Tools for community-oriented primary care: use of key informant trees in eleven practices.

Physicians increasingly need information about their communities to use in care of the individual patient. Busy practitioners need feasible methods for collecting this information before they can begin to gather and use it, however. Our objective was to study key informant trees as a practical approach for practice-based gathering of qualitative data from a community. Following a standard protocol, key informant trees were set up in 11 different practices to study the costs, advantages, and problems with their use for this purpose. Time studies showed that each tree took 7 to 11 hours of physician time and 7 hours of clerical time to organize and conduct. The technique appeared to be best suited for two qualitative informational needs: idea generation and explanatory data gathering. Trees appeared most productive where there was stability of physician staff in the practice, where the practice had been present in the community for some years, and where community residents were relatively stable. Response and selection biases are important considerations in use of this technique.