High-Sensitivity Near Point-of-Care Detection of Asymptomatic and Sub-microscopic Plasmodium Infections in African Endemic Countries

20 The limited diagnostic capacity to detect asymptomatic individuals with low parasite densities continues to 21 hinder malaria elimination efforts across Africa . We adapted a near point -of-care LAMP-based diagnostic 22 platform, originally designed for viral detection in respiratory and skin lesion specimens , for malaria diagnosis 23 using capillary blood. The resulting Pan/Pf malaria test meets the Malaria Eradication Research Agenda (malERA) 24 essential criteria for community-level malaria screening, with an analytical limit-of-detection of 0.6 parasites/μL 25 and a sample -to-result turnaround time under 45 minutes . We evaluated the test using 672 capillary blood 26 samples obtained via finger pricks from individuals enrolled at the community level in The Gambia and Burkina 27 Faso, including 146 positives for P. falciparum confirmed by dried blood spot qPCR. The test achieved a sensitivity 28 of 95.2% [95% CI: 90.4-98.1] and specificity of 96.8% [95% CI: 94.9-98.0]. It also detected 94.9% (130/137) of 29 asymptomatic malaria infections and 95.3% (41/43) of sub-microscopic cases (<16 parasites/µL), outperforming 30 expert light microscopy, which detected 70.1% (96/137) and 0% (0/43), and rapid diagnostic tests, which 31 detected 49.6% (68/137) and 4.7% (2/43), respectively. This field molecular method represents a sensitive and 32 scalable diagnostic solution with the potential to support test-and-treat strategies for malaria elimination across 33 Africa. 34

35 Despite substantial investment , malaria remains a significant global health concern, with an estimated 263 36 million cases and 597,000 deaths reported in 2023.1 The African continent bears the highest burden, accounting 37 for 94% of malaria cases and 95% of deaths , predominately caused by Plasmodium falciparum.1 While high 38 transmission settings rely on universal coverage of standard interventions such as insecticide-treated nets (ITNs), 39 indoor residual spraying (IRS), intermittent preventive treatment for pregnant women (IPTp), and seasonal 40 malaria chemoprevention (SMC), regions with declining transmission urgently require innovative strategies to 41 accelerate malaria elimination.2–4 42 To accelerate progress, it is crucial to target the entire human reservoir of infection, including both symptomatic 43 and asymptomatic Plasmodium-infected individuals.5 Passive case detection at health facilities identifies clinical 44 malaria cases while asymptomatic carriers in the community are unlikely to seek medical care and remain 45 undetected. In contrast, a ctive detection interventions ( ADIs), involving community -level test-and-treat 46 strategies, offer a promising approach to addressing this gap.5,6 ADIs also have the advantage of avoiding the 47 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 14, 2025. ; https://doi.org/10.1101/2025.07.12.25331425doi: medRxiv preprint NOTE: This preprint reports new research that has not been certified by peer review and should not be used to guide clinical practice. 2 widespread drug exposure in Mass Drug Administration (MDA), a strategy that treats entire populations 48 regardless of the infection status. MDA can lead to several challenges, including the risk of drug resistance, the 49 unnecessary treatment of uninfected individuals, and the logistica l burden of implementing large -scale 50 treatments.7,8 However, the success of ADIs hinges on the availability of highly sensitive diagnostic tools capable 51 of detecting all or most malaria infections.2,8,9 52 Asymptomatic carriage is often associated with low parasite densities , typically below 100 parasite/ µL, and 53 frequently fall under the detection threshold of conventional diagnostic methods such as light microscopy (LM) 54 and rapid diagnostic tests (RDTs).5,10,11 These infections, termed sub-microscopic or sub-patent, occur across the 55 whole spectrum of malaria transmission intensity with the highest proportion (60-70%) among infected 56 individuals in low-transmission settings.12 Despite their low density, sub -microscopic infections contribute to 57 ongoing transmission by harbouring gametocytes that can infect mosquitoes .13,14 RDTs and LM have a limit of 58 detection (LOD) around 100-200 and 50-100 parasites/µl, respectively.15 Both tests face challenges in reliably 59 identifying asymptomatic carriers, particularly those with low parasitaemia. LM could be subjective depending 60 on the skills of the microscopist, whereas the increased number of reported deletions of PfHRP2 and PfHRP3 61 genes are compromising the use of HRP -based RDTs .1 In contrast , Polymerase Chain Reaction (PCR) and 62 quantitative PCR (qPCR) detect infections at densities as low as 0.0 02 parasites/µl, but their widespread 63 adoption faces significant challenges due to the complexity of execution, particularly in resource-constrained 64 settings.16,17 PCR-based techniques requires labo ur-intensive procedures, substantial costs, advanced 65 laboratory infrastructure, and highly skilled technicians. Additionally, results can take several hours or even days 66 to produce, significantly delaying diagnosis and treatment.17 Therefore, there is a pressing need for a diagnostic 67 technology that combines both the high analytical sensitivity of molecular methods with the practicality and 68 accessibility required for widespread use in resource -constrained field settings currently achieved by RDTs. As 69 alternatives to standard PCR, several Nucleic Acid Amplification Tests (NAATs) have emerged, including Nucleic 70 Acid Sequence-Based Amplification (NASBA), Recombinase Polymerase Amplification (RPA) and Loop-mediated 71 isothermal amplification (LAMP).18–21 72 LAMP-based technologies offer a promising alternative, combining the high sensitivity of molecular diagnostics 73 with simpler equipment and operational requirements.19,22 Unlike PCR, LAMP allows the amplification of target 74 nucleic acid sequences at a constant temperature, an advantageous feature for field deployment, enabling the 75 use of less expensive and more portable battery-powered block heaters . Unfortunately, similar to PCR , to 76 achieve sufficient sensitivity LAMP requires high quality nucleic acid extraction to be performed prior to 77 amplification.19 As a result, many LAMP assays still rely on lengthy, multi-step nucleic acid extraction kits, making 78 them less practical for community -based screening that demands high throughput and a shorter time -to-79 result.23–25 Moreover, liquid LAMP reagents, as for other molecular approaches, require cold-chain storage which 80 represents a challenge for deployment in resource-constrained settings. To our knowledge, only two LAMP-81 based malaria diagnostic platforms are commercially available , the LoopampTM Malaria Detection Kits (Eiken 82 Chemical Co., Tokyo, Japan) 26 and the Alethia ® Malaria ( Meridian Bioscience Inc., Cincinnati, OH, USA) , 83 previously called Illumigene®27. While both platforms eliminate the need for a cold chain through lyophilisation 84 of the LAMP reagents, they also employ shortened sample preparation processes which may increase reaction 85 inhibition and lower nucleic acid recovery. Both of them also rely on instruments that measure turbidimetry or 86 fluorescence emission for result readout , increasing the cost and bulk of each solution .25 In recent years, the 87 Alethia® Malaria assay has been widely deployed in non-endemic high-income countries for the diagnosis of 88 malaria in returning travellers , due to its high diagnostic accuracy relative to RDTs .28–30 However, while the 89 Alethia® platform maintains relative ease -of-use when compared to standard laboratory-based molecular 90 methods, the cost of the instrument (> $20,000) and limited throughput (< 10 samples per instrument every 40 91 minutes) remain significant barriers to its wider adoption and remote deployment in rural African settings. 92 To address the challenges outlined above, we present a novel near point-of-care (POC) molecular approach for 93 detecting Plasmodium genes. This solution, combining the magnetic bead -based nucleic acid extraction 94 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 14, 2025. ; https://doi.org/10.1101/2025.07.12.25331425doi: medRxiv preprint 3 technology from SmartLid and the lyophilised colo urimetric LAMP chemistry from the Dragonfly platform 95 (originally developed for detecting viral respiratory and skin infections)31,32, was optimised for medium to high-96 throughput malaria testing using capillary blood samples obtained via finger pricks in resource-limited settings. 97 First, the analytical performance of this method was compared against the Alethia® Malaria Test, dried blood 98 spot (DBS)-qPCR, and whole blood qPCR (WB-qPCR) using Plasmodium culture spiked into whole blood. Next, its 99 clinical performance was evaluated in standard laborator ies at the Medical Research Council (MRC) Unit The 100 Gambia at LSHTM and at The Clinical Research Unit of Nanoro . This was achieved by collecting capillary blood 101 samples from individuals enrolled in a community-based survey in rural Burkina Faso and The Gambia, 102 benchmarking its accuracy against HRP2 -based RDTs, LM, and DBS-qPCR. A total of 672 whole blood samples 103 were collected from 646 asymptomatic and 26 symptomatic individuals. 104

