PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
Clin Infect Dis. Author manuscript; available in PMC 2010 September 1.
Published in final edited form as:
PMCID: PMC2912215
NIHMSID: NIHMS149124

Superior Diagnostic Performance of Malaria Rapid Diagnostic Tests as compared to Blood Smears in U.S. Clinical Practice

Abstract

Background

Approximately 4 million U.S. travelers to developing countries are ill enough to seek healthcare with 1,500 malaria cases reported in the U.S. annually. The diagnosis of malaria is frequently delayed due to the time to prepare malaria blood films and lack of technical expertise. An easy, reliable rapid diagnostic test (RDT) with high sensitivity and negative predictive value (NPV), particularly for Plasmodium falciparum, would be clinically useful. The study objective was to determine the diagnostic performance of the FDA-approved NOW® Malaria Test in comparison to traditional thick and thin blood smears for malaria diagnosis.

Methods

This prospective study tested 852 consecutive blood samples sent for thick and thin smears with blinded, malaria rapid tests at three hospital laboratories during 2003–2006. Polymerase chain reaction (PCR) verified positive tests and discordant results.

Results

Malaria occurred in 11% (95/852). The rapid test had superior performance than the standard Giemsa thick blood smear (P=.003). The rapid test’s sensitivity for all malaria was 97% (92/95) vs. 85% (81/95) by blood smear, and the RDT had superior NPV of 99.6% vs. 98.2% (P=.001). The P. falciparum performance was excellent with 100% rapid test sensitivity versus only 88% (65/74) by blood smear (P=.003).

Conclusions

This operational study demonstrates the FDA-approved rapid malaria test is superior to a single set of blood smears performed under routine U.S. clinical laboratory conditions. The most valuable clinical role of the RDT is in the rapid diagnosis or the exclusion of P. falciparum malaria, which is particularly useful in outpatient settings when evaluating febrile travelers.

Keywords: malaria, diagnosis, sensitivity and specificity

Introduction

An estimated 4 million returning U.S. travelers are ill enough to seek healthcare annually with more than 1,500 having malaria reported in the United States. [1,2,3] Of travelers who contract malaria, 90–95% will not become ill until after they return home, with 85% developing symptomatic disease within 30 days of return. [4,5] As malaria is unusual in the U.S., patients frequently present to clinicians without tropical medicine expertise and to primary health care facilities lacking expert diagnostic capabilities. Kain et al. found 59% of malaria cases were initially misdiagnosed in non-endemic North American settings.[4] In addition, they found 64% of community-based microscopic diagnosis provided incorrect species identification. [4] Further, Kain reported the average time from ordering a blood smear until the laboratory diagnosis of malaria was 2.5 days, partially responsible for an average time from symptom onset to malaria diagnosis of 7.6 days for Plasmodium falciparum. [4] Even where malaria is more frequently encountered, delay is common due to the necessary time for formal Giemsa-stained thick blood smear preparation and reading (~6–8 hours).

Of those developing P. falciparum in the U.S., ~1% will die. [4,5] Yet, an estimated 80% of deaths are preventable with a significant proportion due to diagnostic delay or error. [6] An easily performed test that is sensitive and reliable would be a desirable tool that could augment malaria diagnosis. A rapid test with high sensitivity and negative predictive value (NPV) for P. falciparum would be of particular use in the acute care setting where the decision for hospitalization is being made. Lastly, it also may benefit severely ill patients by confirming or excluding a malaria diagnosis rapidly and facilitating prompt intervention.

The rapid antigen capture assay, NOW® Malaria Test (Binax, Inc., Inverness Medical Professional Diagnostics, Scarborough, ME, USA), received U.S. Food and Drug Administration approval in June 2007 for diagnosis of symptomatic malaria. Antigens detected by monoclonal antibodies in this rapid antigen capture assay include Histidine-Rich Protein (HRP-2) for specific identification of P. falciparum and Plasmodium aldolase for identification of all malaria species. Reasonable sensitivity and specificity has been reported in clinical research trials outside the U.S., mainly in endemic areas where malaria dynamics and environmental conditions differ greatly from those in the U.S. [7,8] Few studies have investigated rapid diagnostic tests (RDT) in non-endemic settings, and no study has evaluated the operational use of the current FDA-approved version of the test in the U.S. This study investigated the diagnostic performance of a single blood smear versus the NOW® Malaria rapid diagnostic test (RDT) in detecting symptomatic malaria to determine the strengths and limitations of this test in “real world” U.S. clinical use.

