Optimising courier specimen collection time improves patient access to HIV viral load testing in South Africa.

Background: South Africa uses a courier network for transporting specimens to public laboratories. After the daily collection of specimens from the facility by the courier, patients not yet attended to are unlikely to receive same-day blood draws, potentially inhibiting access to viral load (VL) testing for HIV patients. Objective: We aimed to design an optimised courier network and assess whether this improves VL testing access. Methods: We optimised the specimen transport network in South Africa for 4046 facilities (November 2019). For facilities with current specimen transport times (n = 356), we assessed the relationship between specimen transport time and VL testing access (number of annual VL tests per antiretroviral treatment patient) using regression analysis. We compared our optimised transport times with courier collection times to determine the change in access to same-day blood draws. Results: The number of annual VL tests per antiretroviral treatment patient (1.14, standard deviation: 0.02) was higher at facilities that had courier collection after 13:36 (the average latest collection time) than those that had their last collection before 13:36 (1.06, standard deviation: 0.03), even when adjusted for facility size. Through network optimisation, the average time for specimen transport was delayed to 14:35, resulting in a 6% - 13% increase in patient access to blood draws. Conclusion: Viral load testing access depends on the time of courier collection at healthcare facilities. Simple solutions are frequently overlooked in the quest to improve healthcare. We demonstrate how simply changing specimen transportation timing could markedly improve access to VL testing.

Monitoring viral load for the last mile: what will it cost?

INTRODUCTION: Routine viral load testing is the WHO-recommended method for monitoring HIV-infected patients on ART, and many countries are rapidly scaling up testing capacity at centralized laboratories. Providing testing access to the most remote populations and facilities (the "last mile") is especially challenging. Using a geospatial optimization model, we estimated the incremental costs of accessing the most remote 20% of patients in Zambia by expanding the transportation network required to bring blood samples from ART clinics to centralized laboratories and return results to clinics. METHODS: The model first optimized a sample transportation network (STN) that can transport 80% of anticipated sample volumes to centralized viral load testing laboratories on a daily or weekly basis, in line with Zambia's 2020 targets. Data incorporated into the model included the location and infrastructure of all health facilities providing ART, location of laboratories, measured distances and drive times between the two, expected future viral load demand by health facility, and local cost estimates. We then continued to expand the modelled STN in 5% increments until 100% of all samples could be collected. RESULTS AND DISCUSSION: The cost per viral load test when reaching 80% patient volumes using centralized viral load testing was a median of $18.99. With an expanded STN, the incremental cost per test rose to $20.29 for 80% to 85% and $20.52 for 85% to 90%. Above 90% coverage, the incremental cost per test increased substantially to $31.57 for 90% to 95% and $51.95 for 95% to 100%. The high numbers of kilometres driven per sample transported and large number of vehicles needed increase costs dramatically for reaching the clinics that serve the last 5% of patients. CONCLUSIONS: Providing sample transport services to the most remote clinics in low- and middle-income countries is likely to be cost-prohibitive. Other strategies are needed to reduce the cost and increase the feasibility of making viral load monitoring available to the last 10% of patients. The cost of alternative methods, such as optimal point-of-care viral load equipment placement and usage, dried blood/plasma spot specimen utilization, or use of drones in geographically remote facilities, should be evaluated.

Optimizing viral load testing access for the last mile: Geospatial cost model for point of care instrument placement.

INTRODUCTION: Viral load (VL) monitoring programs have been scaled up rapidly, but are now facing the challenge of providing access to the most remote facilities (the "last mile"). For the hardest-to-reach facilities in Zambia, we compared the cost of placing point of care (POC) viral load instruments at or near facilities to the cost of an expanded sample transportation network (STN) to deliver samples to centralized laboratories. METHODS: We extended a previously described geospatial model for Zambia that first optimized a STN for centralized laboratories for 90% of estimated viral load volumes. Amongst the remaining 10% of volumes, facilities were identified as candidates for POC placement, and then instrument placement was optimized such that access and instrument utilization is maximized. We evaluated the full cost per test under three scenarios: 1) POC placement at all facilities identified for POC; 2)an optimized combination of both on-site POC placement and placement at facilities acting as POC hubs; and 3) integration into the centralized STN to allow use of centralized laboratories. RESULTS: For the hardest-to-reach facilities, optimal POC placement covered a quarter of HIV-treating facilities. Scenario 2 resulted in a cost per test of $39.58, 6% less than the cost per test of scenario 1, $41.81. This is due to increased POC instrument utilization in scenario 2 where facilities can act as POC hubs. Scenario 3 was the most costly at $53.40 per test, due to high transport costs under the centralized model ($36 per test compared to $12 per test in scenario 2). CONCLUSIONS: POC VL testing may reduce the costs of expanding access to the hardest-to-reach populations, despite the cost of equipment and low patient volumes. An optimal combination of both on-site placement and the use of POC hubs can reduce the cost per test by 6-35% by reducing transport costs and increasing instrument utilization.

Impact of a borderless sample transport network for scaling up viral load monitoring: results of a geospatial optimization model for Zambia.

INTRODUCTION: The World Health Organization recommends viral load (VL) monitoring at six and twelve months and then annually after initiating antiretroviral treatment for HIV. In many African countries, expansion of VL testing has been slow due to a lack of efficient blood sample transportation networks (STN). To assist Zambia in scaling up testing capacity, we modelled an optimal STN to minimize the cost of a national VL STN. METHODS: The model optimizes a STN in Zambia for the anticipated 1.5 million VL tests that will be needed in 2020, taking into account geography, district political boundaries, and road, laboratory and facility infrastructure. We evaluated all-inclusive STN costs of two alternative scenarios: (1) optimized status quo: each district provides its own weekly or daily sample transport; and (2) optimized borderless STN: ignores district boundaries, provides weekly or daily sample transport, and reaches all Scenario 1 facilities. RESULTS: Under both scenarios, VL testing coverage would increase to from 10% in 2016 to 91% in 2020. The mean transport cost per VL in Scenario 2 was $2.11 per test (SD $0.28), 52% less than the mean cost/test in Scenario 1, $4.37 (SD $0.69), comprising 10% and 19% of the cost of a VL respectively. CONCLUSIONS: An efficient STN that optimizes sample transport on the basis of geography and test volume, rather than political boundaries, can cut the cost of sample transport by more than half, providing a cost savings opportunity for countries that face significant resource constraints.