105 Malaria detection workflow overview 106 An overview of the adapted Dragonfly workflow is illustrated in Fig. 1 and the entire standard operating 107 procedure detailed in Supplementary Methods. On the front-end, an extraction method based on silica-coated 108 superparamagnetic beads (TurboBeads, ProtonDx) and SmartLid technology was optimised to extract parasite 109 DNA simultaneously from up to 12 whole blood samples in under 1 5 minutes, without us ing a centrifuge. 110 Lyophilised colourimetric LAMP chemistry was then used for the rapid isothermal amplification of both pan -111 Plasmodium species and Plasmodium falciparum targets in a single reaction well , requiring only a simple low -112 cost (~£100), and portable (160 x 110 x 130 mm, <1 kg), dry-bath heat block.31,33,34 Finally, results readout and 113 interpretation were accomplished entirely visually, with a distinct colour change from pink (negative) to yellow 114 (positive), avoiding expensive and bulky instrumentation required for fluorescent detection. Altogether, the 115 entire sample-to-result workflow from EDTA-anticoagulated capillary blood was accomplished for up to 12 116 samples within 45 minutes by a single user. 117 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 14, 2025. ; https://doi.org/10.1101/2025.07.12.25331425doi: medRxiv preprint 4 118 Figure 1. Schematic representation of our Pan/ Pf malaria test workflow. The integrated system combines the 119 SmartLid whole blood extraction process with LAMP-based isothermal amplification and colourimetric readout. 120 The SmartLid extraction process for malaria detect ion from finger prick whole blood involves a four -step 121 workflow (Lysis, Wash 1, Wash 2, and Elution) including a 5 -minute heat-activated enzymatic incubation, and 122 enables DNA purification and elution in under 10 minutes for a single sample. A total of 20 µL of eluted DNA is 123 transferred using a fix-volume pipette into each reaction tube, followed by a maximum of 40-minute incubation 124 at 63.5°C. Upon completion, LAMP results are qualitatively assessed by visually evaluating the colour change 125 within the tubes, a pink colour indicating a negative result, while yellow a positive result. The validity of the test 126 is confirmed by verifying that all control s exhibit the expected colour changes , as described below . Created in 127 BioRender. Cavuto, M. (2025) https://BioRender.com/rwbh4tn. 128 SmartLid blood DNA/RNA extraction kit 129 SmartLid-based nucleic acid extraction technology leverages a disposable lid with a removable magnetic key to 130 quickly and easily transfer magnetic beads and attached nucleic acids through multiple buffers and steps in the 131 extraction and purification process. After binding nucleic acids to the silica -coated magnetic beads (Fig. 2a), 132 collection onto the lid is performed through multiple inversions of the tube with the magnetic key inserted (Fig. 133 2b). The SmartLid, along with the collected magnetic beads, can then be removed from the tube and transferred 134 into the subsequent tube. Release of the magnetic beads into the new buffer is accomplished by removing the 135 magnetic key and briefly shaking the tube. This entire process, transferring magnetic beads from one tube to 136 another, is illustrated in Fig. 2c. Note that while the clear plastic (polypropylene) lid component of SmartLid is 137 disposable, the green magnetic key is reusable, cutting down on plastic and rare-earth waste. 138 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 14, 2025. ; https://doi.org/10.1101/2025.07.12.25331425doi: medRxiv preprint 5 139 Figure 2. Illustrated overview of SmartLid technology. a) SmartLid is composed of two main components, a 140 disposable clear plastic lid , designed to press-fit into most 2 mL flip -cap or screw -cap tubes, and a removable 141 magnetic key, housing a 4 mm x 4 mm N42 neodymium magnet. b) Magnetic beads are collected onto SmartLid 142 when the magnetic key is inserted, and the tube is inverted . A fluid wicking spike on the underside of the lid 143 reduces buffer carry-over from tube to tube. c) The entire magnetic beads collection, transfer, and resuspension 144 process is illustrated, which occurs multiple times throughout the SmartLid extraction process. Created in 145 BioRender. Cavuto, M. (2025) https://BioRender.com/rwbh4tn. 146 The originally developed SmartLid protocol to extract viral DNA and RNA from swabs stored in guanidium 147 thiocyanate-based buffer, eNAT transport/inactivation media (Copan, Italy), was adapted for this study to 148 extract human and Plasmodium genomic DNA from 100 µL of EDTA-anticoagulated whole blood. Notably, a 5 -149 minute heat-activated (65°C) enzymatic (proteinase K) lysis step was added to help break down the protein-rich 150 sample matrix, along with an additional wash step to reduce contaminant carry -over into th e elution. Finally, 151 vortex mixing was utilised instead of manual shaking to agitate the magnetic beads in each extraction buffer, 152 which is encouraged due to the higher viscosity and more inhabitant -rich sample matrix when compared to 153 typical respiratory or skin swab eluent. For a single sample, the entire extraction process can be completed in 154 approximately 10 minutes. All kits used for the study (SmartLid Blood DNA/RNA Extraction Kit) were custom 155 produced in collaboration with ProtonDx Ltd (https://www.protondx.com/). 156 SmartLid adaptation for medium-high throughput sample processing 157 The original SmartLid extraction method was developed for use at the POC and assumed processing only a single 158 sample at a time with single use cardboard trays utili sed both as the kit component packaging and as a 159 workstation. For this study, as shown in Fig. 3, the SmartLid method was adapted for medium-high throughput 160 sample processing by developing two key tools which together enabled the simultaneous processing of up to 12 161 samples at a time by a single user. First, a 3D printed (X1 Carbon with AMS, Bambu Labs) tube rack was created, 162 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 14, 2025. ; https://doi.org/10.1101/2025.07.12.25331425doi: medRxiv preprint 6 with 12 columns (one for each sample) of four rows (one for each step in the sample extraction process), labelled 163 1-12 and A-D for the columns and rows, respectively (Fig. 3a). The spacing of each column and row was optimised 164 to enable easy transfer of SmartLids from tube to tube within a column without obstruction from the open flip-165 cap lids. Users were also encouraged to write the column number on top of each SmartLid to further reduce the 166 likelihood of accidentally mixing up two samples. Next, a multi-tube vortex tool was developed to conveniently 167 hold up to 12 sample tubes (with attached SmartLids) , enabling simultaneous mixing (Fig. 3b-f). The central 168 column on the underside of the vortex tool is depressed into the centre of the vortex mixer, while a screw -on 169 lid with a handle clamp down on the top of each SmartLid. Finally, each tube location in the tool is numbered to 170 allow easy correlation with the number on each SmartLid and/or column number in the tube rack. Combined, 171 these two new tools enabled 12 whole blood samples to be extracted in parallel in under 15 minutes by a single 172 user. 173 174 Figure 3. Summary of SmartLid accessories to enable medium -high throughput sample processing. a) The 175 SmartLid Rack, with numbered columns (1-12) to identify the sample and lettered rows (A-D) for each step in 176 the extraction process. b) A tube being transferred from the SmartLid Rack into the SmartLid Vortex Tool by a 177 user performing six simultaneous extractions. c) A