Methods

Setting & Participants

This prospective operational laboratory study was conducted at the three hospital-based laboratories with the highest incidence of malaria in Minnesota from March 1, 2003 through February 28, 2006. All consecutive samples sent to the laboratories for malaria thick and thin blood smears were included. Participants were returning travelers, of whom significant proportions were visiting friends and relatives (VFRs).

Methods of Measurement

Thick and thin Giemsa-stain blood smears were prepared from EDTA-anticoagulated venous blood. Smears were examined for the presence of malarial parasites by trained technologists or hematopathologists per individual laboratory protocol with a minimum of 500 high-powered fields examined. An expert pathologist or hematopathologist confirmed all positive smears. The NOW® Malaria Test was performed per the manufacturer’s instructions by technologists masked to the blood smear results. The NOW® Malaria Test is a lateral flow immunochromatographic test, similar to a pregnancy test with a visual result in < 15 minutes.

A nested case-control study verified all positive and discrepant results by PCR in a separate laboratory, as previously described. [9] For each positive sample, two samples negative by both blood smear and rapid test served as masked, negative controls for PCR. Finally, unusual or discrepant PCR results were tested at separate laboratory by a second PCR method utilizing different primers [10] and using a second, independent DNA extraction to prevent contamination. For verifying discrepant results between RDT and blood smear, PCR served as the reference standard for validation. All personnel were masked to other test results, and no person performed more than one test type.

Outcome Measures & Data Analysis

The primary outcome measure was the test performance of the RDT and traditional blood smear. Specifically, we were primarily interested in the sensitivity for malaria diagnosis and the NPV to exclude malaria of each diagnostic modality. Differences between the diagnostic performances were analyzed by using the two-tailed McNemar’s test to compare the paired, nominal data. The STARD checklist for the reporting of studies of diagnostic accuracy was utilized. [11] This study was approved as exempt by the institutional review boards of all five institutions involved. The only patient data collected were de-identified data provided to the laboratory upon thick/thin smear order as would occur in routine clinical practice.

Results

There were 103 individual specimens that tested positive for malaria by either blood smear or RDT (12%) with 95 confirmed by PCR (11%) among 852 specimens tested (Figure 1). Patients with malaria were 56% male and 44% female with a mean ±SD age of 33 ±1.9 years (range: 18 months to 67 years).

Figure 1
Performance of Malaria Diagnostic Tests

Overall, in all species of malaria, the RDT was superior to traditional blood smear examination (Table 1) with better sensitivity 97% [92/95] vs. traditional Giemsa blood smear 85% [81/95] (P=.003) and better NPV of 99.6% vs. 98.2% (P=.001). The largest discrepancy in performance was for the diagnosis of P. falciparum with 100% RDT sensitivity (74/74) compared to only 88% (65/74) sensitivity for blood smear (P=.003).

Table 1
Test performance characteristics of Rapid Diagnostic Test vs. blood smear for malaria diagnosis.

For non-falciparum malaria (P. vivax, P. ovale, and P. malariae), the sensitivity of the RDT was less at 86% [18/21] though comparable to smear 76% [16/21] (P=.63). However, even with non-falciparum the RDT had a high NPV (>98%). In addition, although there were few P. ovale cases (n=5) both the RDT and blood smear had poor sensitivity (60%). One noteworthy anomaly was RDT false positives which occurred in 1% [8/757] of specimens negative by smear and PCR; however in 7 of the 8 cases recent, documented anti-malaria treatment had been given. As the RDT is an antigen-based test, antigen may persist in the blood long beyond viable Plasmodium parasites. An example of the range of RDT tests results for P. falciparum and non-P. falciparum malaria is provided [Appendix 1].

