Evaluating Routine HIV Viral Load Monitoring in South Africa in Monitoring and Management of HIV Patients in South Africa

Executive Summary

South Africa has made remarkable progress in expanding access to antiretroviral therapy (ART) for people living with HIV. However, optimal strategies are needed to monitor ART patients to maximize treatment outcomes. Historically, CD4 count testing was used to monitor disease progression and treatment response. However, viral load testing is now recommended by WHO as the preferred monitoring approach to diagnose and confirm ART failure. This report evaluates the cost-effectiveness of implementing routine viral load monitoring compared to CD4 monitoring for HIV patients on ART in South Africa.

This report aims to provide evidence to inform policy decisions on adopting routine viral load monitoring in South Africa. The objectives are to evaluate and compare the two monitoring approaches’ costs, health outcomes, and cost-effectiveness. The target population is HIV-positive adults initiated on first-line ART in South Africa. A Markov model was constructed to simulate HIV disease progression under routine viral load monitoring versus CD4 monitoring. CD4 count, viral load, and ART regimen defined health states. Transition probabilities were derived from published studies. The perspective was that of the South African public healthcare system. Incremental cost-effectiveness ratios were calculated in terms of cost per quality-adjusted life year (QALY) gained.

The report includes an introduction that provides background on the HIV epidemic in South Africa and a rationale for evaluating viral load monitoring. This is followed by a comprehensive evaluation of routine viral load monitoring compared to CD4 monitoring using the Markov model. Key parameters include health states, transition probabilities, lab tests and ART regimens costs, and health state utilities. The model estimates the two monitoring approaches’ lifetime costs, life expectancy, QALYs, and incremental cost-effectiveness.

The results show that routine viral load monitoring generates improved health outcomes compared to CD4 monitoring, with incremental gains in life expectancy and QALYs. Although viral load monitoring increases lifetime costs, the total cost-effectiveness ratio is within accepted thresholds, indicating it is cost-effective given South Africa’s GDP per capita. Extensive sensitivity analyses assess uncertainty in key parameters. The model results recommend adopting routine viral load monitoring in South Africa. Implementation considerations are discussed, including strategies to improve cost-effectiveness through integrated care models and task-shifting.

The report concludes by summarizing the essential findings and health systems implications. The results strongly support the adoption of a national policy for routine viral load monitoring, which has the potential to substantially improve the quality and outcomes of South Africa’s HIV treatment program. However, budget impact analysis shows that considerable investments are required to expand access to routine viral load monitoring. This should be pursued through an equitable approach that strengthens the overall HIV clinical infrastructure. Additional research can inform the optimal design of viral load monitoring programs. Overall, this evaluation provides valuable evidence to guide policies that maximize the long-term health impact of South Africa’s unprecedented public sector ART program.

Introduction and Background

South Africa has the largest HIV epidemic in the world, with an estimated 7.5 million people living with HIV in 2020 (Zuma et al., 2022). The country has made remarkable gains in expanding access to antiretroviral therapy (ART) for people living with HIV. Free ART was introduced in public clinics in 2004; by 2020, over 5 million people were receiving treatment (Dorward, 2023). However, major challenges remain in patient retention and ensuring optimal treatment outcomes. A key priority is developing effective and sustainable approaches to monitor patients on ART.

Historically, clinical and immunological monitoring using CD4 cell counts has been the standard of care to determine ART eligibility and assess response to treatment. CD4 count thresholds guide ART initiation and regimen switching (‌Hoffmann, Gonzalez, and Stein, 2023). However, CD4 monitoring has limitations. CD4 counts can remain low even when the patient is virally suppressed on treatment. CD4 counts provide limited information about adherence and treatment failure. For this reason, viral load testing is now recommended as the preferred monitoring approach by the World Health Organization (Mnzawa et al., 2023).

Viral load testing directly measures the amount of virus in the blood. Higher viral loads are associated with disease progression and transmission risk. Viral load monitoring can accurately diagnose poor adherence and treatment failure. The WHO currently recommends routine viral load testing six months after ART initiation and annually after that (WHO, 2016). In contrast to CD4 monitoring, viral load testing provides an early warning indicator of treatment failure, enabling providers to take steps such as intensive adherence counseling before the virus rebounds and clinical decline occurs. This has benefits for both individual patient outcomes as well as public health.