screw-on plate locks all tubes and maintains all SmartLids in 178 place while also providing a handle. d) All samples are mixed simultaneously and equally by depressing the top 179 of the vortex mixer with the central under neath column of the tool. e) All tubes are fully mixed and magnetic 180 keys are inserted into each SmartLid. f) All magnetic beads are collected simultaneously through inverting the 181 Vortex Tool. Created in BioRender. Cavuto, M. (2025) https://BioRender.com/rwbh4tn. 182 Multi-patient malaria Pan/Pf test panel design 183 The Pan/Pf malaria test panel presented here was adapted from the single-patient, multi-pathogen format to a 184 format designed to test multiple patients for a single pathogen in order to increase throughput and reduce cost 185 per patient.32 Each flip-cap tube within the 8-tube strip panel contained lyophilized colourimetric LAMP reagents 186 which are stable at room temperature for extended periods and provide a clear visual colour change from pink 187 to yellow to indicate a positive result. The tandem Pan/ Pf assay was designed to target both Pan -malaria DNA 188 sequences, conserved across multiple Plasmodium species responsible for malaria, and Plasmodium falciparum-189 specific sequences. A comprehensive list of the p reviously published primer sequences used in this study is 190 provided in Supplementary Table 1.20,35 191 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 14, 2025. ; https://doi.org/10.1101/2025.07.12.25331425doi: medRxiv preprint 7 In addition to six Pan/ Pf target reactions, allowing for the simultaneous screening of six individuals, two 192 additional control reactions are included in the 8-tube strip panel to ensure validity of results . First, as the 193 reaction utilises an unbuffered LAMP system and pH sensitive d ye to detect amplification, a colour reference 194 control reaction was included, which omits polymerase enzymes to prevent amplification. This reaction will 195 always remain pink with the exact shade of pink varying based on the starting pH of the eluted nucleic acids . 196 Depending on the sample type ( e.g. capillary blood from finger pricks versus dried blood spot eluates ), this 197 starting pH and subsequent colour can vary slightly. Therefore, this reaction provided a reference colour against 198 which all other reaction outcomes were compared. An internal control reaction was also included, targeting an 199 exogenous DNA template lyophilized with the rest of the reaction reagents , which amplified if the correct 200 incubation temperature was reached, the reaction was incubated for sufficient time, and reagents were not 201 damaged in storage. 202 Reaction positions in the 8 -tube strip, from left to right, were as follows: colour reference control (tu be 1), six 203 independent Pan/Pf reactions (tubes 2 –7), and the internal control (tube 8). Each reaction was reconstituted 204 with 20 µL of sample eluate. Isothermal a mplification was conducted using a portable dry bath heat block 205 (ProtonDx, UK) set at 63.5 °C. Finally, while the heat block was powered by mains electricity in this study, it is 206 also compatible with portable batteries, standard 12-volt supplies, or solar panels. It requires less than 20W of 207 continuous power to maintain the set temperature, supporti ng its suitability for decentralized testing in 208 resource-limited settings. 209 Assessment of analytical specificity and incubation time 210 To assess the risk of false positive, the analytical specificity of our test was evaluated at different incubation time 211 points up to 60 minutes. At 50 minutes, the test demonstrated a specificity of 98.3% (95% CI: 96.0–100%), with 212 2 false positive results out of 120 negative samples tested , as shown in Supplementary Table 2. A maximum 213 incubation time of 40 minutes was required to confirm negative results, although most positive reactions turned 214 yellow within 20 minutes. 215 Comparison of analytical sensitivity using in vitro cultured ring stages (3D7 strain) 216 Using s erial dilution s of 3D7 parasite culture s, both LAMP platforms , Dragonfly (input volume 100 µL) and 217 Alethia® (input volume 50 µL), consistently detected all replicates down to a parasitaemia of 0.9 parasites/μL, 218 outperforming DBS-qPCR (input equivalent to 3 discs of 3 mm diameter), which showed decreased sensitivity 219 below 3.8 parasites/μL. WB-qPCR (input volume 100 µL) demonstrated the highest analytical sensitivity among 220 the four molecular-based methods (Fig. 4a). The LODs (parasites/μL) estimated by probit analysis were 2.9 [95% 221 CI: 1.8 -4.8], 0.7 [95% CI: 0.3-1.3], 0.6 [95% CI: 0.3-1.4], and 0.4 [95% CI: 0.2-0.6] for DBS -qPCR, Alethia®, 222 Dragonfly, and WB-qPCR, respectively (Fig. 4b). 223 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 14, 2025. ; https://doi.org/10.1101/2025.07.12.25331425doi: medRxiv preprint 8 224 Figure 4. Comparison of analytical sensitivity of malaria detection using spiked whole blood across Dragonfly, 225 Alethia®, DBS-qPCR, and WB-qPCR. Experiments were conducted in collaboration with the London School of 226 Hygiene and Tropical Medicine (LSHTM) using spiked EDTA-blood with ring-stage Plasmodium falciparum 3D7 227 strain. Results are shown in terms of (a) number and percentage of successfully detected replicates at each spike 228 concentration as well as (b) the resulting empirically determined LOD through probit analysis. The LOD is defined 229 as the parasite density at which the probability of a positive result is ≥95%. Created in BioRender. Cavuto, M. 230 (2025) https://BioRender.com/rwbh4tn. 231 232 Validation of the Dragonfly platform against RDT and light microscopy using DBS-qPCR as the gold standard 233 First, a total of 50 capillary blood specimens from febrile malaria patients, purposively selected as positive 234 controls based on concordant Plasmodium detection by LM, RDT, and DBS-qPCR, were also tested using the 235 Dragonfly platform. All 50 samples tested positive, confirming the compatibility of our method with real, field-236 collected capillary blood samples prior to evaluation on unlabelled community survey samples . Next, 672 237 capillary blood samples collected at the community-level in The Gambia and Burkina Faso were used to evaluate 238 the performance of the Dragonfly platform against DBS-qPCR as the reference method. All samples were also 239 assessed using expert LM and RDTs to benchmark the performance of our me thod against the two standard 240 diagnostic approaches for malaria. Of the 672 samples, 27.1% (146/672) were positive for P. falciparum by DBS-241 qPCR. These positive samples represented a broad range of parasite densities, including both microscopically 242 detectable (n=103) and submicroscopic infections (n=43). A breakdown of the sample categories is presented in 243 Fig. 5. Detailed results obtained using Dragonfly, Alethia®, DBS-qPCR, and WB-qPCR for each of the 50 positive 244 controls from confirmed malaria patients, as well as the 672 capillary blood samples, are provided in the 245 Supplementary Data. 246 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 14, 2025. ; https://doi.org/10.1101/2025.07.12.25331425doi: medRxiv preprint 9 247 Figure 5. Clinical sample s selection to evaluate the performance of Dragonfly Pan/Pf malaria platform. 248 Submicroscopic parasit aemia is defined as a parasite density of < 16 parasites/μL , corresponding to the 249 theoretical L OD for an expert microscopist, based on the ability to detect one asexual parasite among 500 250 leukocytes, assuming a white blood cell count of 8,000 leukocytes/μL. Created in BioRender. Cavuto, M. (2025) 251 https://BioRender.com/rwbh4tn. 252 As depicted in Fig. 6, considering the 672 samples, the overall sensitivity and specificity of our method against 253 DBS-qPCR was 95.2% [95% CI: 90.4–98.1] and 9 6.8% [95% CI: 9 4.9–98.0], respectively. Comparatively, the 254 Dragonfly method demonstrated a higher sensitivity than both RDT ( 50.7% [95% CI: 42.3-59.0]) and expert LM 255 (70.5% [95% CI: 62.4-77.8]). All three methods achieved specificities above 9 6%, with expert LM recording the 256 lowest false-positive rate. 257 258 Figure 6. Comparison of the clinical performance of Dragonfly, RDT, and LM using whole blood finger prick 259 samples, with DBS -qPCR as the gold -standard comparator. For each group, the number of true positives (TP), 260 total positive cases, sensitivity rate with 95% CI, number of true negatives (TN), total negative cases, and 261 specificity rate with 95% CI are provided. FP= false positive, FN=false negative. Created in BioRender. Cavuto, M. 262 (2025) https://BioRender.com/rwbh4tn. 263 When considering samples from asymptomatic individuals only (N=646, including 137 malaria positive and 509 264 negative samples determined by DBS-qPCR) the sensitivity gap remained similar, with Dragonfly, LM, and RDTs 265 detecting 94.9%, 70.1%, and 49.6% of posi tive samples, respectively. Confusion matrices for this subset of 266 samples, as well as the smaller subset of symptomatic cases, are provided as Supplementary Fig. 1. 