Community consultation in emergency research.

CONTEXT

Recent federal regulations allow for emergency research without prospective consent provided that additional protections for the subjects are provided. One of these is community consultation.

OBJECTIVE

To determine how to feasibly consult the community and what to do with the findings.

DESIGN, SETTING, AND PARTICIPANTS: In connection with an ongoing study of l-arginine for brain injury at a public hospital's trauma center, we consulted three sets of community representatives using different methods. To sample the entire population of the county, we conducted a random-digit dialing survey (456 residents). To sample individuals served by the hospital, we interviewed a convenience sample of 566 patients and individuals in the waiting area. To sample particularly interested individuals, we conducted a series of public meetings (114 participants). In each case, the same instrument was used to ascertain their attitudes toward the study in general and toward its major features (randomization, waiver of consent, location, risks/benefits). To control for framing effects, items were randomly presented in a positive or negative fashion.

RESULTS

All methods proved feasible, but telephone surveys were most efficient and guaranteed the desired demography. Even for a very low-risk study, only 79.75% of respondents would be willing to participate. The rate of approval for the major features was lower, with only 67.78% approving of the risk/benefit ratio, 53.7% approving of randomization, 57.66% approving of the consent waiver, and 44.45% approving of the location. The most significant factors affecting the rate of approval were the method of consultation and the framing of items (both p < .001), age, ethnicity, and previous research participation (all p < .01).

CONCLUSIONS

Community consultation is feasible, but its results depend heavily on method of consultation, framing of questions, and choice of community. It may well demonstrate unexpectedly substantial opposition. There needs to be a better definition of the process and a better understanding of how to respond to its results.