Although the study design used to estimate the benefits associated with clinical follow-up interventions may result in important biases due to the lack of controls, ethical reasons appear to have prevented the use of a better, controlled design to that purpose. The prospective nature of the study, however, supports to some extent the validity of the results. The low attrition rate also appears to act in favour of the validity of the results, whilst the large sample used reduces the possibility of chance being a crucial explanatory factor in the study results.