267 As summarised in Fig. 7, the 146 DBS-qPCR-positive specimens were stratified into four parasite-density groups: 268 200 parasites/µL 269 (n = 62). Across the three lower-density categories Dragonfly markedly outperformed RDT. Dragonfly detected 270 41 of 43 specimens in the <16 parasites/µL group (95.3 %), 25 of 28 specimens in the 16–100 parasites/µL group 271 (89.3 %), and all 13 specimens in the 100–200 parasites/µL group (100 %). In contrast, the RDT detected 2 of 43 272 specimens (4.7 %), 9 of 28 specimens (32.1 %), and 8 of 13 specimens (61.5 %) in the corresponding groups, 273 respectively. Dragonfly also demonstrated significantly higher sensitivity than expert LM for sub -microscopic 274 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 14, 2025. ; https://doi.org/10.1101/2025.07.12.25331425doi: medRxiv preprint 10 parasitaemia ( 0.05) were observed between 275 Dragonfly and expert LM in the 16–100 and 100–200 parasites/µL groups. When parasite density exceeded 200 276 parasites/µL, all three methods showed comparable performance (p > 0.05). 277 278 Figure 7. Detection of DBS-qPCR positive samples by Dragonfly, RDT, and expert LM, stratified by four parasite 279 density categories: 200 parasites/µL. Dragonfly correctly identified 95.3% of 280 submicroscopic infections, 41 out of 43 samples detected with <16 parasites/µL, significantly outperforming 281 both expert LM and RDT. Dragonfly also significantly outperformed RDT at densities ranging from 16 –200 282 parasites/µL, with no significant difference observed between Dragonfly and expert LM in this range. At parasite 283 densities >200 parasites/µL, no statistically significant differences were observed among the three methods (p 284 > 0.05). Created in BioRender. Cavuto, M. (2025) https://BioRender.com/rwbh4tn. 285

286 Through collaborative efforts between UK and West African institutions, this study present ed a near -POC 287 colourimetric LAMP-based extracted molecular solution to achieving accurate detection of sub-microscopic 288 Plasmodium infections from whole blood at the community level in Sub -Saharan Africa. Our approach 289 demonstrated high analytical performance, achieving an LOD of 0.6. parasites/µL [95% CI: 0.3-1.4] with spiked 290 samples. Field evaluation demonstrated sensitivity and specificity of 95.2% [95CI: 90.4-98.1%] and 96.8% [95CI: 291 94.9-98.0%] from individuals enrolled at community level , most of them (96%) asymptomatic. This high 292 diagnostic accuracy, along with the ability to detect 95.3% of the submicroscopic infections (<16 parasites/µL), 293 most of which were missed by both RDT and LM, suggests that the approach may be considered a valuable tool 294 for community-based ADIs in malaria-endemic regions. Moreover, all readings in our study were performed by 295 expert microscopists to ensure accurate identification of parasites, especially for low parasite density infections. 296 This rigorous approach likely contributed to the higher sensitivity by LM observed in our study compared to 297 routine clinical practice, where blood slide may be read by less experienced technicians. 298 Since the introduction of LAMP technology for malaria detection, more than 26 LAMP assays have been 299 developed and evaluated, demonstrating an estimated pooled sensitivity of 97.1% [95CI: 95.7 -98.0%] as 300 reported in a previous meta -analysis study including both symptomatic and asymptomatic individuals .39 301 However, to date only two LAMP-based diagnostic tests (LoopampTM Malaria and Alethia® Malaria) are currently 302 commercialised for malaria , suggesting that high technical performance alone is insufficient to ensure the 303 sustainable deployment of a diagnostic test in its intended target settings.19,24. Performance studies conducted 304 in both health facilities and community settings demonstrated varying sensitivities of these two commercial 305 options, ranging from 97.2% to 40.8% for symptomatic and asymptomatic cases, respectively.40–44 In our study, 306 Alethia® demonstrated comparable high analytical performance, which combined with its user-friendliness likely 307 contributes to its widespread adoption in high-income countries to guide malaria diagnosi s in returning 308 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 14, 2025. ; https://doi.org/10.1101/2025.07.12.25331425doi: medRxiv preprint 11 travellers.28–30 However, its deployment in rural African settings, especially for large-scale community-based 309 malaria screening interventions is constrained by its high cost and the limited capacity of the Alethia incubator 310 accommodating only 10 samples per run. In contrast, Dragonfly combines high diagnostic accuracy with several 311 key advantages for field deployment in resource constrained settings, meeting all the essential (and many of the 312 desirable) technical and health systems criteria outlined in the malERA Target Product Profiles for diagnostics 313 intended for malaria screening and surveillance , as detailed in Supplementary Table 3.17 These specifications 314 include, ease of sample collection, high sensitivity and specificity, rapid turnaround time, ease of use and 315 portability. For example, the SmartLid DNA extraction metho d is compatible with capillary finger prick whole 316 blood, enabling field sampling using well accepted procedures (finger pricks blood samples) in routine healthcare 317 and malaria testing across Sub -Saharan Africa . Furthermore, the SmartLid technology and highl y optimized 318 protocol ensures quality nucleic acid extraction and purification in a fraction of the time (< 15 minutes for up to 319 12 samples at a time) of gold-standard PCR methods which can take well over an hour and rely on bulky and 320 expensive centrifuges. On the detection side, the lyophilised isothermal colourimetric LAMP chemistry forgoes 321 the cost and bulk of thermocyclers and devices that rely on LED illuminators or fluorescent detectors to 322 determine the result. Altogether, 12 whole blood samples can be extracted and amplified from start to finish in 323 as little as 35 minutes for high parasitaemia samples, relying only on a vortex mixer and two low-cost isothermal 324 heat blocks for powered equipment, and room temperature storage for all consumables. Further comparison 325 of the characteristics of the Dragonfly platform with the two commercial Malaria LAMP technology is shown in 326 Supplementary Table 4. 327 ADIs such as Mass-Testing-and-Treatment (MTAT) and Focused Testing and Treatment (FTAT) are currently not 328 recommended by WHO due to their limited or negligible impact on malaria prevalence and incidence of clinical 329 malaria.36,37 However, such recommendations are primarily based on intervention trials that used RDT and/or 330 LM for malaria diagnosis. 36,37 Recent mod elling studies suggested that deploying a diagnostic test with 331 sufficiently high sensitivity, such as one that reduces the limit of detection below 200 parasites/µL, can 332 accelerate malaria elimination, provided a high coverage of sufficient duration is achieved, and the treatment is 333 efficacious.8,9,38 It has been further shown that by reducing the LOD below 2 parasites/µL, MTAT strategies could 334 substantially increase the identification of s ub-microscopic cases, leading to a reduction in Plasmodium 335 falciparum PCR-based prevalence and a decrease in the required number of intervention rounds.9 336 Since our platform is still in a prototype development stage , a comprehensive cost analysis of the final 337 manufactured version is currently unavailable but will be a significant factor in the assay's suitability for 338 deployment in sub-Saharan Africa. However, a preliminary cost analysis is provided in Supplementary Table 5, 339 based on prototype quantities and estimations at low - and medium -scales, and is compared to the current 340 market costs of the Alethia® Malaria Test and associated required instrument. Another limitation of our study 341 is that Dragonfly malaria testing was performed in standard laboratories in The Gambia and Burkina Faso which 342 do not reflect the real -life conditions of field sampling and testing . Therefore, future studies will assess the 343 robustness of the device in more decentralized environments such as community settings and gather detailed 344 insights into user experiences and device usability in the field. In preparation for this investigation, the SmartLid 345 extraction method described here is being adapted into the true -POC single -use format of the previously 346 presented Dragonfly sample-to-result platform,32 with all reagent s and buffers pre-aliquoted and laboratory 347 micro-pipettes replaced with disposable exact volume pipettes. Finally, although the technology development 348 and adaptation were performed in laboratories in London through an active collaboration with African 349 Institutions, future initiatives should focus on extending the concept of transferability to local development and 350 production in Africa. This approach would facilitate the sustainable manufacturing and distribution of the 351 diagnostic platform, thereby ensuring its availability in malaria-endemic regions. 