRA to understand how PLHIV and health workers perceive the test. We draw on qualitative data collected during the piloting of the UTRA, to describe the perspectives of PLHIV and health workers about integrating the UTRA into routine adherence support systems.

Materials and methods

Study design and context

This qualitative evaluation is a nested sub-study in the ongoing cross-sectional study that investigates the sensitivity of the UTRA to detect virologic failure (Gandhi et al., 2020). The parent study is currently enrolling 120 patients on ART who are at risk of virologic failure (i.e., patients re-engaging in care or with a prior episode of a viral load (VL) > 400 copies/mL). The study site is a primary care health facility in Gugulethu, Cape Town, South Africa. Gugulethu (∼400,000 residents) is a high-density urban and low socio-economic neighbourhood where public sector health services are the main source of ART. More than 8,500 PLHIV were enrolled in HIV care at this facility at the time of data collection, with approximately 130 patients receiving HIV care per day (Kaplan et al., 2014 ).

Sample

We included PLHIV and health worker participants. We recruited PLHIV from the parent study and purposively sampled for diversity in age and gender until we reached saturation. Inclusion criteria were PLHIV who had adherence challenges over the past 12 months (confirmed by pharmacy refill collection data), or those with one or more prior episodes of VL > 400 copies/ml. Health workers were purposively sampled for diversity by profession and experience in delivering ART services. Among the participants approached for interviews, all were willing to take part in the study.

Data collection

A trained graduate researcher (AM) conducted a series of once-off in-depth qualitative interviews with participants between October 2020 and March 2021 at Desmond Tutu Health Foundation site offices in Gugulethu. These were at the same premises as the health facility where recruitment took place. PLHIV were recruited during their regular clinic visit check-ups, while health workers were recruited from the health facility, both by study nurses and the researcher (AM). The interview guide included open-ended questions exploring participants’ (a) treatment journeys and understanding of HIV/ART, (b) adherence challenges, monitoring tools, support needs, and (c) perceptions about the UTRA. Interviews were audio-recorded and were approximately 45–60 minutes each.

Data analysis

The researcher (AM) compiled structured case summaries for each interview from the audio recordings of discussions and field notes. Case summaries, the written narratives detailing interactions step-by-step, were refined through multiple revisions and yielded emergent themes supported by illustrative case excerpts from interviews (Rutakumwa et al., 2020). Interpretation of findings was iteratively discussed throughout the data collection process among all authors.

Conceptual framework

We applied a deductive, outcomes-driven analytic approach to the summaries using the Implementation Outcomes Framework proffered by Proctor et al. (2011). This framework describes “the effects of deliberate and purposive actions to implement new treatments, innovations, practices, and services” (Proctor et al., 2011, p. 1). The three relevant concepts from this framework that guided our evaluation were: acceptability, appropriateness, and feasibility. “Acceptability” is the perception among implementation stakeholders that a given treatment, service, practice, or innovation is agreeable, or satisfactory. “Appropriateness” refers to the perceived fit, relevance, or compatibility of the innovation to address a particular issue or problem. “Feasibility” is defined as the extent to which a new treatment, or an innovation, can be successfully used or carried out within a given agency or setting.

Ethics approval

The study received ethics approval from the University of Cape Town, Faculty of Health Science, ref – 102/2020, and reciprocal approval by the Stellenbosch University Health Research Ethics Committee. The study was also approved by the Western Cape Department of Health. All methods of this manuscript were carried out following the relevant guidelines and regulations that protect human subjects and minimise the risks of harm. All participants were informed about the study in a private room and asked to provide voluntary written informed consent before the start of each interview. Identifying information of participants was removed from the study material and participant IDs were assigned at enrolment.

Results

We interviewed a total of 25 PLHIV and 5 health workers before reaching data saturation. Of PLHIV, 10 were men and 15 were women, with ages ranging between 18 and 55 years. The health workers interviewed included two professional nurses, one enrolled nurse, and two counsellors, all of whom had been involved in delivering ART services in public health facilities. In Table 1, we show the patient participants’ IDs, sex, age, year of HIV diagnosis, year of initiating ART, their UTRA results, and self-reported adherence (i.e., frequency of taking medication and frequency of missing a dose) in the past 30 days prior to testing with the UTRA.