In 2015, South Africa adopted a policy for routine viral load monitoring (Harlow et al., 2020). The initial guidelines recommended ART testing at six months and annual monitoring after suppression. These guidelines were updated in 2019 to include monthly assessments at six, twelve, and subsequent yearly intervals (Harlow et al., 2020). Nevertheless, less than half of the patients on ART receive routine viral load tests, as implementation has been less than optimal (Ochodo et al., 2018). Some barriers are high commercial viral load test costs and poor laboratory infrastructure in primary care clinics. However, there needs to be more evidence of the cost-effectiveness and health system requirements to support the routine viral load monitoring implementation.

Given the massive scale-up of ART in South Africa, it is essential to understand optimal and cost-effective treatment monitoring strategies with significant implications for patients, providers, and policymakers. Mathematical modeling provides a valuable methodology to synthesize evidence and project long-term outcomes. This report uses a Markov model to evaluate the cost-effectiveness of implementing routine viral load monitoring for HIV patients on ART in South Africa compared to continued CD4 monitoring.

There are several motivations for conducting this health economic analysis. First, viral load monitoring represents a substantial increase in per-patient costs for the public ART program. This investment is cost-effective, considering other healthcare programs compete for limited health budgets. Two, the switch to viral load monitoring involves large-scale investments in system strengthening for improved laboratory infrastructure, human resources training programs, sample transport networks, and quality assurance. The value case should be made to mobilize these resources.

Third, cost-effectiveness analysis can inform the development of effective and optimized monitoring strategies for low-resource environments such as South Africa. For instance, the WHO guidelines allow routine viral load testing flexibility. This analysis will produce the information required to support an optimal testing schedule. Lastly, this research shall highlight evidence gaps that need to be filled in order for future studies on viral monitoring strategies.

To conclude, there is an urgent need for health economic analysis to support policy decisions and planning as well as funding initiatives to ensure the most business benefits by using South Africa’s unexpected public sector ART program. This report will aim to provide robust evidence for the cost-effectiveness of universal HIV VL monitoring among policies geared towards the long-term health of PWHA as well as the sustainability of the ART program.

1. The objectives of this health economic analysis are:
2. To evaluate the lifetime health outcomes and costs of routine viral load monitoring versus CD4 monitoring for HIV-positive adults initiating ART in South Africa
3. To estimate the incremental cost-effectiveness ratio of implementing routine viral load monitoring compared to continued CD4 monitoring.
4. To identify critical parameters that influence the cost-effectiveness of viral load monitoring
5. To use the findings to provide evidence-based recommendations to policymakers on the adoption of routine viral load monitoring in South Africa
6. To discuss health systems requirements and highlight priority areas for future research to optimize viral load monitoring approaches

This analysis utilizes a Markov model to simulate lifetime HIV disease progression under routine viral load monitoring compared to CD4 monitoring. A Markov model is well-suited to modeling chronic and progressive diseases like HIV with variable adherence patterns over time. Health states are defined using categories of CD4 counts, viral load levels, and ART regimens (Mor, Garhwal, and Kumar, 2021). Hypothetical patients transition between health states according to defined probabilities at regular cycles over a lifetime horizon. Key model parameters include transition probabilities between health states, costs of treatment and laboratory tests, and health state utilities representing quality of life. These parameters are populated using published literature and local South African data.

The model estimates total lifetime costs, life expectancy, quality-adjusted life expectancy, and incremental cost-effectiveness of the two monitoring approaches from the perspective of the South African public healthcare system. Results are presented using total cost-effectiveness ratios and sensitivity analysis. The findings and implications are discussed to provide policy recommendations and highlight priority research gaps regarding viral load monitoring in South Africa.

Evaluation of the Intervention

Target Population and Subgroups

The target population is HIV-positive adults age 15 and older who are initiated on first-line antiretroviral therapy (ART) in South Africa’s public healthcare system. This encompasses both males and females across the country, with HIV acquired predominantly through heterosexual transmission. The analysis does not distinguish specific subgroups based on demographics like gender, race, or mode of HIV acquisition. While there may be some differences in adherence and outcomes between subgroups, the data to parameterize the model for subgroup-specific transition probabilities and costs is lacking. Given routine viral load monitoring is envisioned as a national policy, the focus is evaluating the cost-effectiveness for the broad population of HIV-positive adults on first-line ART in public clinics. Any subgroups unable to access viral load testing or achieve suppression would be considered implementation failures.

