Cost comparison of community-based vector surveillance using VectorCamTM, an AI-enabled mosquito identification tool, versus routine entomological surveillance in Uganda

Abstract

Abstract Background National malaria elimination programs depend on timely and reliable vector surveillance data to guide interventions. Current systems rely on analog data collection and microscopy-based morphological identification of mosquitoes by entomologists (vector control officers, VCO), a process constrained by limited entomologists, high specimen volumes, and specimen transport from rural collection sites. VectorCam TM is a smartphone-based device that uses AI to classify mosquito species, sex, and abdominal status within seconds, enabling automated workflows and the task-shifting of some entomologist responsibilities to community health workers (CHWs) or village health teams (VHTs). This study evaluated the unit cost of integrating VectorCam TM into routine surveillance in Uganda. Methods Data were extracted from a 12-month pragmatic parallel randomized trial conducted across two districts in Uganda from 2023 to 2024. This analysis accounted only for costs directly associated with field activities, including personnel, transport, and supplies. Three VectorCam TM implementation models were analyzed. Implementation Model 1 represented full task-shifting of mosquito surveillance responsibilities to village health teams, where they were collected, analyzed, mosquito identified and reported data without the involvement of VCO and the use of VectorCam TM . Implementation Model 2 incorporated quarterly supervision by VCO to support village health team-led activities. Implementation Model 3 included monthly VCO supervision, aligning with routine surveillance workflows, to assess the cost implications of adding digital surveillance to existing protocols. Results The annual cost of routine surveillance (control arm) per district was USD $10,615. In comparison, VectorCam TM reduced costs to USD $8,852 in Model 1 (16.6% savings), USD $9,200 in Model 2 (13.3 savings), and USD $9896 in Model 3 (6.8% savings). These reductions were driven primarily by reductions in personnel and transport expenditures, although they were partially offset by higher supply costs associated with devices and service fees. The per-device analysis demonstrated that each VectorCam TM unit represented an annual cost of USD 144.51 but replaced USD 438.29 worth of surveillance functions, resulting in net savings of USD 293.78 per device under Implementation Model 1. Sensitivity analyses demonstrated increased savings in scenarios with existing digital infrastructure or the individual usage of VectorCam TM by one VHT. Conclusion Digitalizing vector surveillance with VectorCam TM offers a cost-saving and scalable approach for low- and middle-income countries. By shifting resources from specialized, labor-intensive methods such as microscopy to technology operable by CHW. VectorCam TM supports more efficient data collection while reducing costs. These findings suggest that integrating digital task-shifting tools into national malaria surveillance programs can enhance vector surveillance in resource-constrained settings.