A novel outreach approach for identification of familial hypercholesterolemia: Interview-based formative evaluation to improve healthcare access and quality.

Background: Familial hypercholesterolemia (FH) is a genetic condition which elevates cholesterol levels and increases risk of premature cardiac events. Medical treatment greatly reduces those risks. However, in the United States, FH is markedly underdiagnosed. We aimed to design and evaluate direct outreach and referral to specialty care for patients with an elevated risk of FH identified through application of a machine learning model and expert review of the electronic health record in a rural United States health system. Methods: Two sets of interviews: (1) seeking advice for designing outreach from a convenience sample comprising health professionals and members of the public both with and without FH and (2)) evaluating the outreach with a convenience sample of health professionals and patients who received the outreach. Two researchers conducted each interview. Thematic analysis included investigator triangulation. Results: We conducted 15 pre-outreach interviews and 32 post-outreach interviews. Most members of the public felt the outreach should be initiated by the patient's primary care clinician, while health professionals recommended outreach directly from a lipid specialist after notifying the primary care clinician. Outreach ultimately included primary care clinician notification; a mailed letter from lipid specialists to the patient indicating partnership with primary care; a message sent through the online patient portal; and a telephone call(s) from a lipid specialist to the patient. Phone calls were most impactful in prompting a clinical evaluation for FH. We identified 4 themes: (1) Both patients and clinical team members supported direct-to-patient outreach about FH; (2) Phone calls from lipid specialists to patients were considered high-value; (3) The importance of primary care team member involvement was perceived differently between pre- and post-outreach phases; and (4)) Outreach had a broader impact beyond the individual patients reached, including family screening. Innovation: This study provides key insights into the acceptable design and use of machine learning and electronic health record data for direct-to-patient outreach. Conclusions: Partnership with the target population led to direct-to-patient FH outreach that was acceptable to most recipients. High-touch engagement by lipid specialists included repeat phone calls, which maximized patients' response but are unlikely to be sustained in routine practice.