352 In addition to demonstrating the strong potential of the SmartLid extraction and Dragonfly colourimetric LAMP 353 technologies towards malaria elimination strategies, this work highlighted the rapid adaptability of the 354 combined approach and potential for assay transfer , which should be a key criterion for selecting molecular 355 diagnostic tests in Sub-Saharan Africa.45 Our approach’s significant use of off -the-shelf consumables and 356 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 14, 2025. ; https://doi.org/10.1101/2025.07.12.25331425doi: medRxiv preprint 12 flexibility enabled the rapid prototyping and deployment of the presented multi-patient Pan/Pf malaria test from 357 capillary finger prick blood samples. Finally, the potential for future digital and cloud integration of the Dragonfly 358 platform, as explored in previous works demonstrating solutions for viral respiratory and skin infection 359 diagnostics, aligns with the growing need for connected diagnostics solutions in Africa.32,46 360 ONLINE METHODS 361 Comparison of analytical sensitivity using in vitro cultured ring stages (3D7 strain) 362 The analytical sensitivity of Dragonfly was evaluated in comparison with Alethia®, DBS-qPCR and WB-qPCR. The 363 comparison was performed using serial dilutions of ring -stage Plasmodium falciparum (3D7 strain) parasite 364 culture at the malaria parasitology lab of the LSHTM in London. Parasites were cultured in-vitro and synchronized 365 using a magnetic separation procedure.47 To minimise the occurrence of red blood cells ( RBCs) infected by 366 multiple parasites, cultures were maintained under gentle agitation (60RPM), resulting in 85% of singly infected 367 RBC among the total infected cells (Supplementary Fig. 2). The final parasitaemia of the culture, measured at 368 5.8% of RBCs, was confirmed by expert LM. A Plasmodium-negative blood sample (tested by WB-qPCR) was then 369 spiked to generate the initial infected sample, yielding a parasitaemia of 6,000 parasites/µL. Serial dilutions were 370 subsequently performed using the same negative whole blood sample to generate samples with decreasing 371 concentrations, calculated based on the dilution factor applied. Each dilution point was tested multiple times 372 across the four molecular-based methods. The concentrations of each dilution point, as well as the number of 373 replicates per method, are summarized in Fig. 4. 374 Alethia® is a commercially available LAMP -based technology that utilises a Plasmodium genus-specific assay. 375 The system consists of an initial DNA extraction process using a passive gel filtration column, followed by an 376 amplification step with lyophilised reagents in a dedicated LAMP incubator. The test provides on a LCD screen a 377 qualitative result (positive or negative ), which is automatically interpreted by a reader integrated within the 378 incubator. As per the manufacturer's guidelines, 50μL of whole blood was used as the input volume in our 379 evaluation. 380 DNA was extracted from DBS (3 discs of 3 mm) and whole blood samples (100 µL) using the QIAamp DNA Mini 381 Kit according to the manufacturer’s instructions. DNA was eluted in AE Buffer with a final volume of 100 µL for 382 DBS and 200 µL for whole blood samples. The Plasmodium falciparum-specific PCR assay applied to both sample 383 types has been previously described.48 The PCR reaction volume was 20 µL, comprising 5 µL of DNA extract, 10 384 µL of GoTaq qPCR Master Mix 2X, 2 µL of P. falciparum primer/probe mix (F/R/P] 10X, and 3 µL of PCR grade 385 water. Amplification was performed on a LightCycler® using the following cycling conditions: an initial 386 denaturation at 95 °C for 2 minutes, followed by 45 cycles of denaturation at 95 °C for 15 seconds and 387 annealing/extension at 60 °C for 1 minute. 388 Dragonfly malaria field validation against RDT and light microscopy versus DBS-qPCR 389 The compatibility of the adapted Dragonfly method with finger prick clinical samples was assessed using 50 390 blood specimens purposively selected based on their positivity for Plasmodium by RDT, LM and DBS-qPCR. These 391 specimens were obtained from febrile patients attending rural health facilities in the Central Region in Burkina 392 Faso. Parasite densities, determined by expert LM, ranged from 149 to 87,500 parasites/µL with a median [IQR] 393 of 714 parasites/µL [95% CI: 128-5,688 parasites/µL]. 394 A total of 672 blood specimens were tested to evaluate the performance of the presented method compared to 395 RDT and LM using DBS-qPCR as gold standard . Blood specimens were collected from individuals enrolled in a 396 community-based survey in two malaria endemic sites, i.e., The Central Region in Burkina Faso and The Upper 397 River Region in The Gambia, characterised by a predominance of P. falciparum. Baseline characteristics of study 398 participants are summarized in Table 1 and show a slight predominance of females (57%) and a fair 399 representation of all age categories. Most participants (96%) were asymptomatic at the time of sample 400 collection. 401 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 14, 2025. ; https://doi.org/10.1101/2025.07.12.25331425doi: medRxiv preprint 13 Table 1. Characteristics of study Participants. 402 Characteristics n (%) Sex Male 289 (43.0) Female 383(57.0) Age groups in years 0.5–4 78 (11.6) 5–14 198 (29.5) 15–29 126 (18.7) 30–59 188 (28.0) ≥60 82 (12.2) Symptomatic status Symptomatic 26 (3.9) Asymptomatic 646 (96.1) Sample collection sites The Gambia 281 (41.8) Burkina Faso 391 (58.2) 403 All malaria tests were performed using capillary blood collected by finger prick. RDT testing was performed on-404 site, either in a health facility (50 blood specimens used for the compatibility assessment) or at community level 405 (672 blood specimens used for the diagnostic performance evaluation) . LM readings were performed on thick 406 smears. Clinical blood specimens used for the Dragonfly method collected into 200 μL EDTA microtainers were 407 stored between 3-5 °C if they were tested the same day as sample collection, or otherwise stored at –20 °C until 408 testing. Blood spots from the same finger prick were collected on DBS card and used for qPCR analysis. 409 Malaria confirmation by DBS-qPCR 410 DBS-qPCR was selected as the gold standard method for evaluating the performance of the developed platform. 411 The DBS -qPCR testing was conducted at the MRCG following a standardized protocol. 49 The gold standard 412

