A direct comparison of real time PCR on plasma and blood to detect Plasmodium falciparum infection in children.

BACKGROUND: Estimation of Plasmodium falciparum parasitaemia can vary with the method used and time of sampling. Quantitative real time PCR (qPCR) on whole blood or plasma samples has previously been shown to be more sensitive than thick film microscopy. However the efficiencies of each method have not been compared using samples obtained from infants less than one year old. METHODS: A multiple of statistical approaches were used to compare the performance of qPCR on whole blood or plasma to detect the 18 S ribosomal gene of P. falciparum in 548 samples from children aged 2.5 or 24 months. Parasite prevalence in matched samples was compared using Mcnemar's test and agreement of positive results quantified as Kappa scores. Parasite prevalences between different age groups were compared by Fisher's test. Results from analyses by thick film microscopy were also available from children at 24 months and their correlation to each qPCR method examined by the Spearman's test. Finally the association of P. falciparum infection with the incidence of multiple malaria episodes from contact to 24 months of age was evaluated using negative binomial regression. RESULTS: These analyses showed that qPCR from whole blood detected approximately 3-fold more cases of infection than plasma qPCR. Both qPCR methods agreed well with each other although qPCR from plasma had a greater agreement with microscopy (96.85%) than did qPCR from blood (69.7%). At 24 months the prevalence of infection detected by all methods was associated with anaemia (p<0.05). CONCLUSIONS: The data presented here demonstrates that low levels of parasitaemia are better detected by qPCR using parasite DNA from whole blood than from plasma. However plasma samples provide a viable substitute when parasite smears are unavailable.