The fact that many people do not take medicines as prescribed has long been regarded as a problem by health care professionals (HCPs). The term concordance is used to signify the need to adopt a different model of the patient - prescriber relationship.

Concordance, with its focus on the relationship between HCPs and patients, is not simply another word for compliance, which is entirely concerned with patients' medicine taking behaviour. The strength of this approach to medicine-taking lies in the assumption of respect for the patient's agenda and the creation of openness in the relationship, so that both HCP and patient can proceed to an understanding, although not necessarily complete agreement, without distrust and concealment.

Concordance is still being developed and discussed in a range of arenas. Yet misinterpretations suggest that the full implications have not necessarily been understood either by those who have embraced, or by those who have criticised, the proposed model.

This paper draws on data from a study entitled Improving doctor- patient communication about drugs. Patients were followed through the consultation process. They were interviewed before and after the consultation and the consultation was audio-taped. The doctor was also interviewed. In this way we gained an understanding of patients' expectations before the consultation, what happened in the consultation and also both doctors' and patients' views of the consultation. We also have data on dispensing and adherence.

An analysis of communication about drugs in the consultation provided a number of examples that had an actual, or potential, effect on adherence. We suggest that pharmacists' involvement would have been of benefit in all these cases.

Medicines were generally not named. This led to non-adherence when a patient did not realise that he had been prescribed a different medicine from the one he had previously tried, and therefore did not dispense the prescription. Failure to name medicines is often accompanied by a lack of information about how the medicine should be used. Interestingly this was most common when the medicine prescribed was also available without a prescription from the pharmacy. It is possible that in such circumstances GPs assume that patients will either already have knowledge about the medicine in question, or that the pharmacist will provide it. Failure to answer patients' queries in sufficient detail and to explore patients' feelings about their medicine was also shown to lead to non-adherence.

Finally, on occasion, pharmacists and GPs presented contradictory information. This demonstrates the need for HCPs to work together to ensure they provide a consistent message and do not confuse patients.

Concordance was developed as an ideal to aspire to and needs to make the transition into practice. This data suggests possible areas in which the pharmacist's role could be further developed. It is however important to note that any interventions to encourage concordance in practice will require the support, enthusiasm and commitment of pharmacists if they are to succeed.