During the internet search eight RDT products were encountered and ordered. Three of these products were not included in the final selection: one (Malapack Travel test, http://www.vaccinations.com.au/product.htm
) was not marketed anymore, the manufacturer of the second product (EZ-Trust Malaria Rapid Screen Test Kit, CS Innovation Pte Ltd, Singapore) replied not to start up the production for an order less than 10,000 tests. The third product that was delivered (Unitest malaria cassette Pf-Pv, Ciriano global S.L., Zaragoza, Spain) appeared to be an RDT detecting malaria antibodies instead of antigens (), whereas the product ordered from the manufacturer’s website clearly mentioned an antigen-detecting product.
Received product of Unitest (Ciriano global S.L., Zaragoza, Spain).
In addition to the five ordered RDT products, the three tests directly provided by the manufacturers (as explained above) were also included, so that a total of eight RDT products was evaluated (). Six RDT formats were cassettes, there was one hybrid format (OptiMAL) and one RDT (Immunoquick) consisted of a dipstick.
The number of samples detected and the number of false positive lines for malaria negative samples for each RDT product are displayed in . For 6/8 RDT products 100% sensitivity was reached for P. falciparum diagnosis, but Labstix and Onestep showed false positive lines in respectively 80% and 70% of malaria negative samples. Three RDT products detected all P. vivax samples. P. ovale and P. malariae detection was in general poor, Labstix was the only one diagnosing all samples but this was at the expense of poor specificity as this product generated visible pan-pLDH lines in 8/10 Plasmodium negative samples as well.
Test characteristics of the different RDT products.
Cross-reactions of non-falciparum samples with the P. falciparum test line occurred for the majority of samples in Labstix (94.3%) and OneStep (80.0%). In addition for OneStep 47 P. falciparum samples and 12 P. ovale/P. malariae samples showed a visible P. vivax test line.
Faint Test Lines
For P. falciparum diagnosis, the median percentage of correctly identified test lines with faint line intensity per RDT product was 3.6% (range 0.0%–14.0%). For the correctly identified non-falciparum samples, faint line intensities occurred at a median frequency of 23.8% (range 10%–58.3%).
Median κ for positive/negative readings for the P. falciparum
test line was 0.95 (range 0.50–1.00, OneStep κ
0.50), for the non-falciparum
test line this was 0.91 (range 0.66–1.00). For line intensity readings median κ were 0.83 (range 0.70–0.96) and 0.76 (range 0.71–0.85) for the P. falciparum
Reading Beyond the Recommended Reading Time
For all but one (SDFK63) products false positive test lines were seen after reading beyond the recommended reading time among 10.0%–90.0% of the malaria negative samples and 9.4%–100% of the non-falciparum samples with initially no false positive test lines. Immunoquick, TODA and Labstix were mainly affected with nearly half (>44.4%) of malaria negative samples erroneously diagnosed as malaria and the majority (>88.2%) of non-falciparum samples diagnosed as P. falciparum malaria upon reading beyond the recommended time.
RDT Packaging, Content and Design
Half of RDT products assessed had inconsistencies in their names displayed on the outer packaging, the device packaging and the IFU. Moreover, 3/8 RDT products did not mention their target antigens and 2/8 RDT products did not display recommended storage temperature (). In general, the RDT products contained all material needed to perform the test, however to open the buffer vial of Labstix, a scissor was needed but this was not mentioned among the required materials, only in the procedural steps of the IFU. Only 2/8 (25.0%) RDT products contained a correct desiccant – i.e. provided with a color-based humidity indicator allowing to control for humidity saturation. Two products did not contain a transfer device, the drop of blood had to be applied directly to the test strip. The pipettes supplied with Labstix and OneStep did not contain a mark for indication of the correct volume of blood.
For 4/7 RDT devices test line labeling consisted of acronyms (‘Pf’, ‘P’, ‘Pan’), the others used either letters (T) or numbers. Immunoquick contained no labeling at all as it is a dipstick without cassette. Incorrect labeling of the reading window was found for Sanitoets: only the symbols ‘C’ and ‘T’ were displayed at either side of the reading window while the strip contains a control line (“C”) and two test lines. For Labstix discordances were observed between labeling on the test device (Pf, Pan) and the pictures displayed in the IFU (T2, T1).
Instructions for Use
OneStep did not deliver an IFU with the tests, nor a link to the online version. On the website where the product was ordered an IFU was found. The IFU of OptiMAL consisted only of pictures. None of the IFUs fulfilled requirements for correct font size (>12), median font size was 5.0 (range 4.5–9). Only two products used an interline distance of 2 and six IFUs used an open character. Median Flesh Kincaid grade level was 8.32 (5.86–9.65), only Immunoquick had a Flesh Kincaid grade level ≤6th grade.
Shortcomings of the IFU with regard to test procedure and interpretation are displayed in . Several critical procedural steps were missing in more than half of the IFUs. For three RDT products, the correct use of the blood lancet was not clearly described or depicted. Sanitoets depicted another type of pipette in the IFU than was delivered in the kit. Moreover, Sanitoets mentioned to cut the end of the sealed pipette that contained the reagent while in the kit a buffer vial with a screw cap was included. Four IFUs failed to mention that the test should not be read beyond the recommended reading time and one IFU (TODA) did not display a reading time at all (for this product, information about reading time was withheld from the box that contained the single packages used for self-diagnosis).
Presence of important items that need to be addressed in the instructions for use.
OneStep did not contain any pictures of possible test results in the interpretation section. OptiMAL and TODA did not mention a visible control line and single visible P. falciparum test line as a possible result. Following their instructions, a P. falciparum infection always generates both a visible P. falciparum and pan-pLDH test line. In the present study however, OptiMAL and TODA showed a single visible P. falciparum test line among one and three P. falciparum samples respectively, leading to 49/55 (89.0%) and 52/55 (94.5%) correctly identified P. falciparum samples respectively. IFUs that did not display all possible combinations for invalid results generally only depicted a test in which none of the lines were visible. For TODA, an error was observed on the figure showing the invalid results, i.e. a cassette was displayed with a visible control line and an absent test line. Except for Labstix that depicted a less visible test lines among the positive results, none of the RDT products mentioned to consider a faint test line as a positive one. Likewise, causes of false negative and false positive results were not mentioned. SDFK63 was the single product mentioning that a negative test does not rule out malaria. Hardly any advices about test policy were made, except for TODA mentioning to use the ‘pan’ line for treatment follow-up of a P. falciparum infection and Labstix and Immunoquick advised to start treatment in case of a positive test or in case of a negative test and persistence of symptoms ().
Blood Lancets and Transfer Devices
Blood lancets and transfer devices for each RDT product are depicted in . Five lancets were so-called safety lancets with a retractable needle. For two of them reuse was possible and for the other three the needle could inadvertently be launched before removal of the protective cap. Two products did not contain a transfer device, the drop of blood had to be applied directly to the test strip. The pipettes supplied with Labstix and OneStep did not contain a mark for indication of the correct volume of blood.
Lancets and transfer devices delivered with the different RDT products.