Sensitive near point-of-care detection of asymptomatic and submicroscopic Plasmodium falciparum infections in African endemic countries

Abstract

Limited diagnostic capacity for detecting asymptomatic malaria infections with low parasite densities hinders elimination efforts in Africa. Here, we adapt a near point-of-care, LAMP-based diagnostic platform for malaria diagnosis using capillary blood. This Pan/Pf detection method meets the Malaria Eradication Research Agenda (malERA) criteria for community-level screening, with a limit of detection of 0.6 parasites/μL and a sample-to-result time under 45 minutes. We evaluate its performance on 672 capillary blood samples collected at the community level in The Gambia and Burkina Faso, including 146 Plasmodium falciparum positives confirmed by qPCR. The diagnostic platform achieved 95.2% sensitivity (95% CI: 90.4-98.1) and 96.8% specificity (95% CI: 94.9-98.0). It also detected 94.9% (130/137) of asymptomatic infections and 95.3% (41/43) of submicroscopic cases (<16 parasites/μL), outperforming expert microscopy (70.1% and 0%) and rapid diagnostic tests (49.6% and 4.7%). This field-deployable molecular diagnostic method offers a sensitive, scalable solution to support test-and-treat strategies for malaria elimination across Africa.

Fig. 1. Schematic representation of our Pan/Pf malaria workflow.

The integrated system combines the SmartLid whole blood extraction process with LAMP-based isothermal amplification and colourimetric readout. The SmartLid extraction process for malaria detection from finger prick whole blood involves a four-step workflow (Lysis, Wash 1, Wash 2 and Elution) including a 5-min heat-activated enzymatic incubation, and enables DNA purification and elution in under 10 min for a single sample. A total of 20 µL of eluted DNA is transferred using a fix-volume pipette into each reaction tube, followed by a maximum of 40-min incubation at 63.5 °C. Upon completion, LAMP results are qualitatively assessed by visually evaluating the colour change within the tubes, a pink colour indicating a negative result, while yellow a positive result. The validity of the test is confirmed by verifying that all controls exhibit the expected colour changes. Created in BioRender. Cavuto, M. (2025) https://BioRender.com/rwbh4tn.

Fig. 2. Illustrated overview of SmartLid technology.

a SmartLid is composed of two main components, a disposable clear plastic lid, designed to press-fit into most 2 mL flip-cap or screw-cap tubes, and a removable magnetic key, housing a 5 mm × 5 mm N42 neodymium magnet. b Magnetic beads are collected onto SmartLid when the magnetic key is inserted, and the tube is inverted. A fluid wicking spike on the underside of the lid reduces buffer carry-over from tube to tube. c The entire magnetic beads collection, transfer, and resuspension process is illustrated, which occurs multiple times throughout the SmartLid extraction process. Created in BioRender. Cavuto, M. (2025) https://BioRender.com/rwbh4tn.

Fig. 3. Summary of SmartLid accessories to enable medium-high throughput sample processing.

a The SmartLid Rack, with numbered columns (1–12) to identify the sample and lettered rows (A–D) for each step in the extraction process. b Tube being transferred from the SmartLid Rack into the SmartLid Vortex Tool by a user performing six simultaneous extractions. c Screw-on plate locks all tubes and maintains all SmartLids in place while also providing a handle. d All samples are mixed simultaneously and equally by depressing the top of the vortex mixer with the central column underneath the tool. e All tubes are fully mixed and magnetic keys are inserted into each SmartLid. f All magnetic beads are collected simultaneously through inverting the Vortex Tool. Created in BioRender. Cavuto, M. (2025) https://BioRender.com/rwbh4tn.

Fig. 4. Comparison of analytical sensitivity of malaria detection using spiked whole blood across Dragonfly, Alethia^®, DBS-qPCR, and WB-qPCR. Experiments were conducted in collaboration with LSHTM using spiked EDTA-blood with ring-stage P. falciparum 3D7 strain.

Results are shown in terms of a number and percentage of successfully detected replicates at each spike concentration as well as b the resulting empirically determined LOD through probit analysis. The solid curves represent the predicted probabilities of positive results as a function of parasite density, with shaded areas indicating 95% CIs. The red dashed line denotes the LOD or the parasite density at which the probability of a positive result is 95%. Created in BioRender. Cavuto, M. (2025) https://BioRender.com/rwbh4tn. Source data are provided as a Source Data file.

Fig. 5. Clinical sample selection to evaluate the performance of Dragonfly Pan/Pf malaria platform.

Submicroscopic parasitaemia is defined as a parasite density of <16 parasites/μL, corresponding to the theoretical LOD for an expert microscopist, based on the ability to detect one asexual parasite among 500 leucocytes, assuming a white blood cell count of 8000 leucocytes/μL. Created in BioRender. Cavuto, M. (2025) https://BioRender.com/rwbh4tn. Source data are provided as a Source Data file.

Fig. 6. Clinical performance comparison of malaria diagnostic methods used in this study.

Comparison of the clinical performance of Dragonfly (a), LM (b) and RDTs (c) using whole blood finger prick samples, with DBS-qPCR as the gold-standard comparator. For each group, the number of true positives (TP), total positive cases, sensitivity rate with 95% CI, number of true negatives (TN), total negative cases, and specificity rate with 95% CI are provided. FP indicates the number of false positives, FN indicates the numbers of false negatives. Created in BioRender. Cavuto, M. (2025) https://BioRender.com/rwbh4tn. Source data are provided as a Source Data file.

Fig. 7. Detection of DBS-qPCR positive samples by Dragonfly, LM and RDTs, stratified by four parasite density categories: <16, 16–100, 100–200 and >200 parasites/µL.

For each category, two-sided McNemar’s tests were applied to compare paired diagnostic outcomes between Dragonfly and LM, and Dragonfly and RDTs. Dragonfly correctly identified 95.3% of submicroscopic infections, 41 out of 43 samples detected with <16 parasites/µL, significantly outperforming both expert LM (p < 0.0001) and RDTs (p < 0.0001). Dragonfly also significantly outperformed RDTs (p = 0.0001) at densities ranging from 16–100 parasites/µL, while no significant difference was observed between Dragonfly and expert LM (P = 0.0833). Similarly, for the 100–200 parasites/µL range, Dragonfly significantly outperformed RDTs (p = 0.0253), while detecting the same number of samples as expert LM. Finally, at parasite densities >200 parasites/µL, no statistically significant differences were observed between Dragonfly and either expert LM (p = 0.1573) or RDTs (p = 0.0956). Created in BioRender. Cavuto, M. (2025) https://BioRender.com/rwbh4tn. Source data are provided as a Source Data file.