Table 1. Demographics of participants (PLHIV).

						Self-reported adherence in the past 30 days
Participant ID	Sex	Age	Year of HIV diagnosis	Year of initiating ART	UTRA results (TFV)	Frequency of taking medication	Frequency of missing a dose
P008	F	48	2006	2006	Present	Almost always	2
P009	F	47	2002	2016	Present	Always	0
P011	F	33	2013	2014	Present	Sometimes	1
P003	F	39	2002	2009	Present	Almost always	0
P012	M	35	2010	2010	Present	Sometimes	2
P014	F	28	2014	2015	Present	Always	1
P015	F	38	2009	2009	Present	Always	2
P016	M	22	2017	2017	Present	Always	0
P017	F	35	2014	2014	Present	Almost always	1
P018	M	28	2014	2014	Present	Always	3
P019	F	53	2011	2012	Present	Almost always	1
P020	M	31	2014	2015	Present	Almost always	3
P021	F	32	2011	2011	Present	Sometimes	3
P028	M	35	2018	2018	Present	Sometimes	0
P031	F	28	2016	2016	Present	Always	1
P032	F	37	2010	2010	Present	Almost always	0
P030	M	36	2012	2013	Present	Sometimes	0
P037	F	39	2010	2010	Present	Always	0
P036	F	41	2001	2012	Present	Almost always	0
P039	M	52	2009	2011	Present	Usually	0
P042	M	51	2013	2013	Present	Almost always	0
P048	F	50	2012	2016	Present	Sometimes	6
P043	F	49	2007	2015	Present	Almost always	0
P045	M	39	2017	2017	Present	Almost always	0
P046	M	43	2016	2016	Present	Almost always	1

The majority of PLHIV self-reported that they were either “almost always” or “always” taking ART medication in the past 30 days and did not recall missing a dose more than three times. We found that, while all our patient participants had UTRA results indicating TFV present in their urine, their self-reported adherence and experiences of UTRA varied substantially. Below, we report on the acceptability, appropriateness, and feasibility of the UTRA based on the initial reactions of health workers and PLHIV who were first-time users of this novel technology. Quotations supporting claims were selected to best illustrate the overall point and the range of experiences in the sample.

Acceptability

We found that the UTRA was perceived as acceptable by many PLHIV participants. Many PLHIV reported that they were interested in and willing to be tested with the UTRA because they felt entitled to know their adherence performance. One of our patient participants spoke about the novelty of having access to such a test as part of her ART care:

I was so excited when they recruited and told me about the test because I really wanted to know where I stand [in terms of adherence]. I was happy to be checked as long as the results come back right [showing the presence of TFV from the urine]. (P031, Female, 28-year-old)

Few PLHIV who self-reported to have missed a dose(s) in the past 30 days relayed that testing adherence with the UTRA made them feel nervous. They reported that they would only support the UTRA’s implementation if health workers offered genuine support with their adherence challenges in case(s) their urine results showed absent TFV. A patient participant who self-reported her frequency of taking medication as “sometimes” said:

I was a bit scared. I don’t care what the results say but they must be reported to me in a respectful manner that won’t discourage me, but push [motivate] me so I won’t avoid providing urine to test adherence next time. (P021, Female, 32-year-old)

All PLHIV reported that they were comfortable providing urine to test adherence, as they perceived urine sampling to be a harmless, non-intrusive procedure. A patient participant, diagnosed with HIV in 2002, told us that in her experience of testing adherence, the UTRA was the least invasive monitoring tool that she encountered thus far:

Nothing is bothering [annoying] about it. There are no itching needles involved, you just pee. That’s something we do all the time. (P009, Female, 47-year-old)

The majority of health workers also perceived the UTRA to be acceptable and relayed that it is less challenging to administer compared to other available modalities. One health worker reported that her work mostly involves writing ART prescriptions for patients, but she would willingly run the test as part of her routine responsibilities:

In terms of administering this UTRA test. I don’t see why I cannot perform the test. It’s not something that looks exhausting or get to a point where I tell myself I won’t go to work because of it. (30-year-old, enrolled nurse)

Among our participants, many PLHIV and health workers relayed that they found the test satisfactory for use and conveyed that the UTRA is more user-friendly than other conventional adherence monitoring modalities.

