This study showed that the highest frequency of treatment for malaria (73.6%) was recorded by the presumptive method of diagnosis followed by RDT (66%) and microscopy (43.2%). Results from this study are similar to those in Tanzania which showed that over 50% of patients for whom antimalarials were prescribed may not have had the disease 
. As shown in , close to half of the children who were presumptively treated for malaria were microscopy negative. Over-diagnosis of malaria and consequent treatment is a public health problem because it leads to increased reporting of the malaria burden with resultant misallocation of resources to manage the disease, wastage of antimalarials and increased threat of resistance to ACTs. It also results in increased attendance to health facilities due to poor response to treatment (potential misdiagnosis of serious non-malarial infections) and consequent increased workload on the already under staffed and inadequately resourced health facilities 
. With major concerns about parasite resistance development to the ACTs and the high costs of the ACTs, the judicious use of these drugs needs to be given high priority.
The high levels of agreement recorded between RDT and microscopy as diagnostic methods means in transmission areas comparable to ours, one of these diagnostics methods if available, is sufficient as a diagnostic method for malaria. The World Health Organization (WHO) recommends laboratory confirmation (either by microscopy or RDT) of all suspected malaria cases before treatment is commenced and that presumptive treatment should only be considered where such confirmation cannot be done 
. Of the two laboratory methods, RDTs appear to be the method receiving the more prominent attention as they are perceived as having the potential to make a significant impact on improving the diagnosis of malaria. This is because RDTs produce quicker results; do not require any high level of skills to perform them, as opposed to microscopy, which requires more time and reagents, equipment and well-trained/dedicated staff to produce quality results 
In this study, the RDT had a high sensitivity (97.7%) but a rather low specificity (58.1%) for detecting malaria. In terms of diagnosing malaria there was moderate agreement between RDT and microscopy (Kappa
. The Parascreen® RDT used in this study satisfies one of the criteria for a useful diagnostic tool for RDTs with its high sensitivity (97.7%) but not for specificity. In spite of its low overall specificity, the marked increase in its specificity from 51.2% in the rainy season to 88.0% in the dry season showed that it could be a valuable tool to use to improve the diagnosis of malaria during the dry period, especially as there was not much decrease in its sensitivity during the same period. This means that the specificity of the RDT is critical in the dry season when the prevalence of malaria is relatively lower. On the other hand, a high sensitivity of the RDT will be required in the wet season when the malaria prevalence is very high.
As earlier stated, almost a quarter of study subjects diagnosed as malaria cases by RDT were not confirmed by microscopy (). This was likely due to prior treatment with antimalarials with consequent clearing of parasitaemia and persistence of HRP2 antigenemia 
With regards to the cost of treatment in the various groups, overall the cost of treatment (per subject) () did not differ among the three groups. The total cost of treatment was higher in the RDT and microscopy groups even though fewer subjects were diagnosed with malaria in those groups than in the presumptive group. These results seem to suggest that RDTs are not cost-saving when used in the management of malaria in children less than five years, a result that is similar to that of Msellem et al. (2009), which found that cost-reduction using RDTs was not achieved among patients under 5 years but rather among those who were 15 years and above 
. The low positive predictive value and specificity of the RDT lends possible credence to this negative cost-saving effect though the latter property of the RDT contrasts sharply with the 100% for both sensitivity and specificity recorded in a study by the manufacturer 
. The cost of antibiotics was not included in this study. It is likely that children who had a false negative RDT were treated with antibiotics, however, this cost may not be lower than the cost of over-treatment of malaria with ACTs as suggested by Shillcott et al 
Notwithstanding the high total cost of malaria treatment in the RDT group, a potential limiting factor for its use, the RDT can still be said to be an effective tool in reducing the over-diagnosis of malaria and the consequent use of ACTs in non-malaria cases, due to its high negative predictive value. The quality of care of such children will therefore be improved 
One limitation of the study was that since it was a cross-sectional study, there was no follow up of the subjects. It was therefore not possible to ascertain the clinical outcomes of the children who were presumptively treated for malaria (especially those for whom the diagnosis of malaria was not confirmed by RDT or microscopy) and those who were not treated for malaria even though they had been diagnosed as having the disease by the two laboratory methods.
Another possible limitation that since the study was conducted only in children less than five years of age and not in participants across all age groups, the possible cost-saving effect of the RDT could not be determined conclusively.