Meta-Analysis: Accuracy of Rapid Tests for Malaria in Travelers Returning from Endemic Areas

17 May 2005 | Volume 142 Issue 10 | Pages 836-846

Background: Microscopic diagnosis of malaria is unreliable outside specialized centers. Rapid tests have become available in recent years, but their accuracy has not been assessed systematically.

Purpose: To determine the accuracy of rapid diagnostic tests for ruling out malaria in nonimmune travelers returning from malaria-endemic areas.

Data Sources: The authors searched MEDLINE, EMBASE, CAB Health, and CINAHL (1988 to September 2004); hand-searched conference proceedings; checked reference lists; and contacted experts and manufacturers.

Study Selection: Diagnostic accuracy studies in nonimmune individuals with suspected malaria were included if they compared rapid tests with expert microscopic examination or polymerase chain reaction tests.

Data Extraction: Data on study and patient characteristics and results were extracted in duplicate. The main outcome was the likelihood ratio for a negative test result (negative likelihood ratio) for Plasmodium falciparum malaria. Likelihood ratios were combined by using random-effects meta-analysis, stratified by the antigen targeted (histidine-rich protein-2 [HRP-2] or parasite lactate dehydrogenase [LDH]) and by test generation. Nomograms of post-test probabilities were constructed.

Data Synthesis: The authors included 21 studies and 5747 individuals. For P. falciparum, HRP-2–based tests were more accurate than parasite LDH–based tests: Negative likelihood ratios were 0.08 and 0.13, respectively (P = 0.019 for difference). Three-band HRP-2 tests had similar negative likelihood ratios but higher positive likelihood ratios compared with 2-band tests (34.7 vs. 98.5; P = 0.003). For P. vivax, negative likelihood ratios tended to be closer to 1.0 for HRP-2–based tests than for parasite LDH–based tests (0.24 vs. 0.13; P = 0.22), but analyses were based on a few heterogeneous studies. Negative likelihood ratios for the diagnosis of P. malariae or P. ovale were close to 1.0 for both types of tests. In febrile travelers returning from sub-Saharan Africa, the typical probability of P. falciparum malaria is estimated at 1.1% (95% CI, 0.6% to 1.9%) after a negative 3-band HRP-2 test result and 97% (CI, 92% to 99%) after a positive test result.

Limitations: Few studies evaluated 3-band HRP-2 tests. The evidence is also limited for species other than P. falciparum because of the few available studies and their more heterogeneous results. Further studies are needed to determine whether the use of rapid diagnostic tests improves outcomes in returning travelers with suspected malaria.

Conclusions: Rapid malaria tests may be a useful diagnostic adjunct to microscopy in centers without major expertise in tropical medicine. Initial decisions on treatment initiation and choice of antimalarial drugs can be based on travel history and post-test probabilities after rapid testing. Expert microscopy is still required for species identification and confirmation.

Editors' Notes Context What is the accuracy of rapid tests for detecting malaria? Contribution This meta-analysis of 21 studies in nonimmune travelers with suspected malaria summarizes the accuracy of various rapid tests compared with expert microscopic examination or polymerase chain reaction tests. Histidine-rich protein-2 (HRP-2) tests had high sensitivities (range, 88% to 99%) and specificities (range, 95% to 100%) for detecting Plasmodium falciparum malaria. Positive 3-band HRP-2 test results detected malaria better than 2-band HRP-2 tests. Negative HRP-2–based test results excluded malaria better than parasite lactate dehydrogenase–based tests. Cautions Few studies evaluated 3-band HRP-2 tests for malaria species other than P. falciparum. –The Editors

 

Author and Article Information

From University of Berne and MediX General Practice Network, Berne, Switzerland; MRC Health Services Research Collaboration, University of Bristol, Bristol, United Kingdom; and Basel Institute for Clinical Epidemiology and Swiss Tropical Institute, University of Basel, Basel, Switzerland.

Acknowledgments: The authors thank Faïza Ajana, Martin O. Bubb, Peter L. Chiodini, Robert Gasser, Martin Grobusch, Eric Hernandez, Tomas Jelinek, Kevin Kain, Diana Lockwood, Carol Palmer, François Peyron, E. Geoffrey Playford, Joachim Richter, and Jose Miguel Rubio for additional information on their diagnostic accuracy studies; Tatiana Spicher for translations; Manuel Bichsel and Aijing Shang for statistical support; and Lucas M. Bachmann, Markus Battaglia, Christoph Minder, Patricia Schlagenhauf, and Marcel Zwahlen for helpful comments.

Grant Support: In part by the Swiss Federal Office of Public Health (grant 03.000048). Dr. Jüni was supported by grants from the Swiss National Science Foundation (grants 3233-066377 and 3200-066378).

Potential Financial Conflicts of Interest: None disclosed.

Requests for Single Reprints: Peter Jüni, MD, Department of Social and Preventive Medicine, University of Berne, Finkenhubelweg 11, CH-3012 Bern, Switzerland; e-mail, juni{at}ispm.unibe.ch.

Current Author Addresses: Dr. Marx: Department of General Internal Medicine, Inselspital, University of Berne, 3010 Berne, Switzerland.

Dr. Pewsner: MediX General Practice Network, Practice Brückfeld, Bonstettenstrasse 1, 3012 Berne, Switzerland.

Dr. Egger: MRC Health Services Research Collaboration, Department of Social Medicine, University of Bristol, Canynge Hall, Whiteladies Road, Bristol BS8 2PR, United Kingdom.

Dr. Nüesch: Outpatient Department of Internal Medicine, University Hospital Basel, Hebelstrasse 32, 4031 Basel, Switzerland.

Dr. Bucher: Basel Institute for Clinical Epidemiology, University Hospital Basel, Hebelstasse 10, 4031 Basel, Switzerland.

Drs. Genton and Hatz: Swiss Tropical Institute, University of Basel, Socinstrasse 57, 4002 Basel, Switzerland.

Dr. Jüni: Department of Social and Preventive Medicine, University of Berne, Finkenhubelweg 11, CH-3012 Bern, Switzerland.