Appropriateness

The UTRA was considered a suitable tool for improving adherence by many PLHIV and health workers. Some PLHIV reported that the UTRA would encourage them to take treatment more consistently because it is an opportunity to demonstrate good adherence on their next clinic visit. A patient participant told us that being tested with the UTRA at every clinic visit would work as a constant reminder that he needed to adhere to treatment, despite facing practical challenges such as travelling with work:

The test will push [motivate] me because if I did not take my meds now, when I come to the clinic, they will see that I didn’t drink my tablets. So, when I go home, I will tell myself ‘I don’t want to go back there to get two lines [no TFV drug present] and be embarrassed’. So, I must drink [take] my pills. (P045, Male, 39-year-old)

Many PLHIV indicated that, with the immediate results offered by the UTRA, they would be able to have constructive discussions with health workers on how to “deal” with adherence challenges in real-time. For instance, a patient participant said:

I want my results now, as we speak. It’s better sooner than later and, with this urine test, I’d know my poor results now and be advised what to do, than finding out later when the damage is already done. (P042, Male, 51-year-old)

In rare cases, some PLHIV reported that although they take medication frequently, they also drink alcohol regularly and were concerned that the UTRA might not detect the TFV from their urine for this reason. A patient participant reported that he is worried that the UTRA could be ineffective because it might report inaccurate results when his urine is “diluted” with alcohol:

What if they also see that I drank alcohol? Maybe the test will not show that I took medication because there’s also alcohol in my urine. (P012, Male, 35-year-old)

The majority of health workers reported that the real-time results offered by the UTRA could inform and enable them to intervene and, through counselling, support PLHIV at risk of treatment failure. For instance, a health worker shared that many patients struggle to accurately recall their adherence which is a major challenge in improving adherence:

I think it’s a good thing because you can monitor someone’s adherence on the spot and act now. You don’t have to send blood samples away and wait. (54-year-old, professional nurse)

The majority of health workers in this evaluation reported that integrating the UTRA into routine HIV care could potentially minimise the workload at the clinic. These health workers reported that introducing the UTRA into standard care could help them identify patients at risk of treatment failure and utilise conventional adherence measures more reservedly:

This might mean I won’t get tired of drawing blood today and again next month in cases where the blood lags at the lab, because we will only test patients needing that [blood tests]. The urine test will help me to know who needs blood tests urgently. (28-year-old, professional nurse)

The responses by our participants convey that the UTRA was perceived as an effective tool in that it could minimise the workload for health workers and encourage PLHIV to consistently adhere to treatment in a timely fashion.

Feasibility

The majority of health workers reported that they did not anticipate any challenges to the countrywide implementation of the UTRA into routine HIV services. These health workers reported that they felt confident and capable of administering the test:

We can easily do this. This is small water [a minor task]. We are experienced in doing pregnancy and HIV tests, so at least this one looks like a familiar thing. (49-year-old, counsellor)

There were also a few concerns reported by some health workers who relayed that the practicality of implementing the UTRA could only be possible if the health sector could provide more staff members to ease the workload in this busy clinic:

There are many patients we see in this clinic a day and we need to think of that. If we are to use this device among other things, we’d need more people to help with it, maybe these non-profit organisations. (30-year-old, enrolled nurse)

However, many health workers further reported that integrating the UTRA into standard HIV care appears feasible because it seems cheaper compared to other expensive, laborious adherence monitoring modalities:

You just ask for someone’s urine, you put a drop of it there, right? And the results show immediately. I think this could be cheaper than other tests. (54-year-old, professional nurse)

Many PLHIV shared similar views on the feasibility of implementing the UTRA into health services. They reported that the implementation of the UTRA could be successful if health workers could run the test while patients are waiting to be seen for other health services:

If the test doesn’t take long to do, the nurses will still be able to test patients while we are waiting for our folders or pills. (P019, Female, 35-year-old)

The majority of participants included in our evaluation expressed support for the feasibility of the UTRA both in the delivery and receipt of HIV care, perceiving it to be time effective.