Specificity by thin blood smear was excellent (100%) with correct malaria speciation in 84% [68/81] as referenced by PCR. However, the inability by blood smear to identify the species occurred in 14% [10/74] of P. falciparum, two mixed P. ovale/P. malariae, and one misidentification of P. malariae as P. ovale. The larger clinical concern was the 12% [9/74] of P. falciparum cases read as negative on blood smear but positive by both RDT and PCR.

PCR was used to validate all positive samples confirming malaria species of: P. falciparum (n=74), P. vivax (n=8), P. ovale (n=5), P. malariae (n=6), and mixed P. malariae/P. ovale (n=2). Percent parasitemia ranged from < 0.1% to 18%. Malaria cases were exposed in multiple regions which were Africa (93%), Asia/Pacific (6%), and Latin America (1%). No malaria occurred in newly arrived refugees. This was likely due to institution of pre-departure, presumptive anti-malaria therapy for migrating refugees by the CDC in 1999. [12]

Discussion

Malaria is the leading infectious cause of mortality among travelers, yet the vast majority of returning febrile travelers will not have malaria. [2] Thus, the diagnosis is frequently delayed or missed in non-endemic regions. It is likely a sensitive, reliable, and rapid diagnostic test would improve malaria diagnosis and management in the U.S. This study compared the NOW® Malaria Test with the current clinical practice of Giemsa-stained blood smears in returning travelers. This was an operational study to compare the performance of each test in a typical U.S. setting to assess the utility of the RDT in clinical practice. Thus, we believe the results are generalizable to other U.S. acute care settings.

There are many rapid tests available on the global market, and although they have shown variable results in the past, the newer generation assays have been refined, particularly for acute P. falciparum infections. [7,8] Since P. falciparum malaria is the most common life-threatening infection in travelers, and infected persons may deteriorate rapidly, [2,13] outpatient clinicians must be able to reasonably exclude P. falciparum prior to allowing outpatient management in non-immune patients. Therefore a rapid test with a high NPV would be of particular use in the outpatient clinical setting when a provider must decide in a timely manner if a non-toxic appearing, non-immune patient with exposure to P. falciparum should be hospitalized. Such a test with a high NPV may fundamentally alter the current approach to managing the outpatient febrile returning traveler.

A pivotal clinical trial evaluating the performance of the NOW® Malaria Test in endemic areas among >4,000 patients was recently presented at the American Society of Tropical Medicine and Hygiene annual conference. [14] For P. falciparum, the sensitivity was 95% and specificity 94% overall. In those with higher parasitemia (>5,000 parasites/μL), the sensitivity was 99.7%. Travelers, presenting in non-endemic regions to acute care settings, have relatively high malaria parasitemia, typically on the order of >10,000 parasites/μL (0.2% parasitemia). [15] Thus, the slightly better RDT performance in our study is most likely due to two facts. First, the study population consisted of non-immune, returning U.S. travelers who typically have higher parasitemia levels than other studies’ populations in malaria endemic areas. Second, we compared the RDT to single blood smears performed under “real world” conditions (rather than the gold standard three tests read by expert malariologists).

Few other studies evaluating this newer generation RDT assay have been presented or published. Farcas et al. in a laboratory research protocol evaluated 256 ill Canadian travelers and found comparable results with a sensitivity of 95.5% for P. falciparum, 87% for P. vivax, and 83.5% for all non-falciparum malaria. [16] In another study of returning French travelers, the sensitivities were 98.8% for P. falciparum and 80% for P. vivax. [17] An Italian study of 145 persons with malaria from among 171 immigrants and 136 returning travelers, 100% P. falciparum sensitivity was reported. [18] All studies have reported excellent specificity (>98%) for all species. [1418]

In our study, there were false positives for P. falciparum detected in eight RDT specimens, of whom seven had known recent treatment. These false positives are usually due to the persistence of the HRP-2 antigen after treatment, although rheumatoid factor can also yield false positives due to binding IgG. [19]

There are some interesting results and observations in this study that should be mentioned. First, it should be noted that the RDT is particularly insensitive for detecting P. ovale infection (60%) which is consistent with other studies. [16,20] In fact, two of the five cases in this study were missed by both microscopy and RDT and were detected in negative samples sent for control purposes. These were independently confirmed by two distinct PCR techniques in separate laboratories. Another interesting case was a mixed infection with P. ovale and P. malariae, a very rare combination, on two separate samples from the same patient which was PCR confirmed by both laboratories.