consisted of capillary blood spotted onto Whatmann filter paper followed by a qPCR detection using a 413 TaqMan assay targeting the var gene acidic terminal sequence of Plasmodium falciparum. The number of copies 414 of the var gene is approximately 59 per genome. Genomic DNA extraction was performed using the QIAamp 96 415 DNA QIAcube HT Kit, in accordance with the manufacturer’s instructions. Three discs of 3mm diameter were 416 punched out of the DBS and the DNA extracts were eluted in a volume of 80 µL of AE Buffer. DNA amplification 417 was conducted using the Bio-Rad CFX96 real-time PCR machine. Each reaction utilized 5 µL of sample DNA. The 418 VarATS qPCR master mix consisted of 1 µL of PCR-grade water, 10 µL of 2X TaqMan Master Mix, 1.6 µL of 10 µM 419 Var forward primer, 1.6 µL of 10 µM Var reverse primer, and 0.8 µL of 10 µM Var probe. The cycling conditions 420 were: 1 cycle at 50 °C for 2 min and 1 cycle at 95 °C for 10 min followed by 45 cycles at 95 °C and 55 °C for 421 respectively, 15sec and 1min. If the result was positive for either PCR or Dragonfly alone, the PCR was repeated 422 in duplicates starting from a new DNA extraction. If at least one of the repeats was positive, the sample was 423 considered positive for Plasmodium falciparum. 424 425 Light microscopy 426 The thick blood smear was prepared from fresh capillary blood and air dried. Giemsa staining was performed 427 with a 3% solution for 30 minutes. All slides were independently read by two qualified microscopists. Parasite 428 density was determined by counting the number of asexual parasites per 500 leukocytes on the thick smear, 429 assuming a leukocyte count of 8,000/µl, and using 100x magnification with optical LM. In case of discrepancy 430 between the two readers such as one reporting negative and the other positive for malaria or a density 431 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 14, 2025. ; https://doi.org/10.1101/2025.07.12.25331425doi: medRxiv preprint 14 difference of > 50%, or different species reported, a third expert microscopist would read the slide as well. The 432 reported parasite density was the geometric mean of the two readers' results or that of the two closest readings 433 if a third reading was done. 434 Rapid-diagnostic tests 435 Plasmodium falciparum infection status of all 672 community-collected samples was assessed using the SD 436 BIOLINE Malaria P. falciparum Ag Test™ (Abott), an HRP2-based immunolateral flow assay accepted on the WHO 437 list of prequalified in vitro diagnostics.50,51 438 The Dragonfly Pan/Pf malaria test workflow 439 Extractions were performed manually using the SmartLid sample preparation in bulk reagents format allowing 440 for efficient processing of multiple samples simultaneously. A total of 100 µL of capillary blood sample was used 441 as input for each extraction. Purified samples were eluted in a total elution volume of 50 µL. LAMP reactions 442 were performed by adding 20 µL of the eluted sample into each tube of the colorimetric LyoLAMP Pan/Pf Test 443 Panels (ProtonDx). The Pan-Plasmodium (Pan18s) and P. falciparum ( PfK13, Pf mtDNA) assays's genes and 444 primer sequences used in this study are listed in Supplementary Table 1. Amplification was conducted in the 445 portable thermal block (ProtonDx) for 40 min at 63.5 °C. Immediately following amplification, results were 446 visually assessed by the technician based on the colour change in each tube, with a change from pink (negative) 447 to yellow (yellow) in case of P. falciparum infection. 448 All testing was carried out in standard molecular biology laboratories at MRCG Unit The Gambia and Clinical 449 Research Unit Nanoro, Burkina Faso. All steps, from DNA extraction to result -read-out, were conducted in a 450 single workspace without the need for a laboratory hood. Prior to testin g, a half -day training program was 451 conducted for users including both theoretical instruction s and hands-on practices using 3D7 -infected whole 452 blood and negative controls. 453 Data Analysis 454 All samples were fully anonymized prior to analysis. Data from the specificity and analytical sensitivity 455 experiments are provided in Supplementary Data. Socio-demographics, DBS-qPCR, RDT, LM and Dragonfly data 456 of the total clinical samples (n=672) were merged into a unique database and statistical analyses conducted 457 using Stata 18 (StataCorp, College Station, TX, USA) and R (R Core Team, Vienna, Austria) (Supplementary Data). 458 Summary statistics were performed using the median value and IQR for continuous variables, and proportions 459 and 95%C Is for categorical variables. Sensitivity and specificity with the ir 95% confidence intervals were 460 calculated for Dragonfly, LM and RDT using DBS-qPCR as the gold standard according to formulas shown in 461 Supplementary Table 6. The LOD was estimated using probit analysis defining LOD as the concentration with a 462 95% probability of obtaining a positive result.52 To assess the differences between groups, McNemar’s Test was 463 employed. A p-value < 0.05 was considered statistically significant. 464 Ethics 465 Collection of samples during community-based surveys in The Upper River Region in The Gambia was approved 466 by the LSHTM Ethics Commit tee and The Gambia Government/MRC Joint E thics Committee ( Ref. 29611). In 467 Burkina Faso, sample collection from symptomatic patients enrolled at health facilities in the Central West 468 Region received approval from The Comite d’Ethique pour la Recherche en Sante -Burkina Faso (ref: 469 DELIBERATION N°2021-04-084) and the Comité d’Éthique Institutionnel pour la Recherche en Sciences de la 470 Santé of IRSS (N/Réf. A07 -2021/CEIRES). T he approval for collecting samples from participants enrolled at 471 community level was granted by The Comite d’Ethique pour la Recherche en Sante -Burkina Faso (ref: 472 DELIBERATION N° 2024-0361). Written informed consent was obtained from all research participants and/or 473 guardians before recruitment. 474

475 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 14, 2025. ; https://doi.org/10.1101/2025.07.12.25331425doi: medRxiv preprint 15 This work was supported by the Dep artment of Health and Social Care -funded Centre for Antimicrobial 476 Optimisation (CAMO) at Imperial College London; the London School of Hygiene and Tropical Medicine (LSHTM), 477 the Wellcome Trust CAMO -Net programme [226691/Z/22/Z] ; the Wellcome Trust Innovator Award 478 [215688/Z/19/Z]; and the Imperial College Research Fellowship to KMC [WDPI.G09074]. In addition, this work 479 was funded by the NIHR [NIHR134694] using UK aid from the UK Government to support global health research. 480 Infrastructure support for this research was funded by the NIHR Imperial Biomedical Research Centre (BRC). The 481 views expressed in this publication are those of the authors and not necessarily those of the NIHR or the UK 482 Department of Health and Social Care. P.G. and J.R.M. are affiliated with the NIHR Health Protection Research 483 Unit (HPRU) in Healthcare Associated Infections and Antimicrobial Resistance at Imperial College London in 484 partnership with the UK Health Security Agency, in collaboration with Imperial Healthcare Partners, the 485 University of Cambridge and the University of Warwick. 486 AUTHOR CONTRIBUTIONS 487 D.R.R., I.P. and M.L.C. contributed equally to the work. Study concept and design: D.R.R., I.P., A.E., U.D.A., H.T., 488 A.C., and J.R.M. Acquisition, analysis, or interpretation of data: D.R.R., I.P., M.L.C., F.K., M.C., D.Y.S., E.Q., K.M.C., 489 M.O.N., S.C., B.D., L.B.S., P.G., M.B., H.T., A.J.C., A.E., U.D.A., and J.R.M. Drafting the manuscript: D.R.R., I.P., 490 M.L.C., J.R.M. Critical revision of the manuscript: D.R.R., I.P., M.L.C., F.K., M.C., D.Y.S., E.Q., K.M.C., M.O.N., S.C., 491 B.D., L.B.S., P.G., M.B., H.T., A.J.C., A.E., U.D.A., and J.R.M. The manuscript was written through the contributions 492 of all authors. All authors have given approval to the final version of the manuscript. 493 COMPETING INTERESTS 494 The authors declare the following competing financial interest(s): I.P., M.L.C., E.Q., K.M.C., P.G. and J.R.M. have 495 financial interest on ProtonDx Ltd, which currently has exclusive license to intellectual property linked to 496 Dragonfly (WO2023131803A1) and SmartLid (WO2022180376A1), and its associated trademark. These authors 497 declare that they do not have any other known competing financial interests or personal relationships that could 498 have appeared to influence the work reported in this paper. The remaining authors declare no competing 499 interests. 500 DATA AVAILABILITY 501 All data supporting the findings of this study are available within the article and its supplementary files. Any 502 additional requests for information can be directed to, and will be fulfilled by, the corresponding authors. Source 503 data are provided with this paper. 504

