Research Article: A Prospective Hospital Study to Evaluate the Diagnostic Accuracy of Rapid Diagnostic Tests for the Early Detection of Leptospirosis in Laos

Date Published: April 26, 2018

Publisher: The American Society of Tropical Medicine and Hygiene

Author(s): Sabine Dittrich, Latsaniphone Boutthasavong, Dala Keokhamhoung, Weerawat Phuklia, Scott B. Craig, Suhella M. Tulsiani, Mary-Anne Burns, Steven L. Weier, David A. B. Dance, Viengmon Davong, Manivanh Vongsouvath, Mayfong Mayxay, Rattanaphone Phetsouvanh, Paul N. Newton, Kate Woods.


Leptospirosis is a globally important cause of acute febrile illness, and a common cause of non-malarial fever in Asia, Africa, and Latin America. Simple rapid diagnostic tests (RDTs) are needed to enable health-care workers, particularly in low resource settings, to diagnose leptospirosis early and give timely targeted treatment. This study compared four commercially available RDTs to detect human IgM against Leptospira spp. in a head-to-head prospective evaluation in Mahosot Hospital, Lao PDR. Patients with an acute febrile illness consistent with leptospirosis (N = 695) were included in the study during the 2014 rainy season. Samples were tested with four RDTs: (“Test-it” [Life Assay, Cape Town, South Africa; N = 418]; “Leptorapide” [Linnodee, Ballyclare, Northern Ireland; N = 492]; “Dual Path Platform” [DPP] [Chembio, Medford, NY; N = 530]; and “SD-IgM” [Standard Diagnostics, Yongin, South Korea; N = 481]). Diagnostic performance characteristics were calculated and compared with a composite reference standard combining polymerase chain reaction (PCR) (rrs), microscopic agglutination tests (MATs), and culture. Of all patients investigated, 39/695 (5.6%) were positive by culture, PCR, or MAT. The sensitivity and specificity of the RDTs ranged greatly from 17.9% to 63.6% and 62.1% to 96.8%, respectively. None of the investigated RDTs reached a sensitivity or specificity of > 90% for detecting Leptospira infections on admission. In conclusion, our investigation highlights the challenges associated with Leptospira diagnostics, particularly in populations with multiple exposures. These findings emphasize the need for extensive prospective evaluations in multiple endemic settings to establish the value of rapid tools for diagnosing fevers to allow targeting of antibiotics.

Partial Text

Leptospirosis is an important zoonotic disease worldwide, with its frequency and severity increasingly recognized.1,2 It has also been shown to be a significant cause of meningoencephalitis in Laos and Thailand.3 Leptospirosis is caused by Leptospira spp. spirochetes contracted by humans through exposure to environments contaminated by urine of infected mammals.2 It is estimated that ∼853,000 people are infected and 48,000 die annually.4 Most of the cases occur in the tropics, particularly in urban slums and rural areas where people are exposed to contaminated water.2 The clinical presentation of leptospirosis is often nonspecific, and as the organism does not grow well in conventional blood cultures, diagnosis is difficult, requiring sophisticated serological and molecular tests. However, vast areas of the tropics where leptospirosis is endemic have extremely limited diagnostic laboratory capacity.5 Even where the laboratory capacity exists, diagnosis using specific culture or serological microscopic agglutination test (MAT) methods2 requires considerable expertise that is not widely available, and results are only available weeks after the initial clinical presentation. At this point, no clear guidance by international bodies such as the World Health Organization (WHO) exists as to which test is recommended for acute detection. Conventionally, the observation of a 4-fold rise between the acute and convalescent sample is considered a clear indication of an acute infection and is therefore considered the gold standard; however, a recent modeling analysis has highlighted the pitfalls of this approach.6 Several manufacturers have developed rapid diagnostic tests (RDTs) for use at the bedside or point-of-care7 of which so far, none has been approved by a stringent regulatory authority. The simplicity and relatively low cost of these tests make them potentially well suited for use in resource-poor settings with limited laboratory and human capacity, as has been achieved with malaria RDTs. Evaluations of RDTs detecting IgM against Leptospira spp. antigens have been conducted, and their diagnostic characteristics have been reported to vary between areas of low and high endemicity.8 Goris et al.8 reported 69% sensitivity and 96% specificity for the LeptoTek lateral flow test when used on admission sera in a Dutch population, whereas the same test used in a Southeast Asian hospital setting (Lao PDR) had only 45% sensitivity and 75% specificity.9 These differences are very important, as a test may be well suited to one setting but not to another. It is likely that the differences, particularly for specificity, are mainly due to background antibody levels in patients who have had multiple exposures to the pathogen, similar to the challenges faced with Orientia tsutsugamushi (scrub typhus) diagnosis in endemic areas.10

Given the global environmental presence of Leptospira spp. and that they have been identified as an important cause of fever in many large non-malarial fever studies,14–16 a simple, rapid diagnostic tool for diagnosing leptospirosis could have a large impact on patient care globally. In this study, we evaluated four RDTs which all detect anti-Leptospira IgM. The “Test-it” and “SD-IgM” are designed as simple lateral flow tests, whereas the “Leptorapide” is an agglutination test and the “DPP” is a lateral flow test with a unique dual path (DPP) technology.8,17 Although the three cassette-based tests represent familiar, supposedly simple-to-interpret, platforms, there was considerable interobserver variability between the three readers in this study. This was less the case for the Leptorapide test, which is an agglutination test. It is conceivable that in some cases, a delay in reading results may have occurred between the three readers that could have contributed to the observed inter-reader variability due to fading/intensifying of bands over time. Although this observation might not be representative because of the very small sample size, it is important to follow our findings up with more research to support product improvement efforts. When using a composite reference standard as comparators for the diagnostic accuracy assessment, no clearly superior RDT could be identified. The DPP assay performed consistently regardless of the days of illness with a sensitivity between 50% and 60% and specificity around 70%, which is in line with what was previously published for mild leptospirosis cases at admission as well as healthy slum habitants.18 In comparison to previously published sensitivity and specificity of more than 90%,17 we found that the “Leptorapide” assay showed a lower sensitivity (< 50%) and specificity (∼80%) combined with an NPV of ∼95%. The “Test-it” assay had a high sensitivity of ∼80% in patients with less than 5 days of fever and the specificity of the test was low at ∼70%. Earlier evaluations3,6 of this assay reported a higher specificity, and the difference can likely be explained by the fact that our study population consisted of individuals who had multiple episodes of exposure to Leptospira spp. It must be noted that one additional reason for the different results in different studies for all the tests could also be due to batch variations related to substandard manufacturing.   Source:


Leave a Reply

Your email address will not be published.