Limitations of Rapid Diagnostic Testing in Patients with Suspected Malaria: A Diagnostic Accuracy Evaluation from Swaziland, a Low-Endemicity Country Aiming for Malaria Elimination

• Published in: Clinical infectious diseases Vol. 64; no. 9; pp. 1221 - 1227
• Main Authors: Ranadive, Nikhil, Kunene, Simon, Darteh, Sarah, Ntshalintshali, Nyasatu, Nhlabathi, Nomcebo, Dlamini, Nomcebo, Chitundu, Stanley, Saini, Manik, Murphy, Maxwell, Soble, Adam, Schwartz, Alanna, Greenhouse, Bryan, Hsiang, Michelle S.
• Format: Journal Article
• Language: English
• Published: United States Oxford University Press 01.05.2017
• Subjects: Accuracy; ARTICLES AND COMMENTARIES; Chromatography, Affinity - methods; Cytochrome b; Demographics; Demography; Deoxyribonucleic acid; Diagnostic systems; Diagnostic tests; Diagnostic Tests, Routine - methods; DNA; Epidemiology; Health care facilities; Histidine; Humans; Major; Malaria; Malaria, Falciparum - diagnosis; Parasites; Patients; Plasmodium falciparum; Polymerase chain reaction; Predictive Value of Tests; Prospective Studies; Regression analysis; Sensitivity; Sensitivity and Specificity; Swaziland; Vector-borne diseases; Switzerland; diagnostic accuracy; malaria; subpatent infection; Rapid Diagnostic Test; low transmission
• ISSN: 1058-4838; 1537-6591; 1537-6591
• DOI: 10.1093/cid/cix131

Background. The performance of Plasmodium falciparum–specific histidine-rich protein 2–based rapid diagnostic tests (RDTs) to evaluate suspected malaria in low-endemicity settings has not been well characterized. Methods. Using dried blood spot samples from patients with suspected malaria at 37 health facilities from 2012 to 2014 in the low-endemicity country of Swaziland, we investigated the diagnostic accuracy of histidine-rich protein 2-based RDTs using qualitative polymerase chain reaction (PCR) (nested PCR targeting the cytochrome b gene) and quantitative PCR as reference standards. To explore reasons for false-negative and/or false-positive results, we used pfhrp2/3-specific PCR and logistic regression analyses of potentially associated epidemiological factors. Results. From 1353 patients, 93.0% of RDT-positive (n = 185) and 31.2% of RDT-negative samples (n = 340) were available and selected for testing. Compared with nested PCR, the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of RDTs were 51.7%, 94.1%, 67.3%, and 89.1%, respectively. After exclusion of samples with parasite densities <100/μL, which accounted for 75.7% of false-negative results and 33.3% of PCR-detectable infections, the sensitivity, specificity, PPV, and NPV were 78.8%, 93.7%, 62.3%, and 97.1%. Deletions of pfhrp2 were not detected. False-positivity was more likely during the second year and was not associated with demographics, recent malaria, health facility testing characteristics, or potential DNA degradation. Conclusions. In the low-transmission setting of Swaziland, we demonstrated low sensitivity of RDT for malaria diagnosis, owing to an unexpectedly high proportion of low-density infection among symptomatic subjects. The PPV was also low, requiring further investigation. A more accurate point-of-care diagnostic may be needed to support malaria elimination efforts.