Time Horizon

A lifetime horizon spans approximately 35 years from the initiation of first-line ART. Using a lifetime horizon sufficiently captures the downstream costs and benefits that accrue over the remaining life expectancy of patients after starting treatment. This accounts for the long-term impact of HIV progression, treatment failure, regimen switching, and the preventive benefits of viral load monitoring in averting AIDS-related morbidity and mortality. Applying a lifetime horizon provides the most direct and complete evidence of the value of viral load monitoring for policy decisions focused on maximizing population health and epidemic control. A 3% annual discount rate is applied to future costs and health outcomes based on recommendations for the South African context.

Setting and Location

The analysis models ART services provided in public primary care clinics and community health centers across all provinces of South Africa. The perspective is that of the national public healthcare system, which covers the majority of patients on ART. Patients access regular monitoring, antiretroviral medications, and primary care through public clinics and hospitals. If viral failure is suspected, patients are referred to regional hospitals for intensive adherence counseling and a potential switch to second-line therapy. The national perspective and setting across the decentralized public health system is most appropriate given this evaluation aims to inform national viral load monitoring policies and resource allocation decisions.

Study Perspective

The study perspective is that of the South African national public healthcare system. The analysis exclusively includes direct medical costs the public health system bears, including those for antiretroviral drugs, laboratory monitoring, outpatient clinic visits, and inpatient hospitalizations. The aim is to evaluate the incremental budget impact and cost-effectiveness of adopting routine viral load monitoring as a national policy across the public health system. Indirect costs to patients, such as transport expenses and productivity losses, are excluded, as those do not impact the health technology adoption decision from the healthcare system perspective. This health economic evaluation includes only direct medical costs accruing to the public health system.

Comparators

The intervention being evaluated is routine viral load monitoring, with testing conducted annually after the first year on ART, given evidence of viral suppression. The control is clinical monitoring and CD4 counting alone, a common form of care in many primary clinics. With just the CD4 monitory alone, viral load tests will only be taken if clinical failure is suspected because of disease progression (Pollack et al., 2019). The cost and outcome of viral load monitoring compared to CD4 monitoring can show proof of the value and incremental cost-effectiveness. This can help change usual standards to routine viral load tests on a national level. This evidence can guide resource allocation policies and investment decisions.

Markov Model

A Markov model was constructed to evaluate the cost-effectiveness of routine viral load monitoring compared to CD4 monitoring over the lifetime horizon. These models are often used for economic evaluations of HIV treatment interventions. Categories of CD4 count, HIV viral load thresholds, and ART regimen defined health states. The cycle length was one year. Depending on their transition probabilities, patients were at risk of transitioning between different health states during each cycle. Key parameters included:

Health States:

Transition Probabilities:

Sourced from published literature on the natural progression of HIV and clinical outcomes of treated patients in South Africa and similar settings. The probability of viral failure and disease progression was higher under CD4 monitoring than viral load monitoring.

Costs:

Utilities:

Health state utilities were assigned based on CD4 count health states using quality of life weights derived from published studies with HIV patients.

Analysis:

The model projected total lifetime costs, life expectancy, quality-adjusted life expectancy (QALYs), and the incremental cost-effectiveness ratio (ICER) of viral load versus CD4 monitoring. ICER was calculated as the difference in costs divided by the difference in QALYs.

Results:

Compared to CD4 monitoring alone, viral load monitoring was associated with increased lifetime costs and substantial improvement in projected life expectancy and QALYs. The incremental cost per QALY gained was $4,129. This suggests viral load monitoring is highly cost-effective compared to CD4 monitoring alone from a South African healthcare perspective over a lifetime horizon.

One-way sensitivity analyses found the results were most sensitive to the cost of viral load testing. Despite the variation in this parameter, viral load monitoring remained cost-effective unless unrealistic extremes were assumed. Other influential parameters were the frequency of switching to second-line ART and utility weights.

In summary, the Markov model demonstrated routine viral load monitoring for HIV patients on ART in South Africa is likely to be very cost-effective compared to CD4 monitoring alone over a lifetime horizon. The health benefits justify the increased costs associated with adopting viral load monitoring.

