Background
The Republic of Djibouti is a malaria endemic country that was in pre-elimination phase in 2006-2012. From 2013, however, malaria has re-emerged in the country, and its prevalence has been increasing every year. Given the co-circulation of several infectious agents in the country, the assessment of malaria infection based on microscopy or histidine-rich protein 2 (HRP2)-based rapid diagnostic tests (RDT) has shown its limitations. This study, therefore, aimed to assess the prevalence of malaria among febrile patients in Djibouti city using more robust molecular tools.
Conclusions
The present study confirmed the high prevalence of falciparum malaria and, to a lesser extent, vivax malaria. Nevertheless, 29% of suspected malaria cases were misdiagnosed by microscopy and/or RDT. There is a need to strengthen the capacity for diagnosis by microscopy and to evaluate the possible role of P. falciparum hrp2 gene deletion, which leads to false negative cases of P. falciparum.
Methods
All suspected malaria cases reported to be microscopy-positive were randomly sampled (n = 1113) and included in four health structures in Djibouti city over a 4-year period (2018-2021), mainly during the malaria transmission season (January-May). Socio-demographic information was collected, and RDT was performed in most of the included patients. The diagnosis was confirmed by species-specific nested polymerase chain reaction (PCR). Data were analysed using Fisher's exact test and kappa statistics.
Results
In total, 1113 patients with suspected malaria and available blood samples were included. PCR confirmed that 788/1113 (70.8%) were positive for malaria. Among PCR-positive samples, 656 (83.2%) were due to Plasmodium falciparum, 88 (11.2%) Plasmodium vivax, and 44 (5.6%) P. falciparum/P. vivax mixed infections. In 2020, P. falciparum infections were confirmed by PCR in 50% (144/288) of negative RDTs. After the change of RDT in 2021, this percentage decreased to 17%. False negative RDT results were found more frequently (P < 0.05) in four districts of Djibouti city (Balbala, Quartier 7, Quartier 6, and Arhiba). Malaria occurred less frequently in regular bed net users than in non-users (odds ratio [OR]: 0.62, 95% confidence interval [CI]: 0.42-0.92). Conclusions: The present study confirmed the high prevalence of falciparum malaria and, to a lesser extent, vivax malaria. Nevertheless, 29% of suspected malaria cases were misdiagnosed by microscopy and/or RDT. There is a need to strengthen the capacity for diagnosis by microscopy and to evaluate the possible role of P. falciparum hrp2 gene deletion, which leads to false negative cases of P. falciparum.