Transplant Nephrologists' Preferences for Clinical Decision Support for APOL1 Genetic Testing of Living Kidney Donors: A Focus Group Study

Abstract

Key Points

• Opportunities exist for clinical decision support (CDS) to help transplant nephrologists counsel living donors receiving apolipoprotein L1 (APOL1) genetic test results.

• CDS for APOL1 genetic test results should provide access to patient education materials and reminders at follow-up appointments.

• Optimal APOL1 CDS would notify providers when results are available, provide information on follow-up appointments, and provide access to patient education materials.

Podcast

This article contains a podcast at https://dts.podtrac.com/redirect.mp3/www.asn-online.org/media/podcast/K360/2023_12_01_KID0000000000000265.mp3

Apolipoprotein L1 (APOL1) gene risk variants are strongly associated with CKD,¹ particularly for patients of African ancestry (AA). Given that living donors (LDs) of AA have greater risk than LDs of European ancestry of developing CKD postdonation,^2,3 mounting evidence suggests that APOL1 risk variants may contribute to this disparity.^4,5 Considering that APOL1 risk variants may elevate risk profiles of potential LDs of AA, transplant clinicians have called for the integration of APOL1 genetic testing into LD clinical evaluation.⁶

The electronic health record (EHR) is one mechanism by which APOL1 genetic testing can be integrated into clinical practice but does not merely entail entering test results in the EHR.^7,8 An important initial step is making APOL1 test results available in a computable form to make results discoverable and accessible to providers at meaningful points in their workflow and facilitate clinical decision support (CDS) for counseling patients on the results.^9–11 Although CDS has been used successfully with APOL1 tests for hypertension,¹² no CDS has been developed to support nephrologists in using APOL1 test results for LD evaluation.

Implementing CDS correctly requires engagement and buy-in from its users. All transplant nephrologists at Northwestern University (NU, n=6) and Georgetown University (GU, n=4) were eligible for study participation. Nephrologists were recruited by email by the study principal investigator (E.J. Gordon) at NU and the research coordinator at GU. Eligible nephrologists were familiar with the research; all NU and one GU nephrologists knew the principal investigator prior to initiating the research. We conducted two focus groups November to December 2021 with five transplant nephrologists at NU and four at GU (90% participation rate) to assess optimal approaches for integrating genetic test results and nephrologists' information needs for CDS. Participants were male (n=6/9), White (n=4/9), Asian (n=4/9), and multirace (n=1/9). The focus groups were conducted as part of a larger clinical trial (ClinicalTrials.gov Identifier: NCT04910867) to integrate APOL1 genetic testing into clinical practice. Institutional Review Boards at NU (STU00214038) and GU (STUDY00003752) (which relied on NU's Institutional Review Board) approved the study, and written informed consent was obtained.

Focus groups assessed nephrologists' preferences for receiving and using APOL1 test results to optimize clinical evaluation of potential LDs and information needs through the APOL1 testing lifecycle, including the interpretation of test results and how to counsel potential LDs about high-risk results. We presented a mock-up APOL1 test report, which included the number of risk variants detected (Supplemental Figure 1), and two mock-ups of potential CDS: an EHR inbox message and a popup alert (Supplemental Figure 2). We sought feedback on mock-ups to ascertain how best to format APOL1 test result reports so that results could be discerned easily; preferences for information to be presented in reports; and preferences for how to be notified when test results become available. After the focus groups, participants completed an online survey to collect demographics. Participants were compensated $50.

Focus group audio recordings were transcribed and analyzed by thematic analysis.¹³ Transcripts were independently coded, deductively based on the test result lifecycle by the focus group comoderators (E.J. Gordon and L.V. Rasmussen) and inductively using Microsoft Word. Annotations were shared, and themes were identified until reaching thematic saturation.¹⁴ In total, six themes emerged about optimal CDS for APOL1 test results across the lifecycle. Table 1 presents representative illustrative quotations.