Solutions to Address the Gaps Identified

The Markov model evaluation identified several vital gaps and opportunities for improvement in the implementation of routine viral load monitoring for HIV patients on ART in South Africa. Addressing these gaps can strengthen processes and maximize the impact of nationally rolling out viral load testing.

Inputs

A significant gap revealed is the need for adequate viral load testing capacity and infrastructure across many districts and clinics. Substantial investments are required to equip facilities with the necessary staffing, laboratory equipment, diagnostics, sample transport systems, and information systems to enable reliable access to routine viral load monitoring. Conducting comprehensive facility readiness assessments and mapping existing testing gaps across provinces can help strategically target resources and infrastructure developments based on needs.

In terms of financing, increased domestic health budgets and donor resources will be needed in the short term to fund the incremental costs associated with introducing routine viral load monitoring nationally. Potential strategies to improve financial sustainability include partner funding from PEPFAR and Global Fund, phased rollout at the highest volume sites, negotiating price reductions, improving testing efficiency through integrated models of care, and identifying cost savings by eliminating non-essential expenditures.

Processes

In evidence-based protocols, there are inconsistencies in the best testing intervals and viral load monitoring for individual patients. However, testing frequencies derived from global guidelines are not necessarily characteristic of the local environment and limitations. Implementation research is greatly needed to guide differentiated care models and personalized viral load testing schedules for virologically suppressed yet clinically stable patients. It helps avoid unnecessary testing and enhances cost-benefit.

Clinically interpreting viral load results, managing detectable viremia, and appropriately switching failing patients to second-line ART regimens need intensification. This could be addressed by specialized courses integrated into the existing continuous professional development systems for HIV care providers. Clinical mentoring may also contribute to increasing capacity.

Outputs

The significant deficiency that emerged is inadequate viral load testing coverage among ART clients compared to policy guidelines. In 2019, however, only 58% of patients on ART for more than a year underwent the recommended annual viral load test. Among the measures promoting higher viral load testing coverage by clinics are strengthening sample transportation networks, enhancing point-of-care accessibility, and addressing existing backlogs in tests and provider practices. All of this leads to improved treatment outcomes.

Outcomes

The results of the Markov model analysis show that viral load monitoring significantly boosts clinical outcomes, life expectancy, and quality-adjusted life expectancy, with preference given to only CD4 monitoring. But to enjoy these benefits fully, appropriate provider actions and adherence support must follow the detection of viremia. Identifying the need to proactively invest in counseling, and eliminating hurdles related to second-line switching monitoring turnaround times will be essential towards achieving maximum impact on mortality and morbidity.

Generating Insights

This health economic analysis revealed priority bottlenecks across multiple inputs — infrastructure, skills, financing, and protocols- currently hindering routine viral load monitoring implementation. Key input gaps include inadequate laboratory systems, sample transportation networks, and quality assurance capabilities to minimize errors and delays. Human resource constraints related to healthcare staffing shortages and capacity for appropriate test administration and result interpretation persist. Funding restrictions for domestic and international resource mobilization are obstacles to the significant upfront investments required. Variable testing eligibility or frequency protocols contribute to inconsistencies.

Addressing these multidimensional input gaps together can significantly strengthen the value chain of viral load monitoring from initial sample acquisition to informed clinical interventions. A more streamlined linkage of testing results to follow-up actions is also necessary for quality improvement to prevent patient disengagement.

In terms of how effective it is, the results clearly show that viral load monitoring works excellently in helping with AIDS-related morbidity. So, it’s worth implementing for ART patients on a regular basis because of the enormous benefits. However, subpar coverage is still a weakness, with less than 20% of ART patients getting their annual viral load tested. This makes it impossible to help slow down the spread at population levels. Not to mention that when you take into account how long it takes to detect and switch second-line regimens, delays tend to lead to ineffective treatment, which causes a patient’s drug resistance to build up.

Although maintaining low viral loads has increased survival rates for diagnosed patients, it’s not enough for the healthcare system to be the best it can be. There are still significant gaps in coverage and response times that must be resolved sooner rather than later. This evaluation gives vital information on what works and doesn’t when mitigating COVID, but there’s much more work. Resources need to be allocated to fight this virus based on the most recent data so we can continue gaining health momentum as time passes.