Clinical Implications

We believe that at point of care settings the NOW® Malaria Test offers a rapid and accurate test for reasonably excluding P. falciparum (high negative predictive value) and a valuable adjunctive test in rapidly confirming the diagnosis of both P. falciparum and non-falciparum malaria. An ill patient who tests positive for P. falciparum should prompt immediate therapy with either intravenous quinidine for severe cases or atovaquone-proguanil (Malarone), artemether-lumefantrine (Coartem), mefloquine (Lariam), quinine/doxycycline, or quinine/clindamycin for non-severe cases. In severe cases if quinidine is unavailable, artesunate may be acquired from the CDC. For P. vivax and P. ovale malaria, typically chloroquine is the drug of choice, although there are increasing areas of resistance, and primaquine therapy is required after G6PD (glucose-6-phosphate dehydrogenase) testing. Refer to www.cdc.gov/malaria for current treatment recommendations. Follow up consultation with a physician knowledgeable in tropical medicine should occur; however, initiation of therapy should not be delayed. When in doubt, treat for the worst case scenario of chloroquine-resistant P. falciparum pending further testing. Persons returning from Southeast Asia, particularly forested regions of Malaysia, may be at risk of P. knowlesi, a primate malaria where humans act as an accidental host. There are no published data on the performance of RDT for the detection of P. knowlesi, and traditional smears should be done in all suspect cases.

We believe that traditional blood smear testing remains valuable, particularly for estimating level of parasitemia and malaria speciation. In most cases blood for thick and thin smears should be collected simultaneously with performance of RDT. The package insert/FDA states that the RDT, when negative, must be confirmed by thick/thin blood smear. We believe this guidance is particularly valuable when non-falciparum malaria and mixed infections are being considered. Clinical judgment must be exercised. Even if a RDT is negative, in the setting of high pre-test probability, patient care should be conservative. Similarly, regardless of test results, all toxic appearing returned travelers require hospitalization and appropriate assistance from trained tropical medicine and/or infectious disease specialists.

Limitations

This study was designed to evaluate the performance of the NOW® Malaria Test during routine practice in U.S. hospital and clinic settings. In keeping with the operational design, although three thick and thin blood smears are recommended to exclude malaria, in reality clinical decisions need to be made in a timely manner in the acute care setting. Therefore, the analysis was performed comparing each specimen separately rather than by case. Due to current U.S. HIPAA regulations, as a laboratory based study, no personal identifiable data were collected and comparison by case was not possible. The smears were performed under normal operational circumstances in each laboratory setting. Therefore, this study reflects the RDT’s use in actual U.S. clinical practice and should be broadly generalizable to other U.S. institutions. Further, due to cost constraints, not all negative samples were tested by PCR. Conceivably, some additional low level malaria infections may have been missed by both microscopy and RDT which would decrease both tests overall sensitivity. However, while this is a possibility, even if present, this would not alter the results as it would decrease the sensitivity of both testing techniques equally.

Caution exists with any negative malaria test, either by RDT or blood smear. A negative test result does not fully exclude infection with malaria. This is particularly true at low levels of parasitemia and with non-falciparum. Subpatent malaria infection (which is not yet detectable) is possible with early testing soon after the onset of symptoms. In Canada, those diagnosed with malaria presented to health care after an average of 3.5 days of symptoms with initial thick blood smears being diagnostic in 99%, but not all. [4] In febrile travelers recently returned from a malaria endemic region with unexplained febrile illness, clinicians should arrange clinical follow up within 24 hours and consider repeat malaria testing.

Conclusions

This operative study comparing traditional blood smear to rapid antigen capture test demonstrates that the RDT, as performed in a routine clinical setting, is superior to a single set of Giemsa-stained blood smears for quickly evaluating a patient for malaria. Importantly, the NOW® Malaria test had a 100% NPV for P. falciparum malaria. While further clinical experience with this test is necessary, this FDA-approved test appears to be an ideal tool for timely decision making when P. falciparum must be considered for patient disposition. This is especially true when reliable, experienced microscopy is not immediately available, such as after-hours or in small or rural settings without expertise in reading malaria blood smears. While we believe blood smears and expert microscopy remain essential, the overall RDT performance provides clinicians with a valuable adjunctive diagnostic tool in the timely evaluation of the ill returning traveler.

