Communicating Results of New Genomic Tests to Physicians

Abstract

Background: New genomic tests are being developed to predict an individual’s risk of cancer recurrence by analyzing the expression of multiple genes. However, it is unclear how to report the test results so that they would be most useful to clinicians. A mail-out questionnaire has the potential to help a) describe physicians’ attitudes towards the clinical use of new genomic tests, b) determine what information physicians prefer to have included in the test reports, and c) explore how physicians think the test results would impact their treatment recommendations.

Objectives: To design such a questionnaire that could be used in the eventual large-scale survey, and to ensure that the questionnaire a) is comprehensible, b) has face validity, c) appears interesting to, and d) does not place undue response burden on, the target population.

Methods: The first draft, based on a specific genomic test for breast cancer recurrence (Oncotype DX) and on two case scenarios, was created. Cognitive interviews with practicing oncologists were conducted to identify problems in the questionnaire. The evaluation involved face-to-face interviews with Kingston oncologists who treat breast cancer, followed by telephone interviews with medical oncologists who treat breast cancer in other places in Ontario. Three-to-four oncologists were included in each round of interviewing after which the questionnaire was revised based on that round’s recommendations. Additional rounds of interviews were conducted until no new problems/issues were raised in one entire round.

Results: A medium-length questionnaire was drafted. Four rounds of interviews were conducted with no new problems/issues being raised in the fourth round. Most of the problems identified in the questionnaire related to comprehensibility, followed by logical issues which detected fundamental problems in the questionnaire design. There was no evidence of fatigue or disinterest in participants and they deemed the response burden reasonable.

Conclusion: The results suggest that the proposed questionnaire is comprehensible and has face validity. Additionally, it appears to be an interesting questionnaire to, and would not place undue burden on, the target population. Thus, the questionnaire is now ready for the field administration.