Perspectives on genetic and genomic technologies in an academic medical center: the duke experience

Abstract

In this age of personalized medicine, genetic and genomic testing is expected to become instrumental in health care delivery, but little is known about its actual implementation in clinical practice.

Methods: We surveyed Duke faculty and healthcare providers to examine the extent of genetic and genomic testing adoption. We assessed providers' use of genetic and genomic testing options and indications in clinical practice, providers' awareness of pharmacogenetic applications, and providers' opinions on returning research-generated genetic test results to participants. Most clinician respondents currently use family history routinely in their clinical practice, but only 18 percent of clinicians use pharmacogenetics. Only two respondents correctly identified the number of drug package inserts with pharmacogenetic indications. We also found strong support for the return of genetic research results to participants. Our results demonstrate that while Duke healthcare providers are enthusiastic about genomic technologies, use of genomic tools outside of research has been limited. Respondents favor return of research-based genetic results to participants, but clinicians lack knowledge about pharmacogenetic applications. We identified challenges faced by this institution when implementing genetic and genomic testing into patient care that should inform a policy and education agenda to improve provider support and clinician-researcher partnerships.

Figure 1

Clinical implementation of genomics tools. (A) In Q24 and Q25 (N = 69 for both), providers were asked how frequently they get questions from patients about genomic tests or about pharmacogenetic tests and were given a scale to indicate frequency: (a) “A lot,” (b) “Some,” (c) “Not much,” or (d) “None.”; (B) In Q18–19, providers were given a scale to indicate how routinely they used various genetic and genomic tools in their practice: (a) “Currently use,” (b) “Would use more frequently if patients requested,” (c) “Would use with timelier results,” (d) “Do not use but would like to implement,” (e) “Do not use,” (f) “Do not use and would not use,” (g) “Not applicable to my field.” Data from (b), (c), and (d) responses were combined into “No, but would” and data from (e) and (f) responses were combined into “No.” In Q19 providers were asked about “Interpretation of patients’ direct to consumer genomic test results” (N = 69) and “referral to genetic counseling” (N = 69), and in Q18 they were asked about pharmacogenetic tests (N = 70) and were provided with the examples of warfarin dosing and selection of antidepressants, genomic tests (N = 71) and were provided with the examples of genome sequencing and microarray profiling, and genetic tests (N = 70) and were provided with the examples of carrier status and diagnostics. For Q17, provider participants were given a similar scale to indicate their use of family history collection (N = 71): (a) “Currently use,” (b) “Would use more frequently if patients requested,” (c) “Do not use but would like to implement,” (d) “Do not use,” (e) “Do not use and would not use,” and (f) “Not applicable to my field.” Data from (b) and (c) responses were combined into “No, but would” and data from (d) and (e) responses were combined into “No.” Decline to respond selections were not included in the figure, but are included in the sample size given for each question. DTC, direct-to-consumer genetic test.

Figure 2

Awareness of pharmacogenetic indications. Respondents (N = 129) were asked how many drugs they were aware of with a pharmacogenetic indication in the labeling and were provided a blank field to enter a numeric answer. At the time of the survey, there were over 130 drugs reviewed by FDA with pharmacogenetic labeling information [36,37].

Figure 3

Attitudes towards return of research results to participants among researchers and non-researchers. (A) In Q14, all participants (N = 186) were asked: “In your opinion, when should genetic results obtained through research be returned?” and were asked to choose one of the following: (a) “According to participant choice,” (b) “When clinically actionable but not necessarily life-threatening (e.g., medication selection/dosing, family planning),” (c) “When clinically actionable and life-threatening,” and (d) “Never be returned.” Examples of genetic research were not provided to gain a general response to the term “genetic research.”; (B) In Q13, all participants (N = 186) were provided a 5-point Likert Scale (strongly agree-strongly disagree) to indicate their agreement with three statements: (1) “Researchers should make an effort to provide participants general information about the progress of the research,” (2) “Research participants have a right to their individual results generated from a research trial,” and (3) “Researchers should make an effort to provide participants personalized information about the research.” Decline to respond selections were not included in the figure, but are included in the sample size given for each question.