A few years ago a colleague who works for a large corporation in Nigeria showed me malaria diagnosis laboratory results from their company clinics.Â The company was able to test workers and family members in their labs after a clinical diagnosis of malaria and before dispensing drugs. Over a span of 5 years only 20-25% of patients were found to be parasitaemic, and without the tests, most people would have been given malaria treatment.Â Those were the days when the company was still dispensing cheap chloroquine. Today with ACTs that cost 10-50 times more than chloroquine, these lab tests are even more important.
Zurovac et al. (2006) found that when both clinical and laboratory skills were improved, major cost savings resulted in malaria treatment. Ochola et al. (2006)Â concluded that dipsticks can perform better than standard microscopy in clinical (field) settings in endemic areas.Â Rennie et al. (2006)Â reported that community health workers can be trained to perform rapid diagnostic tests.
Reyburn et al. (2007) compared RDTs with microscopy in Tanzania among clinic attenders.Â Patients were assigned equally to receive one or the other form of test.Â In both groups approximately 80% of patients had complained of fever within 48 hours.Â Fourteen percent of patients in the microscopy group were found to have malaria parasites, while 16% in the RDT group tested positive.Â Clinical diagnosis might have overestimated the proportion of malaria case by three or four times.
The reality is that few front line health facilities in malaria-endemic have laboratory services, and thus rapid diagnostic tests are being considered. RDTs are estimated to cost between US$0.60 â€“ $1.00, and as implied from research mentioned above, the cost savings from reducing over-prescribed ACTs could be substantial and might offset these costs. Â Â The question is whether malaria control programs will be able to adopt, buy and distribute rapid test materials to front line facilities. Hopefully financial resources like GFATM and PMI can help address this question.