Discussion

We found that the majority of participants – both PLHIV and health workers – perceived our novel urine assay as acceptable and actionable, though some health workers reported that its implementation would require additional staff members in their busy clinic. Nevertheless, the majority of PLHIV and health workers showed strong support for its implementation and relayed that the UTRA could strengthen the adherence of patients through timely identification of patients at risk of virologic failure and allow early intervention through adherence counselling.

Strengths of our study include that our findings contribute to the breadth of knowledge on objective ART adherence monitoring by offering insights on the acceptability, appropriateness, and feasibility of a urine-based point-of-care testing device. The qualitative interviews also pointed out potential uses of the assay, such as triaging patients for further laboratory testing.

Our study has some limitations. Although the interviewer directly noted responses and included open-ended questions, the interpretation of findings could have been influenced by their perceptions despite iterative discussion with other researchers. Another limitation is the inability to extrapolate our findings to other settings since we sampled participants from a single healthcare facility. Therefore, readers should transfer findings from this study cautiously, especially if the local social and health system contexts differ substantively. To mitigate this, we sampled participants for diversity in age and gender, and explored the assay with experienced health workers, allowing us to explore a diverse range of perspectives.

This manuscript reported the first-ever qualitative data on this novel urine assay as a point-of-care tool to evaluate adherence among PLHIV receiving ART as treatment. Tests to monitor drug levels in plasma and dried blood spots have been used in treatment studies, but do not have the advantage of urine-based tests, which are less invasive and provide point-of-care results (Castillo-Mancilla et al., 2013; Phillips et al., 2019). Recent studies found that the real-time urine-based TFV test predicts pre-exposure prophylaxis (PrEP) outcomes and is acceptable and feasible to implement in PrEP programmes (Bardon et al., 2020; Drain et al., 2020; Spinelli et al., 2019). In a qualitative study conducted in the United States among PrEP users, people were at first hesitant about rapid TFV concentration monitoring but were later generally supportive of its implementation and emphasized the potential value of self-administering testing (Bardon et al., 2020). We did not see this pattern of initial hesitation among the majority of PLHIV in the more generalised epidemic setting in South Africa.

Another qualitative study conducted among PrEP users in the United States demonstrated that the rapid TFV testing was highly acceptable, but the participants raised concerns about being tested for illicit drugs and “white coat adherence” (an increase in adherence prior to a clinic visit) (Hunt et al., 2019). Similarly, some PLHIV in our evaluation reported that they drink alcohol regularly and anticipated that it might affect their UTRA results. Nevertheless, many of our PLHIV participants perceived the rapid adherence test as an opportunity to consistently improve their adherence, not just prior to clinic visits. Moreover, the majority of health workers in the study perceived the urine-based test as easy to perform and saw its potential to offer real-time results of patients’ adherence, which could enable them to enact immediate interventions for patients at risk of virologic failure.

Conclusion

Overall, our novel urine-based point-of-care test to monitor tenofovir objectively as an adherence metric was found to be acceptable, appropriate, and feasible in this busy clinic in South Africa, where HIV is a generalised high prevalence epidemic. Future research assessing the acceptability of a monitoring device could include a larger sample of participants from different contexts including those who tested negative with the device for wider transferability of results. However, this qualitative study does reveal the promise of a real-time adherence metric to both monitor and support adherence, and a clinical trial to assess the utility of the UTRA in improving virologic outcomes is underway.

Acknowledgements

The authors thank all the participants for sharing their insightful perspectives and the field team of study nurses for assisting with recruitment.

Data availability statement

The datasets generated and analysed during the current study are not publicly available due to the need to protect participants’ confidentiality but are available from the corresponding author upon reasonable request.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

Research reported in this manuscript was supported by the U.S National Insitutes of Health (NIH) and South African Medical Research Council through its U.S.-SA Program for Collaborative Biomedical Research, Grant R01AI152119 (PIs van Zyl, Orrell, Gandhi). The NIH further supported the development of the urine-based TFV test through R01AI143340 (PI, Gandhi).