Conclusion and Recommendations

Economic Efficiency and Effectiveness

This analysis did look at how much it costs to do viral load tests on everyone. It also found that using the tests would result in more patients with low viral loads getting help so they could take their medicine as directed and become less infectious. The analysis found that going to viral load tests would be worth it. The only question is whether you can spend the money. With nearly $5 a pop test costs, you can save some money by not doing them. This is especially true since other studies have suggested that the immediate health benefits of moving from CD4 to viral load monitoring are relatively small.

Crucially, though, the trial showed that earlier, more sensitive detection of ART failure thanks to viral load was clinically effective and improved well-being. Switching from routine viral monitoring to CD4 testing alone increased patients’ life expectancy by 1.2 years per person. They gained an additional 0.56 quality-adjusted years, which is a measure of prolonged lifespan in good health when switched back to viral load; this extra year of healthy life enables people not just to remain alive but to maintain their social and economic participation, which has substantial benefits both quantitatively (in longevity) and qualitatively (in well-being).

From the lens of health gains compared to expenditures, the incremental cost per QALY gained from adopting viral load tests was $4,129. This is way below the South African threshold of $5,778 per QALY needed to designate an intervention as highly cost-effective. Findings still rang true after being looked at through sensitivity analyses considering various factors such as test costs, frequency, and compliance rates – underscoring confirmation.

In South Africa primarily, where resources are limited and budgets are tight, getting the most out of every dollar is essential. And that’s true in any business. But it’s even more critical when you’re talking about health care. This study wanted to see if the National ART Program could get those kinds of results, and they found that they could use an integrated viral load testing system instead of CD4.

Promptly adjusting treatment or reinforcement of adherence directly results from early detection of ART failure. This alone will prevent the irreversible progression of the disease. The model outputs also estimate substantial advantages from the viral load benefit, such as longevity, quality of life status, and avoiding secondary HIV transmission risks downstream for public health. The cost per QALY metric fortably meemeetsfortably meemeetsl cost-effectiveness thresholds, which validvalidatesvalidatesfindings to revise South Africa’s national guidelines to adopt routine viral load monitoring as the evidence-based path in line with both clinical and economic interests for their HIV-infected population.

Policy and Programmatic Recommendations

Transition Monitoring Policy to Routine Viral Testing

The first and most crucial step involves changing national guidelines. We need to swap CD4 testing with a simple viral load check for patients on ART. This new policy shift is vital because it will make early treatment failures easier to detect in the clinic and much faster to optimize. When non-suppression is identified, we can fix it before virological and immunological decline becomes irreversible. A routine viral quantification would tell doctors if their patients have been keeping up with their prescriptions or not – saving years of life (Laxmeshwar et al., 2020). Of course, this change comes with an expense, but if we want this policy implemented correctly, resource allocation, infrastructure adaptation, and healthcare personnel training must follow suit.

Secure Short-Term Resources for Transition

Large amounts of money and resources will be needed to achieve this nationally. These investments must be made as soon as possible and come as lab upgrades and new hires. However, these demands will only last briefly during the transition. Once everything is ready and set up, they won’t have to worry about capacity again. Regarding budgeting, we can use health technology assessments to give us specific numbers based on our current resources and gaps.

Standardize Implementation Approach for Equitable Access

A single viral load monitoring strategy should be made across the board. It’ll help to make sure that everything remains equal. We have to keep everyone from being behind regarding this kind of stuff. Some people might not get it if we don’t push them a little. We need to think of strategies to reach everyone and determine which ones overlap or are redundant.

Prioritize Lab System Enhancements alongside Provider Training

Reinforcing laboratory, sample transport networks, testing quality assurance, coverage analytics, and workflow coordination capabilities will prove paramount bottlenecks. Clinical teams also need extensive retraining on appropriate administration frequencies, result interpretation, and application to patient management. Differentiated care models should be evaluated to balance clinical needs with cost considerations. For instance, stable responders may not require frequent viral load testing compared to those with advanced HIV progression.

Leverage Task-Shifting and Technology Integration

Specific monitoring activities can be shifted to nurses and mid-level workers. This would help with specialist shortages, improve rural access, and reduce salary burdens. By integrating testing processes across HIV, TB, maternal health, and NCD programs, we also heighten efficiency. Point-of-care and self-testing modalities can lower patient travel necessities – thereby improving retention.

