To the best of our knowledge, the current study for the first time presents data on the analytical performance characteristics of testing HBsAg, anti-HBc, anti-HBs, anti-HCV and HIV-1-p24-antigen/anti-HIV 1/2 in DBS eluates with the Abbott ARCHITECT system, i.e. the currently most frequently used platform worldwide in this field of serological diagnostics (Abbott Diagnostics, personal communication). Our whole analytical evaluation had to be based on elution conditions suitable for routine use and was not designed to achieve an optimisation for every individual parameter. Consequently, the DBS for the five antigen and antibody tests were eluted with a total of 1,000 μl of a PBS-based buffer with Tween 20 overnight at room temperature . The whole blood eluates for the performance of HBV DNA and HCV RNA tests were prepared under identical conditions in a second operation. The analytical tests comprised 1,762 paired serum/whole blood samples. They originated from 726 patients and were the basis for a total of 3,524 determinations.
There was no analytical non-specificity in serological and molecular biological DBS testing.
The detection of HBsAg positive materials from whole blood eluates succeeded with a sensitivity of 98.6% to a similarly high degree as in previous studies, which had in part used a much smaller elution volume of, for example, 100 μl , 250 or 600 μl , or 500 μl . In the two non-detected DBS samples, there were HBsAg concentrations of 749 IU/ml and 6 IU/ml. Such a low antigenaemia would probably hardly ever occur in currently injecting drug users chronically infected with HBV because the majority of those people does not receive antiviral treatment .
Whether the serum anti-HBc/anti-HBs system for the differentiation of previous HBV infections, on the one hand, and antibody titres as a result of vaccination, on the other hand, is generally also appropriate in DBS eluates has been very rarely investigated to date. Komas and co-workers  determined an analytical specificity and sensitivity of 100% for both tests in 15 and 10 corresponding serum/DBS pairs, respectively. In contrast, Tappin et al.  were only able to detect anti-HBc antibodies in DBS eluates from 7-day old neonates in 44 of 56 cases (sensitivity: 79%) with a modified hemagglutination assay. Another investigation indicated in an exemplary manner that the anti-HBc/anti-HBs system definitely reacts critically to dilution . Whereas the anti-HBc antibodies in the whole blood eluates were detected independently of the elution volumes (150 or 600 μl) in twelve patients chronically infected with HBV, twelve resolved HBV infections could only then be correctly recorded if the quantity of buffer used to eluate the DBS did not exceed 250 μl. The observations we made on the Abbott ARCHITECT system with an elution volume of 1.000 μl confirm these results communicated by Villa et al.  in as much as the anti-HBc/anti-HBs system completely failed when it was used on individuals infected with HIV. In addition, our results prove that immunocompetent persons can also be tested false negative for anti-HBc and anti-HBs in rare cases. This applies particularly to subjects with very low anti-HBc and anti-HBs concentrations in serum.
As a result of its comparatively high detection limit of 100 IU HBV DNA/ml, the artus HBV LC PCR Kit  used for DBS testing in this study turned out to be just as reliable as the Cobas Amplicor HBV Monitor Test [23, 24] or an in-house rt-PCR  and was therefore not superior to older procedures with conventional endpoint detection . For the projected study “Drugs and Chronic Infectious Diseases”, in the course of which DBS testing is to occur, this means that the presumably small number of study participants who had received antiviral treatment could be overlooked, just as could those with a chronic HBV infection and a generally low viremia on the basis of the serological finding “anti-HBc alone” .
The detection of anti-HCV antibodies with the ARCHITECT system was a predominantly smooth process for whole blood eluates. The sensitivity of 97.8% determined in the investigation of 179 serum/DBS pairs corresponded to the results obtained in existing studies [16, 28–31], which however had worked with an elution volume that was lower by a factor of 5 to 10. In addition, the protocols for anti-HCV detection had been appropriately optimised, in contrast to our analyses, by, e. g., stipulating their own cut-off points [16, 29, 31] or by increasing the sample volumes used from 20 μl to 100 μl . The fact that in our study four patients with presumably already long since resolved HCV infections were not recognised clearly illustrates the limits of the DBS technique and this fact should be carefully considered when employing this approach in collectives such as intravenous drug users. In this context, due to the high anti-HCV prevalence, a substantial absolute number of people with resolved HCV infections and consequently low anti-HCV concentrations are to be expected ; they could, at least partially, evade detection by DBS testing due to the dilution used.
The excellent analytical sensitivity of the HCV RNA test with TMA (circa 5 IU HCV RNA/ml) [33, 34] is also reflected in the results of our DBS analysis and allows this approach to appear more promising than the use of an in-house real-time PCR procedure  or a multiplex methodology with detection of the amplification products by means of SYBR Green . However, by way of qualification, it must be said that the mean serum HCV RNA concentration in the 100 samples that we examined with approximately 1.4 million IU/ml was relatively high. Since a dilution factor of 16.5 – 18.2 between HCV RNA positive sera and whole blood eluates obtained with a volume of 500 μl was calculated for determinations with TMA in a previous investigation , an approximate TMA detection limit of 165 – 182 IU HCV RNA/ml may be assumed under the conditions that we used (1,000 μl elution volume) for the DBS testing. Therefore, the two samples with HCV RNA concentrations of 178 and 331 IU/ml, which Tuailon et al.  classified as false negative, could possibly also have escaped from our TMA analysis. This means that a difference in sensitivity of 97%  and the value of 100% which we determined could be simply random.
HIV-1-p24-antigen/anti-HIV 1/2 was ultimately detected with the Abbott ARCHITECT system in the eluted whole blood of all 112 infected persons. Thus, the rather unfavourable routine elution conditions used in this investigation proved to be in no way disadvantageous. The determined ideal analytic specificity and sensitivity of 100% each were not only equal or superior to the performance characteristics established with other immunoassays which have been specifically adapted to DBS testing [38, 39] or a stepwise procedure with the combined use of several anti-HIV tests . They, indeed, also exceeded the performance record of an assay that had been specially developed and optimised for the detection of anti-HIV antibodies in DBS eluates (Q-Prevent HIV 1 + 2 DBS kit) .