• (1) APOL1 test result presentation: Regarding presenting results, participants recommended that the report present the interpretation (high risk or not) and visually highlight most relevant parts (e.g., via bolding and underlining). Participants recommended unambiguous language regarding the interpretation of terms (e.g., negative and positive) and section labels (e.g., Interpretation) to succinctly indicate whether the genotype was high risk or not.

• (2) Receiving APOL1 test results: Regarding receiving test results, participants preferred the inbox message over the popup alert for being notified when test results become available. The inbox was consistent with their existing workflow for receiving results, whereas the popup alert mock-up was perceived as not having enough information to be useful.

• (3) Storage of APOL1 test results: Regarding storage, GU nephrologists recommended that the results should be clearly marked as APOL1 test results to ensure their findability among other information stored in the EHR. One NU nephrologist noted that APOL1 results stored in a discrete (computable) format could be leveraged to develop CDS for CKD.

• (4) Reviewing APOL1 test results: Regarding reviewing results, participants reported considering many clinical variables when counseling and making recommendations for a LD about donating. One nephrologist commented that nephrologists may not recall the exact risk associated with APOL1 risk variants. Participants related that transplant nurses are central to their workflow of managing test results and evaluating LD eligibility. One nephrologist noted that nurses may not have comparable familiarity with APOL1 as nephrologists.

• (5) Return of APOL1 test results: Regarding returning results, participants reported that when counseling potential LDs about APOL1 risk variants, providing hardcopy educational materials to LDs could facilitate communication. They recommended that educational materials clearly explain the risk, simplify the complexity of genetics without detracting from the core message of risk, and include visuals (i.e., diagrams).

  Nephrologists acknowledged LDs' different levels of understanding of genetics. They emphasized the importance of providing LDs with contact information of genetic counselors if in-depth counseling is needed, with two caveats. First, the shortage of genetic counselors¹⁵ means that scheduling an appointment can take a long time (it takes 6 months to schedule). Second, LDs may not have insurance coverage for the genetic counseling appointment. Therefore, participants urged providing as much information as possible during their counseling session about APOL1 while ensuring LDs have follow-up information to contact a genetic counselor.

• (6) Subsequent patient encounters: Regarding subsequent encounters, participants described how current encounters with potential LDs are limited in scope. Once the decision regarding donation has been made, long-term care recommendations are directed to the primary care physician. Participants stated they would value an active popup alert to remind them that a LD has two APOL1 risk variants but only for LD follow-up encounters postdonation.

Table 1.

Representative illustrative quotations about clinical decision support and apolipoprotein L1 test results by workflow step and key themes

