Implementation of an automated EMR workflow to facilitate genetic consultation and review of inpatient genetic test requests

Inpatient genetic testing is generally only appropriate when the results will change the management of patients while they are in the hospital. When test results are available after a patient has been discharged, it can be unclear as to which provider is responsible for following up on the results.

Inpatient genetic testing can also create a financial burden to health systems since these costs are not typically covered by current reimbursement practices. A review of select inpatient genetic test orders at UCLA from 2020 to 2022 demonstrated that approximately 80% of the results were not available until after the patient was discharged, so we wanted to implement a workflow whereby select inpatient genetic test requests would be reviewed for appropriateness before the actual test orders could be signed.

Designing and Launching the Electronic Consultation Workflow

We developed an automated electronic consultation workflow within our electronic medical record (EMR) (Epic) that blocks providers from signing an order when they attempt to select a list of inpatient genetic test orders. In these cases, the EMR shows an alert requiring providers to consult the molecular diagnostics laboratories first; for the consult, the provider enters the desired test, the reason for ordering it, and a contact phone or pager number. After the provider signs the consult order, the molecular lab director team is notified by an e-mail with the information.

The molecular lab director on service then contacts the provider by phone by the close of the next business day to discuss whether the test request is appropriate for the inpatient setting. If it is determined to be appropriate, the molecular lab director then signs a note in the EMR that allows the provider to sign the genetic test order by suppressing the alert. If the order is deferred to the outpatient setting, then we recommend the provider document ordering the test in the discharge summary.

We launched this workflow in October 2023 and applied it to seven specific germline genetic test orders (both in-house and sendout). Additional sendout test requests to four dedicated genetics labs were added to the workflow in February 2024. Based on previous ordering practices, the molecular diagnostics group expected approximately two consult requests per week. Consultations with providers focused on two main questions: 1) when is the patient expected to be discharged? and 2) how will the results from the test affect clinical management during admission?

Evaluating the Workflow and Outcomes

In the first seven months since the workflow was launched (up until June 3, 2024), we received 60 consult requests. This is approximately two per week, which is similar to the pre-intervention level. Other test orders have also presumably been prevented where the provider ultimately elected not to order the consult, though this can be difficult to determine.

The types of tests that providers requested are shown in Figure 1. As expected, trainees (residents or fellows) initiated the majority (64%) of the inpatient consults thus far. Approximately 18% of them have been initiated by clinical genetic counselors. Overall, genetics providers have been far more likely to order the inpatient genetic tests appropriately as compared to non-genetics providers, and their requests were less likely to be cancelled (see Table 1). In contrast, many of the requests involving non-genetics providers were canceled/deferred to the outpatient setting, with some of those requests being auto-canceled (due to the patient being discharged before the request could be discussed).

This automated EMR workflow has presumably reduced inappropriate inpatient genetic test orders, though some providers may have ended up ordering tests directly with the reference laboratory using their provider-facing portal outside of Epic. This is information we don’t always have easy access to, and we have already initiated several efforts to reduce direct portal ordering as much as possible in the future.

Moreover, many educational discussions have occurred between providers and molecular lab directors during the time in which the workflow has been live; however, only a few clinical users have been involved in more than one consult, making it difficult to determine how much of an educational impact the intervention has had overall.

We are currently evaluating whether certain specialties, such as clinical genetics, should have the consult alert suppressed, as most of their requests have thus far been deemed appropriate. Other specialties could also have the consult alert suppressed in the future as needed. Revising the requirements for specific specialties is preferred to revising the requirements for specific providers, as the latter option would likely require additional ongoing monitoring and maintenance as providers and trainees come and go.

However, some clinical genetics providers have expressed a preference to retain the alerts for their orders too, as they are sometimes consulted to order tests for other services who are under the assumption that clinical genetics is already excluded from this workflow. Many of those requests are for tests that clinical genetics would not recommend ordering during the inpatient admission, and by keeping the alert in place, clinical genetics should receive fewer inappropriate requests from other providers to place those orders.

The laboratory also regularly reviews a dashboard to evaluate changes in the volumes of specific inpatient genetic test orders, which can help identify additional tests to be added to this workflow in the future. As was previously mentioned, many germline genetic tests are currently being sent directly from clinics to reference laboratories, and one of our overall goals is to continue building discrete tests and integrating additional laboratories into our EMR (e.g., through Epic Aura). As those efforts continue, we will establish better visibility of test utilization, and those additional tests can
also be considered for inclusion in the inpatient genetic testing consult workflow.

Josh Deignan, PhD, FACMG is a clinical professor of pathology and laboratory medicine at the David Geffen School of Medicine at UCLA and associate director UCLA Molecular Diagnostics Laboratories in Los Angeles, California. +Email: [email protected]