Shades of Grey in a Binary World: Clinician Preferences for Appearance of Uncertain Results in the Electronic Medical Record
Laboratory Genetics and Genomics
-
Primary Categories:
- Laboratory Genetics
-
Secondary Categories:
- Laboratory Genetics
Introduction:
Advancements in electronic medical record (EMR) systems have enabled integration of genomic health information. Utilization for personalized health management relies on accurate clinician interpretation of genomic results. Technical limitations in how major EMR systems can discretely categorize results (i.e. “Normal” or “Abnormal”) begs the question of how variants of uncertain significance (VUS) should appear. The passage of the 2020 CARES Act has called attention to the patient perspective as genomic results may be readily accessible in the patient-facing EMR prior to disclosure by the ordering provider. Laboratories integrated with an institutional EMR must balance competing clinician values and priorities when releasing genomic results.
Methods:
A survey was circulated to genetics-trained clinicians at one academic medical center with a molecular laboratory. Questions were developed based on discussions with stakeholders (laboratory personnel, clinical providers) to assess preferences for appearance of VUS results (“Abnormal”/“Normal”) in the EMR as well as rating of priorities related to VUS review/disclosure (1 = not important at all, 5 = very important). Respondents were invited to share comments as free text.
Results:
Clinicians who responded to the survey (N=26) include genetic counselors (69.2%), geneticists (19.2%), other physicians with genetics training (7.7%), and a nurse practitioner (3.9%). Clinicians practice across specialty clinics, some within multiple clinics, most often general pediatric or adult genetics (53.8%) or cancer genetics (38.5%). Most respondents (70.6%) indicate they prefer VUS results to appear as “Abnormal” in the EMR (70.6% of genetic counselors, 60.0% of geneticists, 42.9% of all physicians). Most respondents who provide cancer genetics-related care prefer VUS to appear “Normal” (70.0%) while most who practice in general genetics prefer these to appear “Abnormal” (78.6%). Clinicians who prefer “Abnormal” appearance of VUS results indicated highest (on average) priority of ensuring potentially important uncertain results are not missed (4.6/5). Those who prefer “Normal” appearance prioritize avoiding misinterpretation of uncertain results (4.7/5) and avoiding negative emotional response in patients when viewing “abnormal” results prior to disclosure (4.4/5). Free-text responses emphasize the desire for the EMR to enable an intermediate category for a VUS, or additional documentation within the report text, to distinguish between diagnostic and non-diagnostic results.
Conclusion:
Clinicians providing genetics care at a single institution disagree on how VUS results should be discretely categorized and appear in the EMR, though given a binary choice most prefer “Abnormal”. Preference is informed in part by specialty: general genetics providers more often prefer that VUSs appear “Abnormal” while cancer genetics providers prefer these appear “Normal”, perhaps due to differences in clinical scenarios and genetic test content. For instance, evaluation of rare phenotype in general genetics may involve a focused panel where VUS findings are more likely to be relevant, while genetic evaluation of cancer (relatively more common, and less often genetic) may assess genes associated with broad tumor types where VUSs are less likely to be phenotype-specific. EMR architects and laboratories should engage with clinicians to build systems and improve reporting strategies accounting for ambiguity in interpretation. Future assessments may explore clinician preference for result appearance based on other factors (i.e. type of genetic test, number of tests ordered per year).
Advancements in electronic medical record (EMR) systems have enabled integration of genomic health information. Utilization for personalized health management relies on accurate clinician interpretation of genomic results. Technical limitations in how major EMR systems can discretely categorize results (i.e. “Normal” or “Abnormal”) begs the question of how variants of uncertain significance (VUS) should appear. The passage of the 2020 CARES Act has called attention to the patient perspective as genomic results may be readily accessible in the patient-facing EMR prior to disclosure by the ordering provider. Laboratories integrated with an institutional EMR must balance competing clinician values and priorities when releasing genomic results.
Methods:
A survey was circulated to genetics-trained clinicians at one academic medical center with a molecular laboratory. Questions were developed based on discussions with stakeholders (laboratory personnel, clinical providers) to assess preferences for appearance of VUS results (“Abnormal”/“Normal”) in the EMR as well as rating of priorities related to VUS review/disclosure (1 = not important at all, 5 = very important). Respondents were invited to share comments as free text.
Results:
Clinicians who responded to the survey (N=26) include genetic counselors (69.2%), geneticists (19.2%), other physicians with genetics training (7.7%), and a nurse practitioner (3.9%). Clinicians practice across specialty clinics, some within multiple clinics, most often general pediatric or adult genetics (53.8%) or cancer genetics (38.5%). Most respondents (70.6%) indicate they prefer VUS results to appear as “Abnormal” in the EMR (70.6% of genetic counselors, 60.0% of geneticists, 42.9% of all physicians). Most respondents who provide cancer genetics-related care prefer VUS to appear “Normal” (70.0%) while most who practice in general genetics prefer these to appear “Abnormal” (78.6%). Clinicians who prefer “Abnormal” appearance of VUS results indicated highest (on average) priority of ensuring potentially important uncertain results are not missed (4.6/5). Those who prefer “Normal” appearance prioritize avoiding misinterpretation of uncertain results (4.7/5) and avoiding negative emotional response in patients when viewing “abnormal” results prior to disclosure (4.4/5). Free-text responses emphasize the desire for the EMR to enable an intermediate category for a VUS, or additional documentation within the report text, to distinguish between diagnostic and non-diagnostic results.
Conclusion:
Clinicians providing genetics care at a single institution disagree on how VUS results should be discretely categorized and appear in the EMR, though given a binary choice most prefer “Abnormal”. Preference is informed in part by specialty: general genetics providers more often prefer that VUSs appear “Abnormal” while cancer genetics providers prefer these appear “Normal”, perhaps due to differences in clinical scenarios and genetic test content. For instance, evaluation of rare phenotype in general genetics may involve a focused panel where VUS findings are more likely to be relevant, while genetic evaluation of cancer (relatively more common, and less often genetic) may assess genes associated with broad tumor types where VUSs are less likely to be phenotype-specific. EMR architects and laboratories should engage with clinicians to build systems and improve reporting strategies accounting for ambiguity in interpretation. Future assessments may explore clinician preference for result appearance based on other factors (i.e. type of genetic test, number of tests ordered per year).