Workflow Step	Themes and Illustrative Quotations
APOL1 test result presentation	Visual formatting of text “I like the way that you underlined the important phrase there, because sometimes it is a little bit overwhelming when you see a giant block of text” (GU nephrologist) “…we read a lot of lab results and it's sometimes hard to see the result so you need to make it obvious” (NU nephrologist) Clear interpretation text “…it's giving you the actual genotypic results and then it says, "high risk," I feel like that's enough” (GU nephrologist) “Results are results and an interpretation is whether there's significance to that result or not” (NU nephrologist)
Receiving APOL1 test results	Familiarity with inbox/in basket “…the only way that we use [the inbox] and I guess inbox, we see it immediately. So that's the best way.” (NU nephrologist) Pop-up alert is ineffective “…it doesn't have enough data” (GU nephrologist) “… looks like the type of stuff that I clear every day without looking at it” (GU nephrologist)
Storage of APOL1 test results	Importance of labeling results “…make sure the document was titled something like APOL1 in a way that you could type in and we could find it that way” (GU nephrologist) “So it's important to label it, I think, properly so it's easy to search for it in whatever computer system you have.” (GU nephrologist) Future potential for CDS “If you want to build the truly most awesome CDS tool of all times …[take] every identified and established risk factor associated with progression to CKD and of CKD with an APOL1 positive person based upon their BP and their age and any other… risk factors” (NU nephrologist)
Reviewing APOL1 test results	Role of transplant nurse coordinators “…should be directed to the transplant nurse coordinator of which I think there's currently two that handle LDs” (NU nephrologist) “…a lot of these [APOL1 test results] would be reviewed by nurse coordinators” (GU nephrologist) APOL1 in LD evaluation “…one tiny piece of a river of information coming at [nephrologists]” (NU nephrologist) “…we still interpret the results in the context of the patient, who they are and what's going on” (NU nephrologist) “…this [APOL1] test is not be-all and end-all” (GU nephrologist) Clinician familiarity with APOL1 “…I don't also think about APOL1 every day. So I don't necessarily remember off the top of my head what the exact risk increases with two high risk alleles versus one versus zero.” (NU nephrologist) “…a lot of these [APOL1 results] would be reviewed by nurse coordinators. So just keep them in the back of your mind. And they might not have the level of sophistication that some of our training would have.” (GU nephrologist)
Return of APOL1 test results	Importance of educational materials “…patients don't have the capacity to absorb all the information during the discussion … so that we often send patients with educational materials or resources to read” (GU nephrologist) Use of visuals to convey risk “Risk is a hard thing to explain to people” (NU nephrologist) “…visuals always work way better than talking goes” (GU nephrologist) “A visual abstract of APOL1 risk to a potential LD, literally that simple, that would be huge” (NU nephrologist Need for printed materials “I personally think you can print it out so you can hand it to them but it could be on the computer” (NU nephrologist) “…if we can give [patients] something to leave with like a resource page that you come back to, that would be fantastic” (GU nephrologist) Contact information for genetic counseling “…having information available for referring to genetic counselors, if necessary” (GU nephrologist) “…how to contact…the genetic counselors if the patient wants the genetic counselors, if we have some contact information or where to call” (NU nephrologist) Providing information for primary care providers (PCPs) “We often will send some information back to their PCP if they have some lesion on their kidney, they can't donate or something else that comes up, we usually try to make sure that their PCP, particularly if there's something to follow up on” (NU nephrologist) “You usually have to individualize it, i.e., Sometimes you print out information and try to fax it to a doctor's office or something” (NU nephrologist) “If you're going to be saying this to a PCP, which again, it might cause some angst with a PCP, you might need a cover sheet that just explains what the test result actually means and what it doesn't mean” (GU nephrologist) “A cover sheet that says follow the BP, take the medicine” (GU nephrologist) Use of EHR templates for documentation "…it would be helpful to standardize what we put in the charts” (NU nephrologist) “…having a dot phrase to do that in a standardized fashion probably would be faster and more reliable” (NU nephrologist)
Subsequent patient encounters	Limited relevance “I guess maybe the questions are would we do anything different with this particular donor from our end? So, no reason to really, for us, to take that or I'm not sure.” (GU nephrologist) “[t]hat information popped up for that. It reminded you, ‘oh yeah. Miss Smith donated 2 years ago. That's two high risk alleles. She's here for follow up. Would that change what we would do?’” (NU nephrologist)

APOL1, apolipoprotein L1; GU, Georgetown University; NU, Northwestern University; CDS, clinical decision support; LD, living donor; PCP, primary care physician; EHR, electronic health record.

As the first study to assess optimal approaches for integrating APOL1 genetic test results into the EHR, we identified several novel factors that may aid transplant programs in this process. Both participating institutions had established workflows by which transplant nurses review and summarize relevant EHR data points for the transplant nephrologist. Our primary finding was that transplant nephrologists perceived no clear benefit to replacing this workflow and the role of the transplant nurses with CDS. Given the myriad of factors influencing nephrologists' decisions about LD eligibility, the decision to recommend donation may not be easily summarized or calculated by automated CDS. Instead, optimizing information display and information retrieval can support transplant nephrologists in making this decision effectively.