Summary: The NOW® Malaria rapid diagnostic test had superior performance compared to malaria blood smear under typical U.S. laboratory settings for detecting P. falciparum malaria with high negative predictive value assisting in clinical decision making in acute care settings.

Supplementary Material

Appendix1

Acknowledgments

We wish to thank Aaron Devries MD, MPH for assistance with laboratory testing, David Neitzel for provision of Minnesota malaria statistics. WMS, DRB received support from the NIH NIAID (T32AI055433-04; L30AI066779-03; K12RR023247-04).

Footnotes

Contributions: WMS conceived the study. CCP, DO, BAJ, CMT, SHB, KLH supervised the conduct of the study and data collection. DRB conducted the statistical analysis. WMS and DRB drafted the manuscript, and all authors contributed to the final version. WMS and DRB take responsibility for the paper as a whole.

Financial Disclosure: Binax, Inc. (now, Inverness Medical Professional Diagnostics) supplied the NOW® Malaria tests but no financial support. They were not involved in study design, data analysis, or manuscript preparation.

Contributor Information

William M. Stauffer, University of Minnesota, Department of Medicine, Infectious Diseases and International Medicine, Department of Pediatrics, Infectious Diseases, School of Public Health, Community Health and Epidemiology, 420 Delaware Street S.E., Mayo D407, MMC 250, Minneapolis, Minnesota 55455, Phone 612-624-9996, Fax 612-625-4410.

Charles P. Cartwright, Departments of Laboratory Medicine and Pathology, Hennepin County Medical Center and University of Minnesota Medical School, Clinical Laboratories P4, Hennepin County Medical Center, 701 Park Avenue, Minneapolis, MN 55415.

Douglas Olson, Department of Pathology, Microbiology Laboratory, HeathPartners’/Regions Hospital, 640 Jackson Street, St. Paul, MN 55101.

Billie Anne Juni, Public Health Laboratory Division, Minnesota Department of Health, MDH/MDA Laboratory Building, 601 North Robert Street, P.O. Box 64899, St. Paul, MN 55164-0899.

Charlotte M Taylor, Public Health Laboratory Division, Minnesota Department of Health, MDH/MDA Laboratory Building, 601 North Robert Street, P.O. Box 64899, St. Paul, MN 55164-0899.

Susan H. Bowers, Department of Pathology, Park Nicollet, Methodist Hospital, 6500 Excelsior Blvd, St. Louis Park, MN 55426, Ph: 952 993-5290.

Kevan L. Hanson, Hennepin County Medical Center, 701 Park Ave So. Minneapolis, Minnesota, 55415, Ph:612-873-3037.

Jon E. Rosenblatt, Division of Clinical Microbiology, Mayo Clinic, Rochester MN 55906, Ph: 507-284-3050.

David R. Boulware, University of Minnesota, Department of Medicine, Infectious Diseases and International Medicine, MMC 250, 420 Delaware Street S.E. Minneapolis, Minnesota 55455, Ph: 612-624-9996.

