Lay reporting of elephantiasis of the leg in northern Ghana.

Within a large scale community trial in northern Ghana lay interviewers were trained to inquire about and identify elephantiasis of the leg by the use of local terms and simple examination of respondents. This was repeated a year later after moving the interviewers to different geographical areas. The proportions of extended family compounds reported to have at least one member with elephantiasis of the leg were 12.2% and 12.1 % respectively in the first and second surveys (kappa = 0.60). 'Blind' re-examination of a sub-sample by a physician showed a high level of agreement with the lay interviewer's findings in the first and second surveys (kappa = 0.67 and 0.82 respectively). This study has shown that lay people, even with minimal training, can obtain repeatable and valid estimates of the prevalence of elephantiasis of the leg, at least within an area where local terms for the condition are available. This method could potentially be used for other diseases with visible manifestations.