Parasitological indicators of onchocerciasis relevant to ivermectin control programmes in the Amazonian focus of southern Venezuela

In the previous paper it was concluded that those aged greater than or equal to 15 years of both sexes could comprise the indicator group for rapid epidemiological assessment (REA) of onchocerciasis in the Amazonian focus. This paper explores relationships between community microfilarial (mf) prevalence, intensity, and nodule prevalence in 20 Yanomami communities, that would allow identification of REA methods in the region. The mean nodule ratio (prevalence of nodules/prevalence of mf) was 0.54 when onchocercomata in the indicator group were considered. The Spearman correlation coefficient between mf and nodule prevalence was 0.686 (P = 0.001). Palpation of nodules had 92 % specificity and 32 % sensitivity when compared to skin-snipping for the diagnosis of onchocerciasis. The predictive value positive increased from 75 % to 81 % when the indicator group was used. A microfilarial prevalence > 75 % in this group would be indicative of hyperendemic status in the village, between 30 and 75 % of mesoendemicity, and < 30 % of hypoendemicity. For the assessment of infection intensity, biopsies may be taken from the iliac crest for all endemicity levels. Five of the hyperendemic villages surveyed in this work had a community microfilarial load (CMFL) greater than 10 mf/skin snip; the remaining 5 had a CMFL between 5 and 9. These levels of infection merit high priority ivermectin treatment. In Latin America, communities at both moderate and severe risk are included in mass chemotherapy programmes (i.e. when mf prevalence is over 20 %). Roughly, a nodule prevalence in the indicator group > 10 % would suggest a community mf prevalence > 20 % with a sensitivity of 85 % and a specificity of 71 %. A multiple linear regression model of the are-sine transformed mf prevalence in the village (all ages) on nodule prevalence in those aged greater than or equal to 15 years and altitude of the village explained 72 % of the variance. The model combining nodule and altitudinal information had a sensitivity of 92 % and a specificity of 71 % in comparison to an estimated mf prevalence of 21 % or more. It is suggested that the usefulness of the REA methods proposed be assessed in other areas of the Amazonian onchocerciasis focus.