Participants suggested providing LDs educational materials that effectively convey risks associated with high-risk APOL1 genotypes. The implication for the implementation of CDS is ensuring these materials are readily available during clinical workflow, which passive CDS can accomplish.

Implementing these CDS recommendations may facilitate the integration of APOL1 testing into clinical practice. Although our sample size was small, qualitative research aims to identify the breadth of a phenomenon, rather than be generalizable. The concordance in recommendations across focus groups suggests the findings have application in other transplant centers. Future research should replicate our study, evaluate the impact of these recommendations on clinical practice for APOL1 testing, assess transplant nurses' CDS preferences regarding APOL1 test results, and identify additional decision support needs for pretest workflows.

Supplementary Material

kidney360-4-1610-s001.pdf ^{(169.6KB, pdf)}

Disclosures

E.J. Gordon reports the following: Ownership Interest: My husband is a partner in Halock Security Labs. This is *not* related to health care; Honoraria: Honoraria and travel reimbursements for presentations and meetings: ACBTSA; NHLBI DSMB; NIAID DSMB; NIH ad hoc study section grant reviewer; I also received hotel and per diem paid for by CareDx in March 2022 and in 2023; I received hotel and per diem paid for by the National Kidney Foundation in April 2022; I received an honorarium for a grand rounds presentation from University of Rhode Island in February 2022, from the University of New Mexico in 2023, and from Fresenius in September 2022; Advisory or Leadership Role: Associate Editor for Narrative Inquiry in Bioethics; Associate Editor for American Journal of Transplantation; Associate Editor for Kidney Medicine; Chair of the AST PSECOP; Member, NHLBI DSMB; Member, ACBTSA; Member, NIAID DSMB; and Member of the ESKD Network Board;and Other Interests or Relationships: CareDx sponsored a Consortium led by Duke University, covered dinners for Consortium members; The NKF hosted a workshop in Washington, DC on xenotransplant clinical trials in 2022. NKF covered my travel expenses. Makana and eGenesis provided funds to NKF for the conference, but did not compensate me directly. All remaining authors have nothing to disclose.

Funding

E.J. Gordon: National Institute of Diabetes and Digestive and Kidney Diseases (R01DK128207, to E.J. Gordon, PI) and Northwestern Memorial Hospital Dixon Translational Research Innovation Award (to EJ Gordon, PI). L.V. Rasmussen: National Institute of Diabetes and Digestive and Kidney Diseases (R01DK128207) A.H. Agrawal: National Institute of Diabetes and Digestive and Kidney Diseases (R01DK128207).

Author Contributions

Conceptualization: Akansha H. Agrawal, Elisa J. Gordon, Luke V. Rasmussen.

Data curation: Elisa J. Gordon, Luke V. Rasmussen.

Formal analysis: Elisa J. Gordon, Luke V. Rasmussen.

Funding acquisition: Elisa J. Gordon.

Investigation: Elisa J. Gordon.

Methodology: Elisa J. Gordon, Luke V. Rasmussen.

Project administration: Elisa J. Gordon.

Resources: Elisa J. Gordon.

Supervision: Elisa J. Gordon.

Writing – original draft: Elisa J. Gordon, Luke V. Rasmussen.

Writing – review & editing: Akansha H. Agrawal, Elisa J. Gordon, Luke V. Rasmussen.

Data Sharing Statement

Data are available upon reasonable request by the corresponding author.

Supplemental Material

This article contains the following supplemental material online at http://links.lww.com/KN9/A394.

Supplemental Figure 1. Mock-up of an APOL1 genetic test report with two risk variants.

Supplemental Figure 2. Mock-up display of two displays for alerting transplant nephrologists to the presence of APOL1 genetic test result: (A) an inbox message showing an APOL1 test result and (B) a popup alert using pieces of information from the APOL1 test result.