Objective: To investigate case management of fever among under-fives, performance and
utility of malaria rapid diagnostic test (mRDT) under reduced malaria burden in Rufiji District,
Southeastern Tanzania.
Methods: A quantitative cross sectional study was conducted at primary health facilities in
Rufiji District from April to May 2012. Information on socio-demographic characteristics,
history and duration of fever, fever measurement, clinical diagnosis and drugs prescribed
were recorded. Parasitological malaria confirmation was done by mRDT and microscopy.
Performance of mRDT and utility of mRDT results to guide on the management of malarial and
non-malarial fevers were assessed.
Results: Of the 466 under-fives with fever, 111 (23.8%) were mRDT positive and 100 (21.5%)
were microscopically positive for malaria. Sensitivity and specificity of mRDT were 90% [95%
confidence interval (CI): 82.6%–94.5%] and 94.3% (95% CI: 91.4%–96.2%) respectively;
overall diagnostic accuracy was 93.3% (95% CI: 91.1%–95.6%). A total of 130 (28.5%) underfives
received an antimalarial. Among them, 109 (83.8%) were mRDT positive while 21 (16.2%)
were negative. Of the 100 under-fives with microscopic parasitaemia, 34 had counts > 200000/
µL an indication for quinine but only 5/34 (14.7%) received quinine prescription. Five underfives
with parasitaemia > 200000/µL had negative mRDT results. Being mRDT negative was
significantly associated with receipt of an antibiotic prescription (χ2 = 162.2, P < 0.001).
Conclusions: Use of mRDT reduced unnecessary antimalarial use by 71.5%. However, this
had the potential for over prescribing an antibiotic for non-malarial fevers. The diagnostic
performance of mRDT was still high despite decline in malaria burden.