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

Nikhil Ranadive; Simon Kunene; Sarah Darteh; Nyasatu Ntshalintshali; Nomcebo Nhlabathi; Nomcebo Dlamini; Stanley Chitundu; Manik Saini; Maxwell Murphy; Adam Soble; Alanna Schwartz; Bryan Greenhouse; Michelle S Hsiang

doi:10.1093/cid/cix131

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

Clin Infect Dis. 2017 May 1;64(9):1221-1227. doi: 10.1093/cid/cix131.

Authors

Affiliations

¹ Emory University School of Medicine, Atlanta, Georgia, USA.
² National Malaria Control Programme, Ministry of Health, Manzini, Swaziland.
³ Clinton Health Access Initiative, Mbabane, Swaziland.
⁴ Department of Medicine, University of California, San Francisco, CA, USA.
⁵ Malaria Elimination Initiative, Global Health Group, University of California San Francisco (UCSF), San Francisco, CA, USA.
⁶ Department of Pediatrics, Benioff Children's Hospital and University of California, San Francisco, CA, USA.
⁷ Department of Pediatrics, University of Texas Southwestern Medical Center, Dallas, USA.

Abstract

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.

Keywords: Rapid Diagnostic Test; diagnostic accuracy; low transmission; malaria; subpatent infection.

Publication types

Evaluation Study

MeSH terms

Chromatography, Affinity / methods*
Diagnostic Tests, Routine / methods*
Eswatini
Humans
Malaria, Falciparum / diagnosis*
Plasmodium falciparum
Predictive Value of Tests
Prospective Studies
Sensitivity and Specificity