(1) The authors could have provided more information on the study design,the search strategy and the selection criteria for the clinical trials included in the review.
(2) The authors used a pooled estimate, from separate trials, of the review. There was no evidence that the groups of patientswere comparable.
(3) More information on the model which the authors developed would have been useful.
(4)More information on costing would have been useful. It is not clear how the authors derived the cost of treating a patient per day. Furthermore the costs of side-effects were not considered. The estimates of the cost of GP consultations are varied and the authors give no justification for their choice of estimate. No information is given on the dates to which the costing data referred.