Over-diagnosis of malaria by microscopy in the Kilombero Valley, Southern Tanzania: an evaluation of the utility and cost-effectiveness of rapid diagnostic tests

Abstract

Background: Early and accurate diagnosis of febrile patients is essential to treat uncomplicated malaria cases properly, prevent severe malaria, and avert unnecessary anti-malarial treatments. Improper use of anti-malarials increases the risk of adverse drug reaction and the evolution of drug/parasite resistance. While microscopy is the most common form of malaria diagnosis, concerns over its accuracy have prompted the incorporation of malaria rapid diagnostic tests (RDTs) into many national malaria control programmes. Methods: Over a three-month period, a direct comparison between microscopy and RDTs was made in a rural, private dispensary in the Kilombero Valley, Morogoro District, southern Tanzania, with the aim of estimating the extent of malaria over-diagnosis and over-treatment with anti-malarials. The study cohort was made up of patients referred by the dispensary’s clinician for malaria testing. One hundred percent of patients approached agreed to participate in this study and were then tested using both microscopy and RDTs. Using the results from the comparison of the two tests at this dispensary, the potential cost effectiveness of introducing RDTs to a neighbouring public health centre was estimated on the basis of this centre’s past malaria records spanning December 2007 to August 2011. Results: At the private dispensary, the apparent prevalence of malaria was 78% based on microscopy whereas the true prevalence, calculated using RDTs as the gold standard, was estimated at 14%. This discrepancy indicates that when using microscopy as the sole diagnostic test, malaria is being over-diagnosed by approximately a factor of five in this setting. At the public clinic, apparent malaria prevalence based on microscopy was 74%. If similar rates of over-diagnosis are assumed, 5,285 patients of the 6,769 patients positively diagnosed with malaria using microscopy were likely given unnecessary anti-malarials, and their true cause of illness was not addressed. The introduction of RDTs to the public clinic would be highly cost-efficient, with an estimated net saving of over 96 USD/month. Conclusions: Compared with RDTs, microscopy led to almost four out of five patients being over-diagnosed with malaria in this rural part of Tanzania. A policy that encompasses both the private and public sectors of health care is needed to ensure quality diagnostic testing for febrile patients. With estimated prevalence at 14%, RDT introduction is recommended given WHO findings that RDTs are predicted to be cost-effective in prevalence areas of less than 20%. The use of RDTs in malaria diagnosis would not only reduce government spending but would prove beneficial to ensuring appropriate care and treatment of febrile illness.