Poor Accuracy of Rapid Diagnostic Tests and Misdiagnosis of Imported Malaria: Are PCR-Based Reference Laboratories the Answer?

The recent article by Rubio et al. reports poor accuracy of rapid malaria diagnostic tests (RDTs) and promotes PCR-based reference laboratories to avoid potential misdiagnosis (6). Both points deserve comment.

The authors' reported low sensitivity is mainly due to poor detection of Plasmodium species other than Plasmodium falciparum. Although “pan-specific” antibodies have been included in some RDTs, it is the diagnosis of P. falciparum infection that is crucial for reducing mortality. Here, Rubio et al. confirm the findings of many studies, reporting a sensitivity well above 90% (7). Moreover, an unusually high number of false-positive results are presented, without attempts to identify a reason. For example for an average of more than 17 days, 44.4% of OptiMAL (pLDH-based assay) results were reported to be false positives. However, pLDH is only produced by live parasites, and studies have shown very good correlations with declining parasitemias during follow-up (4). The reported number of pretreatment false positives (>20%) even exceeds numbers reported from studies where the influence of rheumatoid factor on RDT performance was investigated in nonmalarious patients (1).

The authors use this unduly dark picture to justify the need for PCR-based reference laboratories, although they fail to elucidate the clinical usefulness of such tests in the diagnosis of acute malaria. Following their argument of limited accuracy of malaria diagnostic tests, confirmation by a reference center would have to include positive as well as negative samples. Hence, there would be considerable financial and logistic hurdles for such a policy: urgent transport, turnaround time of PCR runs, weekend service, etc. From the authors' country (Spain), a mean annual number of 175 cases per year and for Europe 5,000-8,000 cases per year were reported from 1985 to 1995 (5). However because millions of tourists visit countries where malaria is endemic, each year many of them may present at health service facilities with unspecific symptoms after return, prompting a malaria diagnostic test. For each positive specimen there will be many negative ones. Unfortunately, the feasibility of such a “fast-track” reference laboratory service is not discussed (e.g., in a cost-benefit analysis).

Furthermore, lack of clinical suspicion may be more important in the misdiagnosis of imported malaria than poor performance of laboratory tests. In Canada, more than 50% of malaria cases were clinically not suspected upon first presentation (3). RDTs or PCR-based reference laboratories do not address this problem. Development of common routine laboratory tests that alert to the presence of Plasmodium spp. may be more useful. One example is the detection of hemozoin during automated routine full-blood counts, which allowed the diagnosis of six clinically unsuspected malaria cases in one study (2).

Reference centers play an important role primarily in the posterior confirmation of positive samples, species identification, and comprehensive quality assurance. However, whether the use of presently available RDTs increases the misdiagnosis of imported malaria, as suggested by Rubio et al., remains to be proven by an appropriately designed study. Such a study would have to investigate the routine use of RDTs in normal everyday practice involving non-expert microscopists in peripheral laboratories. Is it not more likely that the judicious use of RDTs would reduce the misdiagnosis of imported malaria in many of these places?