505 (1) WHO. World malaria report 2024. https://www.who.int/teams/global-malaria-506 programme/reports/world-malaria-report-2024 (accessed 2024-12-16). 507 (2) Feachem, R. G. A.; Chen, I.; Akbari, O.; Bertozzi-Villa, A.; Bhatt, S.; Binka, F.; Boni, M. F.; Buckee, C.; 508 Dieleman, J.; Dondorp, A.; Eapen, A.; Sekhri Feachem, N.; Filler, S.; Gething, P.; Gosling, R.; Haakenstad, 509 A.; Harvard, K.; Hatefi, A.; Jamison, D.; Jones, K. E.; Karema, C.; Kamwi, R. N.; Lal, A.; Larson, E.; Lees, M.; 510 Lobo, N. F.; Micah, A. E.; Moonen, B.; Newby, G.; Ning, X.; Pate, M.; Quiñones, M.; Roh, M.; Rolfe, B.; 511 Shanks, D.; Singh, B.; Staley, K.; Tulloch, J.; Wegbreit, J.; Woo, H. J.; Mpanju-Shumbusho, W. Malaria 512 Eradication within a Generation: Ambitious, Achievable, and Necessary. The Lancet 2019, 394 (10203), 513 1056–1112. https://doi.org/10.1016/S0140-6736(19)31139-0. 514 (3) Bhatt, S.; Weiss, D. J.; Cameron, E.; Bisanzio, D.; Mappin, B.; Dalrymple, U.; Battle, K. E.; Moyes, C. L.; 515 Henry, A.; Eckhoff, P. A.; Wenger, E. A.; Briët, O.; Penny, M. A.; Smith, T. A.; Bennett, A.; Yukich, J.; Eisele, 516 T. P.; Griffin, J. T.; Fergus, C. A.; Lynch, M.; Lindgren, F.; Cohen, J. M.; Murray, C. L. J.; Smith, D. L.; Hay, S. 517 I.; Cibulskis, R. E.; Gething, P. W. The Effect of Malaria Control on Plasmodium Falciparum in Africa 518 between 2000 and 2015. Nature 2015, 526 (7572), 207–211. https://doi.org/10.1038/nature15535. 519 (4) WHO. Global technical strategy for malaria 2016-2030, 2021 update. https://www.who.int/publications-520 detail-redirect/9789240031357 (accessed 2023-01-20). 521 (5) Bousema, T.; Okell, L.; Felger, I.; Drakeley, C. Asymptomatic Malaria Infections: Detectability, 522 Transmissibility and Public Health Relevance. Nat. Rev. Microbiol. 2014, 12 (12), 833–840. 523 https://doi.org/10.1038/nrmicro3364. 524 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 14, 2025. ; https://doi.org/10.1101/2025.07.12.25331425doi: medRxiv preprint 16 (6) Sturrock, H. J. W.; Hsiang, M. S.; Cohen, J. M.; Smith, D. L.; Greenhouse, B.; Bousema, T.; Gosling, R. D. 525 Targeting Asymptomatic Malaria Infections: Active Surveillance in Control and Elimination. PLoS Med. 526 2013, 10 (6), e1001467. https://doi.org/10.1371/journal.pmed.1001467. 527 (7) Kim, S.; Luande, V. N.; Rocklöv, J.; Carlton, J. M.; Tozan, Y. A Systematic Review of the Evidence on the 528 Effectiveness and Cost-Effectiveness of Mass Screen-and-Treat Interventions for Malaria Control. Am. J. 529 Trop. Med. Hyg. 2021, 105 (6), 1722–1731. https://doi.org/10.4269/ajtmh.21-0325. 530 (8) Nankabirwa, J. I.; Arinaitwe, E.; Briggs, J.; Rek, J.; Rosenthal, P. J.; Kamya, M. R.; Olwoch, P.; Smith, D. L.; 531 Rodriguez-Barraquer, I.; Dorsey, G.; Greenhouse, B. Simulating the Impacts of Augmenting Intensive 532 Vector Control with Mass Drug Administration or Test-and-Treat Strategies on the Malaria Infectious 533 Reservoir. Am. J. Trop. Med. Hyg. 2022, 107 (5), 1028–1035. https://doi.org/10.4269/ajtmh.21-0953. 534 (9) Slater, H. C.; Ross, A.; Ouédraogo, A. L.; White, L. J.; Nguon, C.; Walker, P. G. T.; Ngor, P.; Aguas, R.; Silal, 535 S. P.; Dondorp, A. M.; La Barre, P.; Burton, R.; Sauerwein, R. W.; Drakeley, C.; Smith, T. A.; Bousema, T.; 536 Ghani, A. C. Assessing the Impact of Next-Generation Rapid Diagnostic Tests on Plasmodium Falciparum 537 Malaria Elimination Strategies. Nature 2015, 528 (7580), S94-101. https://doi.org/10.1038/nature16040. 538 (10) Okell, L. C.; Ghani, A. C.; Lyons, E.; Drakeley, C. J. Submicroscopic Infection in Plasmodium Falciparum-539 Endemic Populations: A Systematic Review and Meta-Analysis. J. Infect. Dis. 2009, 200 (10), 1509–1517. 540 https://doi.org/10.1086/644781. 541 (11) Drakeley, C.; Gonçalves, B.; Okell, L.; Slater, H.; Drakeley, C.; Gonçalves, B.; Okell, L.; Slater, H. 542 Understanding the Importance of Asymptomatic and Low- Density Infections for Malaria Elimination. In 543 Towards Malaria Elimination - A Leap Forward; IntechOpen, 2018. 544 https://doi.org/10.5772/intechopen.77293. 545 (12) Whittaker, C.; Slater, H.; Nash, R.; Bousema, T.; Drakeley, C.; Ghani, A. C.; Okell, L. C. Global Patterns of 546 Submicroscopic Plasmodium Falciparum Malaria Infection: Insights from a Systematic Review and Meta-547 Analysis of Population Surveys. Lancet Microbe 2021, 2 (8), e366–e374. https://doi.org/10.1016/S2666-548 5247(21)00055-0. 549 (13) Tadesse, F. G.; Slater, H. C.; Chali, W.; Teelen, K.; Lanke, K.; Belachew, M.; Menberu, T.; Shumie, G.; 550 Shitaye, G.; Okell, L. C.; Graumans, W.; van Gemert, G.-J.; Kedir, S.; Tesfaye, A.; Belachew, F.; Abebe, W.; 551 Mamo, H.; Sauerwein, R.; Balcha, T.; Aseffa, A.; Yewhalaw, D.; Gadisa, E.; Drakeley, C.; Bousema, T. The 552 Relative Contribution of Symptomatic and Asymptomatic Plasmodium Vivax and Plasmodium Falciparum 553 Infections to the Infectious Reservoir in a Low-Endemic Setting in Ethiopia. Clin. Infect. Dis. 2018, 66 (12), 554 1883–1891. https://doi.org/10.1093/cid/cix1123. 555 (14) Ouédraogo, A. L.; Gonçalves, B. P.; Gnémé, A.; Wenger, E. A.; Guelbeogo, M. W.; Ouédraogo, A.; 556 Gerardin, J.; Bever, C. A.; Lyons, H.; Pitroipa, X.; Verhave, J. P.; Eckhoff, P. A.; Drakeley, C.; Sauerwein, R.; 557 Luty, A. J. F.; Kouyaté, B.; Bousema, T. Dynamics of the Human Infectious Reservoir for Malaria 558 Determined by Mosquito Feeding Assays and Ultrasensitive Malaria Diagnosis in Burkina Faso. J. Infect. 559 Dis. 2016, 213 (1), 90–99. https://doi.org/10.1093/infdis/jiv370. 560 (15) Zimmerman, P. A.; Howes, R. E. Malaria Diagnosis for Malaria Elimination. Curr. Opin. Infect. Dis. 2015, 561 28 (5), 446–454. https://doi.org/10.1097/QCO.0000000000000191. 562 (16) Kamau, E.; Tolbert, L. S.; Kortepeter, L.; Pratt, M.; Nyakoe, N.; Muringo, L.; Ogutu, B.; Waitumbi, J. N.; 563 Ockenhouse, C. F. Development of a Highly Sensitive Genus-Specific Quantitative Reverse Transcriptase 564 Real-Time PCR Assay for Detection and Quantitation of Plasmodium by Amplifying RNA and DNA of the 565 18S rRNA Genes. J. Clin. Microbiol. 2020, 49 (8), 2946–2953. https://doi.org/10.1128/jcm.00276-11. 566 (17) malERA Consultative Group on Diagnoses and Diagnostics. A Research Agenda for Malaria Eradication: 567 Diagnoses and Diagnostics. PLoS Med. 2011, 8 (1), e1000396. 568 https://doi.org/10.1371/journal.pmed.1000396. 569 (18) Kreitmann, L.; Miglietta, L.; Xu, K.; Malpartida-Cardenas, K.; D’Souza, G.; Kaforou, M.; Brengel-Pesce, K.; 570 Drazek, L.; Holmes, A.; Rodriguez-Manzano, J. Next-Generation Molecular Diagnostics: Leveraging Digital 571 Technologies to Enhance Multiplexing in Real-Time PCR. TrAC Trends Anal. Chem. 2023, 160, 116963. 572 https://doi.org/10.1016/j.trac.2023.116963. 573 (19) Oriero, E. C.; Jacobs, J.; Van Geertruyden, J.-P.; Nwakanma, D.; D’Alessandro, U. Molecular-Based 574 Isothermal Tests for Field Diagnosis of Malaria and Their Potential Contribution to Malaria Elimination. J. 575 Antimicrob. Chemother. 2015, 70 (1), 2–13. https://doi.org/10.1093/jac/dku343. 576 (20) Malpartida-Cardenas, K.; Moser, N.; Ansah, F.; Pennisi, I.; Ahu Prah, D.; Amoah, L. E.; Awandare, G.; 577 Hafalla, J. C. R.; Cunnington, A.; Baum, J.; Rodriguez-Manzano, J.; Georgiou, P. Sensitive Detection of 578 Asymptomatic and Symptomatic Malaria with Seven Novel Parasite-Specific LAMP Assays and 579 Translation for Use at Point-of-Care. Microbiol. Spectr. 2023, 11 (3), e05222-22. 580 https://doi.org/10.1128/spectrum.05222-22. 581 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 14, 2025. ; https://doi.org/10.1101/2025.07.12.25331425doi: medRxiv preprint 17 (21) Wei, H.; Li, J.; Liu, Y.; Cheng, W.; Huang, H.; Liang, X.; Huang, W.; Lin, L.; Zheng, Y.; Chen, W.; Wang, C.; 582 Chen, W.; Xu, G.; Wei, W.; Chen, L.; Zeng, Y.; Lu, Z.; Li, S.; Lin, Z.; Wang, J.; Lin, M. Rapid and 583 Ultrasensitive Detection of Plasmodium Spp. Parasites via the RPA-CRISPR/Cas12a Platform. ACS Infect. 584 Dis. 2023, 9 (8), 1534–1545. https://doi.org/10.1021/acsinfecdis.3c00087. 585 (22) Han, E.-T. Loop-Mediated Isothermal Amplification Test for the Molecular Diagnosis of Malaria. Expert 586 Rev. Mol. Diagn. 2013, 13 (2), 205–218. https://doi.org/10.1586/erm.12.144. 587 (23) Britton, S.; Cheng, Q.; McCarthy, J. S. Novel Molecular Diagnostic Tools for Malaria Elimination: A Review 588 of Options from the Point of View of High-Throughput and Applicability in Resource Limited Settings. 589 Malar. J. 2016, 15 (1), 88. https://doi.org/10.1186/s12936-016-1158-0. 590 (24) Lucchi, N. W.; Ndiaye, D.; Britton, S.; Udhayakumar, V. Expanding the Malaria Molecular Diagnostic 591 Options: Opportunities and Challenges for Loop-Mediated Isothermal Amplification Tests for Malaria 592 Control and Elimination. Expert Rev. Mol. Diagn. 2018, 18 (2), 195–203. 593 https://doi.org/10.1080/14737159.2018.1431529. 594 (25) Morris, U.; Aydin-Schmidt, B. Performance and Application of Commercially Available Loop-Mediated 595 Isothermal Amplification (LAMP) Kits in Malaria Endemic and Non-Endemic Settings. Diagnostics 2021, 596 11 (2), 336. https://doi.org/10.3390/diagnostics11020336. 597 (26) Human. LoopampTM Malaria Pan Detection Kit. https://www.human.de/malaria-lamp/detection-assay 598 (accessed 2024-09-20). 599 (27) Alethia Malaria Test | Advanced Malaria Detection. Meridian Bioscience. 