Identifying individuals with inherited cancer susceptibility is critical for targeted cancer prevention, screening, and treatment. Strategies to assess the genetic risk of unaffected individuals are needed. Scalable and sustainable methods to automatically extract and analyze family history information routinely captured in the electronic health record (EHR) can identify primary care patients appropriate for cancer genetic services. Increased patient ascertainment needs to be paired with implementation studies to compare models of delivering genetic services, including patient-directed models. Because access to services continues to be a barrier for those from minority racial and ethnic groups and rural areas, examining responses to different delivery models across population subgroups is essential. This study will employ an implementation science framework to test a replicable EHR-based clinical decision support (CDS) infrastructure to: (i) automatically identify unaffected patients from 48 primary care clinics in two healthcare systems, University of Utah and New York University, who qualify for cancer genetic services (Aim 1); and (ii) compare two models of genetic services delivery for 1,920 primary care patients using a randomized trial design with clinic-level randomization (Aims 2 and 3). We hypothesize that the CDS infrastructure will identify additional patients who have not been previously referred (Aim 1) and that uptake of genetic testing (Aim 2) and adherence to management recommendations (Aim 3) will be equivalent between the models. To address Aim 1, we will evaluate whether the CDS approach identifies patients who have not previously been referred, and whether this varies by race/ethnicity and rurality. To address Aim 2, we will compare: a patient-directed model in which those identified by the CDS infrastructure as meeting testing criteria will be informed of their cancer risks, provided with educational resources, and offered the option to select genetic testing through a patient portal to an enhanced standard of care model in which providers and patients are notified through CDS when criteria are met and of the availability of standard of care genetic counseling. We will compare uptake of genetic testing by model and whether this differs by race/ethnicity and rurality.
In Aim 3, we will compare the effects of the two delivery models on adherence to recommendations 12 months after return of results, examining differences in effects by race/ethnicity and rurality. Innovative features include implementation of population-based CDS assessment of family history information available in the EHR; comparison of outcomes of patient-directed and enhanced standard of care delivery models; and focus on impact of race/ethnicity and rurality. This highly impactful translational research builds on our unique strengths in cancer genetics, clinical informatics, and population sciences, and addresses issues of immediate clinical significance, including increasing hereditary cancer genetic testing in appropriate patients and improving access for underserved groups.
Identifying individuals with inherited cancer susceptibility is critical for targeted cancer prevention, screening, and treatment, and strategies to assess the genetic risk of unaffected individuals before cancer occurs are needed. This study will employ an implementation science framework to test a replicable electronic health record-based clinical decision support infrastructure to automatically identify unaffected patients from 48 primary care clinics in two healthcare systems, University of Utah and New York University, who qualify for cancer genetic services (Aim 1); and, using a randomized trial design with clinic-level randomization, compare the effects of two models of cancer genetic services delivery on uptake of genetic testing (Aim 2) and adherence to management recommendations after 12 months (Aim 3) among 1,920 primary care patients.
