The rapid monitoring of ivermectin treatment: will school-based surveys provide the answer?

The data on ivermectin-treatment coverage recorded in household surveys sometimes conflict with those recorded in school-based surveys or in the relevant treatment registers maintained by community-directed distributors (CDD). An attempt has now been made, in two sites in Nigeria (Enugu and Kaduna states) and one in Sudan (Abu Hamad province), to determine how well these three sets of data are correlated (and to explore the effectiveness of several alternative channels for the delivery of treatment-monitoring forms to schools). Using a cross-sectional approach, data were collected from primary schools, households and treatment registers. Calculation of Pearson's correlation coefficients (r) indicated that, overall, the data from the household surveys were very similar to those collected using the school-based strategy (r=0.66; P<0.0001) or from the treatment registers of the CDD (r=0.86; P<0.0001). The information recorded in the CDD registers also closely matched that recorded in the school-based surveys (r=0.67; P<0.0001). These encouraging results for the pooled data masked some inter-site differences. The correlation between the household-survey and treatment-register data was, for example, only good in Enugu (r=0.89; P<0.001), and was too weak to be statistically significant in Abu Hamad or Kaduna. Although the results of the school-based survey in Kaduna also did not closely correlate with those of the corresponding household survey (r=0.10; P=0.71), the household survey at this site was probably not conducted as well as those at the two other sites. In general, it appears that school-based surveys are an effective means of monitoring community coverage with ivermectin, rapidly, accurately and at relatively low cost. It is therefore recommended that school-based methods of monitoring of coverage are adopted by programme managers.