600 https://www.meridianbioscience.com/diagnostics/disease-areas/other/malaria/alethia-malaria/ 601 (accessed 2024-09-20). 602 (28) Martín-Ramírez, A.; Lanza, M.; Hisam, S.; Perez-Ayala, A.; Rubio, J. M. Usefulness of a Commercial LAMP 603 Assay for Detection of Malaria Infection, Including Plasmodium Knowlesi Cases, in Returning Travelers in 604 Spain. BMC Res. Notes 2022, 15, 147. https://doi.org/10.1186/s13104-022-06037-9. 605 (29) Ljolje, D.; Abdallah, R.; Lucchi, N. W. Detection of Malaria Parasites in Samples from Returning US 606 Travelers Using the Alethia® Malaria Plus LAMP Assay. BMC Res. Notes 2021, 14 (1), 128. 607 https://doi.org/10.1186/s13104-021-05542-7. 608 (30) Payne, R. O.; Edwards, N. J.; Themistocleous, Y.; Silk, S. E.; Barrett, J. R.; Rawlinson, T. A.; Lim, I. W.; 609 Draper, S. J.; Minassian, A. M. Diagnosis of Plasmodium Falciparum Malaria at Very Low Parasitaemias 610 Using a Commercially Available LAMP Assay and RDT. Trans. R. Soc. Trop. Med. Hyg. 2025, traf050. 611 https://doi.org/10.1093/trstmh/traf050. 612 (31) Pennisi, I.; Cavuto, M. L.; Miglietta, L.; Malpartida-Cardenas, K.; Stringer, O. W.; Mantikas, K.-T.; Reid, R.; 613 Frise, R.; Moser, N.; Randell, P.; Davies, F.; Bolt, F.; Barclay, W.; Holmes, A.; Georgiou, P.; Rodriguez-614 Manzano, J. Rapid, Portable, and Electricity-Free Sample Extraction Method for Enhanced Molecular 615 Diagnostics in Resource-Limited Settings. Anal. Chem. 2024, 96 (28), 11181–11188. 616 https://doi.org/10.1021/acs.analchem.4c00319. 617 (32) Manzano, J. R.; Cavuto, M.; Cardenas, K. M.; Pennisi, I.; Pond, M.; Mirza, S.; Moser, N.; Comer, M.; 618 Stokes, I.; Eke, L.; Lant, S.; Szostak, K.; Miglietta, L.; Stringer, O.; Sumner, R.; Bolt, F.; Sriskandan, S.; 619 Holmes, A.; Georgiou, P.; Ulaeto, D. (O); Motes, C. M. de. Dragonfly: A Portable Molecular Diagnostic 620 Platform for Rapid Point-of-Care Detection of Mpox and Other Diseases. Research Square August 20, 621 2024. https://doi.org/10.21203/rs.3.rs-4926828/v1. 622 (33) LyoLamp. ProtonDx. https://www.protondx.com/lifesciences/lyolamp (accessed 2024-09-20). 623 (34) Dragonfly | Point of Care Diagnostic Device. ProtonDx. https://www.protondx.com/dragonfly (accessed 624 2024-01-29). 625 (35) Malpartida-Cardenas, K.; Miscourides, N.; Rodriguez-Manzano, J.; Yu, L.-S.; Moser, N.; Baum, J.; 626 Georgiou, P. Quantitative and Rapid Plasmodium Falciparum Malaria Diagnosis and Artemisinin-627 Resistance Detection Using a CMOS Lab-on-Chip Platform. Biosens. Bioelectron. 2019, 145, 111678. 628 https://doi.org/10.1016/j.bios.2019.111678. 629 (36) WHO. WHO’s recommendations on malaria elimination. https://www.who.int/teams/global-malaria-630 programme/elimination/recommendations-on-malaria-elimination (accessed 2023-04-24). 631 (37) Bhamani, B.; Martí Coma-Cros, E.; Tusell, M.; Mithi, V.; Serra-Casas, E.; Williams, N. A.; Lindblade, K. A.; 632 Allen, K. C. Mass Testing and Treatment to Accelerate Malaria Elimination: A Systematic Review and 633 Meta-Analysis. Am. J. Trop. Med. Hyg. 2024, 110 (4_Suppl), 44–53. https://doi.org/10.4269/ajtmh.23-634 0127. 635 (38) Rosas-Aguirre, A.; Erhart, A.; Llanos-Cuentas, A.; Branch, O.; Berkvens, D.; Abatih, E.; Lambert, P.; Frasso, 636 G.; Rodriguez, H.; Gamboa, D.; Sihuincha, M.; Rosanas-Urgell, A.; D’Alessandro, U.; Speybroeck, N. 637 Modelling the Potential of Focal Screening and Treatment as Elimination Strategy for Plasmodium 638 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 14, 2025. ; https://doi.org/10.1101/2025.07.12.25331425doi: medRxiv preprint 18 Falciparum Malaria in the Peruvian Amazon Region. Parasit. Vectors 2015, 8 (1), 1–12. 639 https://doi.org/10.1186/s13071-015-0868-4. 640 (39) Picot, S.; Cucherat, M.; Bienvenu, A.-L. Systematic Review and Meta-Analysis of Diagnostic Accuracy of 641 Loop-Mediated Isothermal Amplification (LAMP) Methods Compared with Microscopy, Polymerase 642 Chain Reaction and Rapid Diagnostic Tests for Malaria Diagnosis. Int. J. Infect. Dis. 2020, 98, 408–419. 643 https://doi.org/10.1016/j.ijid.2020.07.009. 644 (40) Cook, J.; Aydin-Schmidt, B.; González, I. J.; Bell, D.; Edlund, E.; Nassor, M. H.; Msellem, M.; Ali, A.; Abass, 645 A. K.; Mårtensson, A.; Björkman, A. Loop-Mediated Isothermal Amplification (LAMP) for Point-of-Care 646 Detection of Asymptomatic Low-Density Malaria Parasite Carriers in Zanzibar. Malar. J. 2015, 14 (1), 43. 647 https://doi.org/10.1186/s12936-015-0573-y. 648 (41) Aydin-Schmidt, B.; Morris, U.; Ding, X. C.; Jovel, I.; Msellem, M. I.; Bergman, D.; Islam, A.; Ali, A. S.; Polley, 649 S.; Gonzalez, I. J.; Mårtensson, A.; Björkman, A. Field Evaluation of a High Throughput Loop Mediated 650 Isothermal Amplification Test for the Detection of Asymptomatic Plasmodium Infections in Zanzibar. 651 PloS One 2017, 12 (1), e0169037. https://doi.org/10.1371/journal.pone.0169037. 652 (42) Lucchi, N. W.; Gaye, M.; Diallo, M. A.; Goldman, I. F.; Ljolje, D.; Deme, A. B.; Badiane, A.; Ndiaye, Y. D.; 653 Barnwell, J. W.; Udhayakumar, V.; Ndiaye, D. Evaluation of the Illumigene Malaria LAMP: A Robust 654 Molecular Diagnostic Tool for Malaria Parasites. Sci. Rep. 2016, 6, 36808. 655 https://doi.org/10.1038/srep36808. 656 (43) Polley, S. D.; González, I. J.; Mohamed, D.; Daly, R.; Bowers, K.; Watson, J.; Mewse, E.; Armstrong, M.; 657 Gray, C.; Perkins, M. D.; Bell, D.; Kanda, H.; Tomita, N.; Kubota, Y.; Mori, Y.; Chiodini, P. L.; Sutherland, C. 658 J. Clinical Evaluation of a Loop-Mediated Amplification Kit for Diagnosis of Imported Malaria. J. Infect. 659 Dis. 2013, 208 (4), 637–644. https://doi.org/10.1093/infdis/jit183. 660 (44) Hsiang, M. S.; Ntshalintshali, N.; Kang Dufour, M.-S.; Dlamini, N.; Nhlabathi, N.; Vilakati, S.; Malambe, C.; 661 Zulu, Z.; Maphalala, G.; Novotny, J.; Murphy, M.; Schwartz, A.; Sturrock, H.; Gosling, R.; Dorsey, G.; 662 Kunene, S.; Greenhouse, B. Active Case Finding for Malaria: A 3-Year National Evaluation of Optimal 663 Approaches to Detect Infections and Hotspots Through Reactive Case Detection in the Low-Transmission 664 Setting of Eswatini. Clin. Infect. Dis. 2020, 70 (7), 1316–1325. https://doi.org/10.1093/cid/ciz403. 665 (45) Baldeh, M.; Bawa, F. K.; Bawah, F. U.; Chamai, M.; Dzabeng, F.; Jebreel, W. M. A.; Kabuya, J.-B. B.; 666 Molemodile Dele-Olowu, S. K.; Odoyo, E.; Rakotomalala Robinson, D.; Cunnington, A. J. Lessons from the 667 Pandemic: New Best Practices in Selecting Molecular Diagnostics for Point-of-Care Testing of Infectious 668 Diseases in Sub-Saharan Africa. Expert Rev. Mol. Diagn. 2023. 669 https://doi.org/10.1080/14737159.2023.2277368. 670 (46) Africa CDC. Digital Transformation Strategy. Africa CDC. https://africacdc.org/download/digital-671 transformation-strategy/ (accessed 2024-09-27). 672 (47) Ribaut, C.; Berry, A.; Chevalley, S.; Reybier, K.; Morlais, I.; Parzy, D.; Nepveu, F.; Benoit-Vical, F.; Valentin, 673 A. Concentration and Purification by Magnetic Separation of the Erythrocytic Stages of All Human 674 Plasmodium Species. Malar. J. 2008, 7 (1), 45. https://doi.org/10.1186/1475-2875-7-45. 675 (48) Lazrek, Y.; Florimond, C.; Volney, B.; Discours, M.; Mosnier, E.; Houzé, S.; Pelleau, S.; Musset, L. 676 Molecular Detection of Human Plasmodium Species Using a Multiplex Real Time PCR. Sci. Rep. 2023, 13 677 (1), 11388. https://doi.org/10.1038/s41598-023-38621-9. 678 (49) Hofmann, N.; Mwingira, F.; Shekalaghe, S.; Robinson, L. J.; Mueller, I.; Felger, I. Ultra-Sensitive Detection 679 of Plasmodium Falciparum by Amplification of Multi-Copy Subtelomeric Targets. PLOS Med. 2015, 12 (3), 680 e1001788. https://doi.org/10.1371/journal.pmed.1001788. 681 (50) WHO. Public Report for Bioline Malaria Ag P.f (PQDx 0031-012-01) | WHO - Prequalification of Medical 682 Products (IVDs, Medicines, Vaccines and Immunization Devices, Vector Control). 683 https://extranet.who.int/prequal/WHOPR/public-report-bioline-malaria-ag-pf-pqdx-0031-012-01 684 (accessed 2024-09-25). 685 (51) WHO. Public Report for AdvDx Malaria Pf Rapid Malaria Ag Detection Test, (PQDx 0345-101-00) | WHO - 686 Prequalification of Medical Products (IVDs, Medicines, Vaccines and Immunization Devices, Vector 687 Control). https://extranet.who.int/prequal/WHOPR/public-report-advdx-malaria-pf-rapid-malaria-ag-688 detection-test-pqdx-0345-101-00 (accessed 2024-09-25). 689 (52) Burd, E. M. Validation of Laboratory-Developed Molecular Assays for Infectious Diseases. Clin. Microbiol. 690 Rev. 2010, 23 (3), 550–576. https://doi.org/10.1128/CMR.00074-09. 691 692 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 14, 2025. ; https://doi.org/10.1101/2025.07.12.25331425doi: medRxiv preprint