The effects of introduced procedural errors on malaria rapid diagnostic test performance in a laboratory setting
摘要
Rapid Diagnostic Tests (RDTs) are the primary means of malaria diagnosis in sub-Saharan Africa. Outside large health facilities, individuals performing RDTs have little or no formal laboratory training, and performing RDTs is by job aids derived from manufacturers’ instructions for use (IFU). Furthermore, in many countries, RDT products are interchanged often without associated training or notification to users of differences in characteristics not immediately obvious to non-laboratory workers. This leads to common deviations from IFUs the consequences of which have not been systematically studied. This study investigated how these errors impact RDT results.
MethodsSix RDT products were tested using cultured Plasmodium falciparum diluted to represent infections with a range of parasitaemia. Tests were performed according to IFU (baseline) then with deviations from the IFU including changes in buffer volume, blood volume and incubation time. Effects of the deviations on test validity and overall test result compared to baseline were captured. Also captured, were effects of deviations on test band intensity, and ease of reading result due to test window abnormalities.
ResultsIncreasing sample volume beyond the recommended 5µL impaired RDT performance, with 4 of 6 RDT products showing 83.3% (30/36) invalid results due to faulty sample migration. No invalid results were observed for the remaining two products. The shortest incubation period (5 min) led to the most deviations from baseline, whereas longer periods aligned more with baseline results. Insufficient buffer volumes caused at least one invalid outcome [10/27 (37%)] in over half the products. Conversely, exceeding buffer volumes led to reductions in the proportion of invalid results among all tests with 0 of 45 invalid results. Higher parasitaemia was associated with increased band intensity and resulted in the fewest deviations from baseline among all products. At 1,000 parasites/µL and 5 µL sample, three products achieved 100% agreement with baseline regardless of incubation time and buffer volume.
ConclusionAlthough malaria RDTs are tolerant of some errors, in general, procedural errors adversely affect results, particularly in low parasitaemia samples. Understanding how sample and buffer volumes, alongside incorrect incubation times, influence RDT performance can be incorporated into training and continuous quality improvement.