Field evaluation of a PfHRP-2/pLDH rapid diagnostic test and light microscopy for diagnosis and screening of falciparum malaria during the peak seasonal transmission in an endemic area in Yemen

Lina M. Q. Alareqi, Mohammed A. K. Mahdy, Yee-Ling Lau, Mun-Yik Fong, Rashad Abdul-Ghani, Arwa A. Ali, Fei-Wen Cheong, Rehab Tawfek, Rohela Mahmud

Malaria is a public health threat in Yemen, with 149,451 cases being reported in 2013. Of these, Plasmodium falciparum represents 99 %. Prompt diagnosis by light microscopy (LM) and rapid diagnostic tests (RTDs) is a key element in the national strategy of malaria control. The heterogeneous epidemiology of malaria in the country necessitates the field evaluation of the current diagnostic strategies, especially RDTs. Thus, the present study aimed to evaluate LM and an RDT, combining both P. falciparum histidine-rich protein-2 (PfHRP-2) and Plasmodium lactate dehydrogenase (pLDH), for falciparum malaria diagnosis and survey in a malaria-endemic area during the transmission season against nested polymerase chain reaction (PCR) as the reference method. A household-based, cross-sectional malaria survey was conducted in Mawza District, a malaria-endemic area in Taiz governorate. A total of 488 participants were screened using LM and PfHRP-2/pLDH RDT. Positive samples (160) and randomly selected negative samples (52) by both RDT and LM were further analysed using 18S rRNA-based nested PCR. The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the RDT were 96.0 % (95 % confidence interval (CI): 90.9-98.3), 56.0 % (95 % CI: 44.7-66.8), 76.3 % (95 % CI: 69.0-82.3), and 90.4 % (95 % CI: 78.8-96.8), respectively. On the other hand, LM showed sensitivity of 37.6 % (95 % CI: 29.6-46.3), specificity of 97.6 % (95 % CI: 91.7-99.7), PPV of 95.9 % (95 % CI: 86.3-98.9), and NPV of 51.3 % (95 % CI: 43.2-59.2). The sensitivity of LM dropped to 8.5 % for detecting asymptomatic malaria. Malaria prevalence was 32.8 % (32.1 and 37.5 % for ≥10 and <10 years, respectively) with the RDT compared with 10.7 % (10.8 and 9.4 % for age groups of ≥10 and <10 years, respectively) with LM. Among asymptomatic malaria individuals, LM and RDT-based prevalence rates were 1.6 and 25.6 %, respectively. However, rates of 88.2 and 94.1 % of infection with P. falciparum were found among patients who reported fever in the 48 h prior to the survey by LM and PfHRP-2/pLDH RDT, respectively. The PfHRP-2/pLDH RDT shows high sensitivity for the survey of falciparum malaria even for asymptomatic malaria cases. Although the RDT had high sensitivity, its high false-positivity rate limits its utility as a single diagnostic tool for clinical diagnosis of malaria. On the other hand, low sensitivity of LM indicates that a high proportion of malaria cases is missed, underestimating the true prevalence of malaria in the community. Higher NPV of PfHRP-2/pLDH RDT than LM can give a straightforward exclusion of malaria among febrile patients, helping to avoid unnecessary presumptive treatments.