Performance evaluation of different strategies based on microscopy techniques, rapid diagnostic test and molecular loop-mediated isothermal amplification assay for the diagnosis of imported malaria

Malaria is one of most common tropical diseases encountered in travellers and migrants. It requires an urgent and reliable diagnosis considering its potential severity. In this study, performance of five diagnostic assays were evaluated in a nonendemic region and compared prospectively to quantitative PCR (qPCR). A prospective study was conducted at Toulouse Hospital from August 2017 to January 2018 and included all patients with initial Plasmodium screening. Thin and thick blood smears (TnS, TkS), quantitative buffy coat (QBC), rapid diagnostic tests (RDTs) and commercial loop-mediated isothermal amplification (LAMP) were independently performed on each blood sample and compared to our qPCR reference standard. The study encompassed 331 patients, mainly returning from Africa. qPCR detected 73 Plasmodium-positive samples (including 58 falciparum). Individually, LAMP had a 97.3% (71/73) sensitivity, far ahead of TnS (84.9%, 62/73), TkS (86.3%, 63/73), QBC (86.3%, 63/73) and RDT (86.3%, 63/73). RDT demonstrated a high sensitivity for falciparum (98.3%, 57/58) but missed all ovale, malariae and knowlesi infections. Specificity was excellent for all techniques (99.6–100%). The most sensitive diagnosis strategies were TnS + RDT (95.9%, 70/73), TnS + LAMP (97.3%, 71/73) and TnS + RDT + LAMP (100%, 73/73), about 10% higher than strategies using exclusively microscopy, TkS + TnS (87.7%, 64/73) or QBC + TnS (87.7%, 64/73). TnS remains necessary for Plasmodium species identification and quantification. Adding sequentially TnS only on LAMP-positive samples did not decrease TnS + LAMP strategy sensitivity. In nonendemic countries, the currently recommended microscopy-based strategies seem unsatisfactory for malaria diagnosis considering RDT and LAMP performance, two rapid and sensitive assays that require limited training.