Comparison of methods for estimating the frequency of paralytic poliomyelitis in developing countries.

Standard reporting systems have proved to be unreliable in estimating the frequency of paralytic poliomyelitis in many developing countries. The effectiveness of three survey methods for estimating the prevalence of lameness attributable to poliomyelitis were compared in the Danfa Health Project district of rural Ghana. Lameness was studied because it is easily identified by inexpensive survey techniques. A postal survey of school headteachers was the least costly and most accurate method, giving a prevalence of lameness attributable to poliomyelitis of 7.2 per 1000 schoolchildren. This method is recommended for countries with an extensive network of primary schools. Medical examinations in a village health survey yielded a prevalence of 4.6 per 1000 children aged 6-15 years. This method was relatively expensive but could be justified if the survey was multipurpose. A question added to the annual district census showed a prevalence of 2.2 per 1000 children, a rate still many times higher than estimates from national statistics. The inclusion of such a question in a census or other national survey entails little additional cost and with more careful questioning results should improve. Repeated surveys can be used to assess the success of poliomyelitis vaccination programmes.