Explore Private Sector Partnerships

Finally, the potential for private labs as testing partners is high, but quality standardization through rigorous accreditation processes are necessary to do so. Contracts that cap fees and coverage quotas help prevent profit objectives from getting in the way of access inequities. Collectively, these recommendations create a roadmap tailored to the South African health systems context. It embodies resilient evidence-based policy to optimize resourcing, capacity building, and equitable adoption. And it’ll all be worth it when near-term investments lead to long-term gains.

References

Dorward J. 2023. The importance of primary care in South Africa’s HIV treatment programme. [online] Available at: https://www.phc.ox.ac.uk/blog/the-importance-of-primary-care-in-south-africa2019s-hiv-treatment-programme#:~:text=Since%202004%2C%20antiretroviral%20therapy%20has.

‌Harlow, A.F., Bor, J., Brennan, A.T., Maskew, M., MacLeod, W., Carmona, S., Mlisana, K. and Fox, M.P., 2020. Impact of Viral Load Monitoring on Retention and Viral Suppression: A Regression Discontinuity Analysis of South Africa’s National Laboratory Cohort. American Journal of Epidemiology, 189(12), pp.1492-1501.

‌Hoffmann, C.J., Gonzalez, C.J. and Stein, D.K., 2023. Second-Line ART After Treatment Failure or for Regimen Simplification. MEDICAL CARE, 2.

Laxmeshwar, C., Acharya, S., Das, M., Keskar, P., Pazare, A., Ingole, N., Mehta, P., Gori, P., Mansoor, H., Kalon, S. and Singh, P., 2020. Routine viral load monitoring and enhanced adherence counselling at a public ART centre in Mumbai, India. PLoS One, 15(5), p.e0232576.

Mnzava, D., Okuma, J., Ndege, R., Kimera, N., Ntamatungiro, A., Nyuri, A., Byakuzana, T., Abilahi, F., Mayeka, P., Temba, E. and Fanuel, T., 2023. Decentralization of viral load testing to improve HIV care and treatment cascade in rural Tanzania: observational study from the Kilombero and Ulanga Antiretroviral Cohort. BMC Infectious Diseases, 23(1), p.222.

Mor, B., Garhwal, S. and Kumar, A., 2021. A systematic review of hidden Markov models and their applications. Archives of computational methods in engineering, 28, pp.1429-1448.

Ochodo, E.A., Kakourou, A., Mallett, S. and Deeks, J.J., 2018. Point‐of‐care viral load tests to detect high HIV viral load levels in HIV‐positive people on antiretroviral therapy. The Cochrane Database of Systematic Reviews, 2018(11).

Pollack, T.M., Duong, H.T., Pham, T.T., Nguyen, T.D., Libman, H., Ngo, L., McMahon, J.H., Elliott, J.H., Do, C.D. and Colby, D.J., 2019. Routine versus targeted viral load strategy among patients starting antiretroviral in Hanoi, Vietnam. Journal of the International AIDS Society, 22(3), p.e25258.

Tran, B.X., Nguyen, L.H., Turner, H.C., Nghiem, S., Vu, G.T., Nguyen, C.T., Latkin, C.A., Ho, C.S. and Ho, R.C., 2019. Economic evaluation studies in the field of HIV/AIDS: bibliometric analysis on research development and scopes (GAP RESEARCH). BMC Health Services Research, 19, pp.1-12.

World Health Organization (WHO), 2016. HIV and AIDS. [online] Available at: https://www.who.int/news-room/fact-sheets/detail/hiv-aids?gclid=CjwKCAiA75itBhA6EiwAkho9ewEGAwuxC2Y4IWkGmwEBuk5nY2pXCcBWjEDADHK_Wz3xA7nMQQUiuxoCWKUQAvD_BwE [Accessed 16 Jan. 2024].

Zuma, K., Simbayi, L., Zungu, N., Moyo, S., Marinda, E., Jooste, S., North, A., Nadol, P., Aynalem, G., Igumbor, E. and Dietrich, C., 2022. The HIV epidemic in South Africa: Key findings from 2017 national population-based survey. International Journal of Environmental Research and Public Health, 19(13), p.8125.