authors' reply Drs Douglas and Osman correctly state that we did not perform a systematic review of the efficacy of education programmes, but neither did we claim to do so. Our goal, clearly stated in the title, was to describe the objectives, methods, and content of education programmes. In fact, we renounced conducting a meta-analysis of the effectiveness of programmes when we realised the extent of the variability of educational interventions. Averaging the proverbial apples and oranges did not make much sense. Our study suggests that, not only is the number of fruit species greater than anticipated (variability between programmes), but you cannot always tell one from the other (insufficient description of programmes). The latter finding implies that even a systematic review aimed at identifying features associated with greater effectiveness is not feasible. Such an endeavour would be further complicated by the fact that variables used to assess efficacy vary from one evaluation study to the next. In our opinion, standardisation of both programme descriptions and evaluation methods would foster progress in patient education.

While randomised controlled trials are the gold standard for assessing efficacy, all studies reporting an educational intervention should describe in sufficient detail what that intervention consisted of. We therefore included in our review all studies that had an educational component, regardless of the evaluation design.

We admit that we used a broad definition of education as “any attempt to provide the patient with knowledge or personal skills to reduce the impact of asthma on health”. The educational content varied among programmes (this is one of our main points) and could include drug management, environmental control, relaxation, yoga, etc. The paper by Partridge1-1 provides an interesting description of an education programme in an asthma clinic, its weaknesses, and attempts at correcting these. As for including work published only as a dissertation, this may be considered an advantage rather than a drawback by some meta-analysts. We maintain that all studies that we reviewed included an explicit educational component and doubt that changing eligibility criteria to exclude a small subset of studies would much alter our general conclusions.

We stand corrected about the incomplete reference to the Grassic intervention in the discussion section of our paper.1-2 In our database this programme was described more accurately as follows (partial data): number of training sessions:4 (counting one 10 minute session in person and three mailed booklets); duration of training period: 3 months; delivery of education by:physician andself-help; educational setting:individual; training tools:booklet; training method:lecture/vertical teaching. Had we conducted an effectiveness review we would have no doubt singled out this study as by far the largest trial of asthma education, and one that did achieve clinical benefits for its patients. More such research studies are needed.