Diagnostic network optimization (DNO), a geospatial optimization technique, can improve access to diagnostics and reduce costs through informing policy-makers' decisions on diagnostic network changes. In Zambia, viral load (VL) testing and early infant diagnosis (EID) for HIV has been performed at centralized laboratories, whilst the TB-programme utilizes a decentralized network of GeneXpert platforms. Recently, the World Health Organization (WHO) has recommended point-of-care (POC) EID/VL to increase timely diagnosis. This analysis modelled the impact of integrating EID/VL testing for children and pregnant/breastfeeding-women (priority-HIV) with TB on GeneXpert in Zambia. Using OptiDx, we established the baseline diagnostic network using inputs for testing demand (October 2019-September 2020), referrals, testing sites, testing platforms, and costs for HIV/TB testing (transport, test, device) respectively in Zambia. Next, we integrated priority-HIV testing on GeneXpert platforms, historically only utilized by the TB-programme. 228,265 TB tests were conducted on GeneXpert devices and 167,458 (99%) of priority-HIV tests on centralized devices at baseline, of which 10% were tested onsite at the site of sample collection. With integration, the average distance travelled by priority-HIV tests decreased 10-fold (98km to 10km) and the proportion tested onsite increased (10% to 48%). 52% of EID tests are likely to be processed within the same-day from a baseline of zero. There were also benefits to the TB-programme: the average distance travelled/specimen decreased (11km to 7km), alongside potential savings in GeneXpert device-operating costs (30%) through cost-sharing with the HIV-programme. The total cost of the combined testing programmes reduced marginally by 1% through integration/optimization. DNO can be used to strategically leverage existing capacity to achieve the WHO's recommendation regarding POC VL/EID testing. Through DNO of the Zambian network, we have shown that TB/HIV testing integration can improve the performance of the diagnostic network and increase the proportion of specimens tested closer to the patient whilst not increasing costs.
Copyright: © 2023 Girdwood et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.