Introduction
An extended drug-information and management service from community pharmacists for people with mental health problems (clients) was evaluated using a combination of research methods. The research project comprised three phases: pre-intervention, intervention and post-intervention. Community pharmacists, members of community mental health teams (key workers) and people with mental health problems were involved in data collection at all three phases of the project. Four different research methods were used to generate data for evaluation.

Methods
The methods used during the research project are summarised below.

The clients completed a structured interview after their recruitment. They were asked to keep a record of the dates of their visits to their link-pharmacist using calendars provided by the researcher. After 5 to 7 months of their participation, they were invited to a semi-structured interview. A selected group of 4 clients were also invited to participate in a consensus development panel with the pharmacists and key workers.

Results
One of the most challenging aspects of this research was the recruitment of clients. After 6 months, 26 clients were recruited (one client was recruited separately at a later date, n=27). However, in the majority of cases, once the clients had agreed to participate, they were highly committed. Involvement of their key workers throughout the research project eased the process.

The structured interview was brief and administered without problems. The calendars were kept by minority of clients (n=6) but where they were completed, they offered valuable information. Clients were willing to participate in a semi-structured interview (23 out of 27 were interviewed, remaining four were unable to participate), they were communicative, open and honest. Their contribution to the consensus development panel was exceptional.

Conclusions
This project highlighted the value and difficulties associated with involving of service users, and particularly people with mental problems, as a source of data.

The clients' contributions to data collection were invaluable, widening greatly the researcher's insight into the issues. The clients' commitments were a pronounced feature of the research process. In particular, the clients' contributions to the consensus panel were admirable.

The research project also highlighted issues that need to be addressed during a research involving people with mental health problems in data collection such as their recruitment and safety issues.

This project demonstrated that service users with mental health problems should be considered as willing and valuable primary sources of data.