References

1. World Health Organization (WHO) The global malaria situation: current tools for prevention and control. 55th World Health Assembly. Global Fund to Fight AIDS, Tuberculosis and Malaria. WHO document no. A55/INF.DOC./6 Available at: http://www.who.int/topics/malaria/en.
2. Freedman DO, Weld LH, Kozarsky PE, et al. Spectrum of disease and relation to place of exposure among ill returned travelers. N Engl J Med. 2006;354:119–130. [PubMed]
3. Mali S, Steele S, Slustker, Argin PM. Malaria surveillance - United States, 2006. Morb Mort Wkly Rep Surveill Summ. 2008;57 (SS05):24–39.
4. Kain KC, Harrington MA, Tennyson S, Keystone JS. Imported malaria: prospective analysis of problems in diagnosis and management. Clin Infect Dis. 1998;27:142–149. [PubMed]
5. Ryan ET, Wilson ME, Kain KC. Illness after international travel. N Engl J Med. 2002;347:505–516. [PubMed]
6. Newman RD, Parise ME, Barber AM, Steketee RW. Malaria-related deaths among U.S. travelers, 1963–2001. Ann Int Med. 2004;141:547–55. [PubMed]
7. Wongsrichanalai C, Arevalo I, Laoboonchai A, Yingyuen K, Miller RS, Magill AJ, et al. Rapid diagnostic devices for malaria: field evaluation of a new prototype immunochromatographic assay for the detection of Plasmodium falciparum and non-falciparum Plasmodium. Am J Trop Med Hyg. 2003;69:26–30. [PubMed]
8. Marx A, Pewsner D, Egger M, Nüesch R, Bucher HC, Genton B, et al. Meta-Analysis: Accuracy of Rapid Tests for Malaria in Travelers Returning from Endemic Areas. Ann Intern Med. 2005;142:836–847. [PubMed]
9. Snounou G, Viriyakosol S, Zhu XP. High sensitivity detection of human malaria parasites by the use of nested polymerase chain reaction. Mol Biochem Parasitol. 1993;61:315–320. [PubMed]
10. Swan H, Sloan L, Muyombwe A, Chavalitshewinkoon-Petmitr P, Krudsood S, Leowattana W, et al. Evaluation of a real-time polymerase chain reaction assay for the diagnosis of malaria in patients from Thailand. Am J Trop Med Hyg. 2005;73:850–4. [PubMed]
11. Bossuyt PM, Reitsma JB, Bruns DE, Gatsonis CA, Glasziou PP, Irwig LM, et al. The STARD statement for reporting studies of diagnostic accuracy: explanation and elaboration. Ann Intern Med. 2003;138:W1–12. [PubMed]
12. Collinet-Adler S, Stauffer WM, Boulware DR, Larsen KL, Rogers TB, Williams DN. Financial implications of refugee malaria: the impact of pre-departure anti-malarial presumptive treatment. Am J Trop Med Hyg. 2007;77:458–63. [PMC free article] [PubMed]
13. Bottieau E, Clerinx J, Schrooten W, Van den Enden E, Wouters R, Van Esbroeck M, et al. Etiology and outcome of fever after a stay in the tropics. Arch Intern Med. 2006;166:1642–8. [PubMed]
14. Gasser RA, Jr, Magill AJ, Ruebush T, Miller RS, Sirichaisinthop J, Forney JR, et al. Malaria Diagnosis: Performance of NOW® ICT Malaria in a Large Scale Field Trial [Abstract #2338]. 54th Annual Meeting American Society of Tropical Medicine and Hygiene; Washington, DC. USA. Dec 11–15, 2005.
15. Murray CK, Gasser RA, Jr, Magill AJ, Miller RS. Update on rapid diagnostic testing for malaria. Clin Microbiol Rev. 2008;21:97–110. [PMC free article] [PubMed]
16. Farcas GA, Zhong KJ, Lovegrove FE, Graham CM, Kain KC. Evaluation of the Binax NOW ICT test versus polymerase chain reaction and microscopy for the detection of malaria in returned travelers. Am J Trop Med Hyg. 2003;69:589–92. [PubMed]
17. De Monbrison F, Gerome P, Chaulet JF, Wallon M, Picot S, Peyron F. Comparative diagnostic performance of two commercial rapid tests for malaria in a non-endemic area. Eur J Clin Microbiol Infect Dis. 2004;23:784–6. [PubMed]
18. Gatti S, Gramegna M, Bisoffi Z, et al. A comparison of three diagnostic techniques for malaria: a rapid diagnostic test (NOW Malaria), PCR and microscopy. Ann Trop Med Parasitol. 2007;101:195–204. [PubMed]
19. Iqbal J, Khalid N, Hira PR. Performance of rapid malaria Pf antigen test for the diagnosis of malaria and false-reactivity with autoantibodies. Adv Exp Med Biol. 2003;531:135–48. [PubMed]
20. Bigaillon C, Spiegel A. Ineffectiveness of the Binax NOW Malaria Test for diagnosis of Plasmodium ovale malaria. J Clin Micro. 2005;43:1011. [PMC free article] [PubMed]