[Reliability evaluation in rural Congo of the calibrated blood smear technique to assess Loa loa microfilaremia level]. / Évaluation en milieu rural au Congo de la fiabilité de la méthode de la goutte épaisse calibrée pour évaluer le niveau de la microfilarémie à Loa loa.

Background-Rationale: The diagnosis of Loa loa microfilaremia consists in the observation, using a microscope, of microfilariae in a sample of peripheral blood spread on a slide and subsequently stained (the "blood smear technique"). The accurate quantification of Loa loa microfilaremia is important because the choice of the first intention treatment depends on the patient's microfilaremia: severe adverse events can occur in individuals with high microfilarial densities when treated with ivermectin or diethylcarbamazine, the latter drug being the only one which can definitively cure the infection. However, despite the widespread usage of this technique and its role in guiding clinical management of the patient, estimates of its reliability remain scarce. Materials and methods: We evaluated the reliability (reproducibility and repeatability) of blood smear technique using several sets of 10 L. loo-positive slides, randomly selected, and considered the results with regard to regulatory requirements. The slides had been prepared as part of a clinical trial conducted in Sibiti, Republic of Congo, a region where loiasis is endemic. Results: The estimated and acceptable coefficients of repeatability (NB: the lower, the better) were 13.6% and 16.0%, respectively. The estimated and acceptable coefficients of intermediate reliability (reproducibility) were 15.1% and 22.5%, respectively. The poorest coefficient of intermediate reliability was 19.5% when the tested parameter was related to the technician who performed the readings (10.7% when the reading day was changed). The inter-technician coefficient of variation assessed using 1876 L. loo-positive slides was 13.2%. The coefficient of inter-technician variation considered acceptable was estimated at 18.6%. Discussion-Conclusion. All estimated coefficients of variability were lower than the calculated acceptable coefficients suggesting reliability of the technique, although the lack of laboratory references precludes any conclusion on the quality of this diagnosis. It is imperative to implement a quality system and standardization of procedures for the diagnosis of L. loo microfilaremia, both in endemic countries and in the rest of the world, where the demand for diagnosis has been increasing for years.