The fall in recruitment observed during the first year could be attributable to waning enthusiasm for trial. However, an alternative (and more probable) explanation is the recruitment of existing (prevalent) cases. Once the pool of prevalent patients has been recruited eligible cases are restricted to those with incident disease.

The mixture of incident and prevalent cases recruited to primary care based trials emphasises a further difficulty in analysing and interpreting data on chronic diseases . It may be simpler to restrict the study population to incident cases. However, the results of such trials are less relevant to routine practice.