Over-diagnosis of malaria is not a lost cause

From the beginning of 2005, a randomized controlled trial (RCT) of intermittent preventive treatment for malaria in infants (IPTi) has been conducted at the study hospital. As part of the RCT, the study team offered laboratory quality assurance, and supervision and training of paediatric ward staff using information on malaria epidemiology in the community. Data on clinical and blood slide confirmed cases of malaria from 2001 to 2005 were extracted from the hospital records.The proportion of blood slides positive for malaria parasites had decreased from 21% in 2001 to 7% in 2005 (p < .01). The proportion of outpatient and inpatient cases diagnosed as malaria ranged between 34% and 28% from 2001 to 2004 and this decreased substantially to 17% after the introduction of the package of training and support in 2005 (p < .01). There was no clear trend in the ratio of blood slide examined versus total diagnosis of malaria.It may be possible to change the diagnostic behaviour of clinicians by rigorous training using local malaria epidemiology data and supportive supervision.Recent studies have highlighted the massive problem of over-diagnosis of malaria in malaria endemic countries [1-3]. Over-diagnosis of malaria has moved to the forefront of policy due to the unacceptable increase in mortality associated with an incorrect diagnosis [2] and the cost implications as a result of rising prices of first line drug treatment options [4].This paper reports a case study of rigorous training using local malaria epidemiology data that changed the diagnostic behaviour of clinicians in a rural district hospital in an area of low to moderate malaria transmission.Same District, situated in north-eastern Tanzania, is a plain with an altitude between 600 m and 1,000 m. The lower areas are dry semi desert inter dispersed by large irrigated rice growing areas. In 2001, a cross sectional survey in the district showed that malaria parasite prevalence in under five year olds varied from 25% in