A number of published studies have described major epidemics, community and nosocomial outbreaks, and the ecology of CCHF (4,7,8,27–30). These reports have shown that a distinct epidemiologic situation can arise in regions where the virus is endemic but also that new foci can emerge (10,31). The World Health Organization has listed CCHF among the emerging diseases for which control and prevention measures should be renewed and intensified (13,32). In addition, an assessment of the importance and magnitude of vector-borne diseases initiated by the European Centre for Disease Prevention and Control identified CCHF as a priority disease for the European Union (33).

A strong laboratory capacity, in particular standardized approaches for diagnostic methods and assay validation, has been identified as a short-term priority in CCHF-endemic areas and regions where CCHFV could be expected to circulate (11,12). The aim of this study was to identify and evaluate easily available, simple-to-perform CCHF diagnostic methods considered most suitable for widespread use in CCHF-endemic areas and countries at risk.

We assessed the performances of 2 commercially available IgM and IgG serologic tests, the Vector-Best CCHF ELISA and the Euroimmun CCHF IFA, and 2 assays for viral genome detection, the Altona-diagnostics CCHF qRT-PCR and a CCHF LCD array. The IgM and IgG ELISAs showed a sensitivity of 88% and 80%, respectively, lower than the numbers given by the manufacturer. These assays were validated by the manufacturer by using serum panels from CCHF cases originating from southwestern Russia (S. Suchkov, pers. comm.). Therefore, lower sensitivity for certain serum samples tested in this study may reflect antigenic variation among CCHFVs circulating in other countries.

Observed sensitivities of the IgM and IgG IFA were 93.9% and 86.1%, respectively. Although these estimates are higher than those observed for the IgM and IgG ELISAs, these results may have had a sampling bias because not all serum samples tested by ELISA could be tested by IFA. This bias was, however, minimized, because the tested serum panel included 15/16 false-negative samples observed for the ELISAs.

The sensitivity of the selected molecular assays was found to be more modest (79.6% for qRT-PCR and 83.3% for LCD array) than for serologic methods and to be associated with the patient country of origin. This result is consistent with the finding that the application of molecular assays in different settings is hampered by the high diversity of the CCHFV genomes, whereas serologic methods can have a broader use due to cross-reactivities. In particular, the qRT-PCR seems to be less sensitive for patients originating from the Balkans region and Turkey than for patients from other countries compared with in-house reference molecular methods. The in-house methods developed by reference laboratories are optimized for detection of strains circulating in that area, which may result in a lower detection limit when compared with methods that cover a broader spectrum. However, other factors, such as RNA degradation due to freezing and thawing cycles or inhibition of PCR reactions because of inhibitory compounds in the samples, may have contributed to decreased sensitivity of molecular methods compared with serologic methods.

The observed specificity was excellent for all assays, ranging from 95.5% to 100% when compared with the reference method and equal to 100% when compared with a composite reference standard. However, predictive positive and negative values for the different assays could not be calculated because precise prevalence data are not available for most CCHF-endemic areas, and these data cannot be predicted for areas where the virus could emerge.

The CCHF IFA demonstrated a higher complexity in equipment, technique, and interpretation. However, the interpretation of fluorescence patterns may enable trained users to differentiate positive from cross-reactive serum samples, whereas such false positives may not be avoided with the ELISA. The operational characteristics of the molecular assays were comparable. Both methods required regular equipment maintenance and specific training for appropriate use of the equipment; however, this was more apparent for the real-time qRT-PCR method. The LCD array technique was considered to be complex but acceptable. Training for technique and result interpretation was recommended for each method.

All participating laboratories are reference centers for CCHF laboratory diagnosis and surveillance in their respective countries, and some are World Health Organization collaborating centers. The protocols from reference methods in use at each site have been extensively validated previously (16–20,24). In addition, the laboratories participated in a recent international external quality assessment (34). Therefore, local conditions at the participating sites and validity of reference methods were considered comparable.

The multicenter design of this study enabled the testing of a large sample size, representative of >1 population, so findings could be more generally applicable. In addition, sample panels were constituted without any selection for disease severity, and negative controls included not only healthy patients but also patients who had a wide range of other conditions, thereby avoiding inflated estimates of diagnostic accuracy.

Our study has some limitations. Because archived samples were used for the study, specimen quality could have been affected; however, statistical analysis demonstrated that sample storage time and temperature did not influence sensitivity. Also, the use of different sample panels for serologic and molecular testing did not enable calculation of the added value of combining the serologic and molecular CCHF diagnostic methods evaluated.

The results of this study give additional guidance on the type of CCHF diagnostic tools that could be used in different contexts. During a large outbreak, easily interpretable tests for simultaneous analysis of numerous samples, such as the ELISA and real-time qRT-PCR, might be considered useful tools to identify CCHF cases. Methods available in smaller format size and demonstrating a long shelf life, such as the IFA and LCD array, could be used to identify sporadic cases or to confirm single cases as part of a larger outbreak.

This study demonstrates that efficient, well-characterized serologic and molecular assays and protocols are available for CCHF diagnosis. The on-site use of such assays by outbreak assistance laboratories would greatly diminish the risks posed by the handling, packaging, and shipping of highly infectious samples. Moreover, acquiring diagnostic reagents would be more time- and cost-effective for laboratories than would the organization of compliant packaging and shipment of hazardous biologic material to reference laboratories abroad. Nevertheless, laboratory personnel should receive the appropriate training to perform the different assays (e.g., during international workshops or network meetings). Collaborative evaluations of diagnostic methods remain essential to guide decision-making, especially with emerging diseases, where a standard is frequently missing, and laboratory expertise is rare.