Concordance analysis of three diagnostic tests for malaria in the symptomatic population of Colombian endemic municipalities.

Introduction: Taking into account the difficulty of performing malaria microscopic diagnosis in rural areas, rapid diagnostic tests (RDT) are a good alternative, but it is important to verify their diagnostic performance. Objective: To evaluate the diagnostic performance of the RDTs used in five Colombian departments by comparing them with the microscopic diagnosis and using PCR as the reference standard. Materials and methods: Thick blood film and RDTs were used to diagnose symptomatic individuals; additionally, the filter paper was impregnated with blood for the molecular test. Results: We included 314 samples whose percentage of positivity for malaria was 49% by PCR, 48% by microscopy and 46% by RDT; parasitemia ranged between 180 and 23,800 p/μL of blood. The concordance of the results from the microscopy units and those of the PCR (National Laboratory of Reference) was as follows: Cohen’s kappa coefficient, 0.975 (95% CI: 0.950-0.999); sensitivity, 97% (95% CI 95-100); specificity 100% (95% CI: 100-100), and kappa index of species, 0.958 (IC95%: 0.912-1.00). The concordance between the Pf/Pv RDT (at the microscopy units) and the PCR (National Laboratory of Reference) was as follows: kappa coefficient, 0.878 (95% CI: 0.784-0.973); sensitivity, 94% (95% CI: 87-100); specificity, 95% (95% CI: 90-100), and kappa index of species, 1.0 (95% CI: 1.00-1.00). The concordance between the Pf/Pan RDT versus PCR was: Cohen’s kappa coefficient, 0.920 (95 % CI: 0.865- 0.974); sensitivity, 94% (95% CI: 90-98); specificity, 99% (95% CI 95-100), and kappa index of species, 0.750 (IC95% 0,637-0,863). Conclusion: The results of this study support the use of RDTs in Colombia; however, more training of the personnel is required to accurately differentiate Plasmodium species.