During Becker’s Health IT + Digital Health + RCM Meeting, in a session sponsored by Amgen, clinical pharmacy specialist Ralph J. Riello III, PharmD of Yale University School of Medicine, discussed the utility of EHR alerts in bridging this gap.
Lack of guideline adherence is one of the main causes of care gaps
The time lag between new clinical guidelines and their integration into routine clinical practice can have critical consequences for patients. For example, a recent retrospective cohort study of pharmacy and medical claims data in the U.S. showed that only about 50% of patients with atherosclerotic cardiovascular disease, such those with a history of myocardial infarction, are prescribed any statin therapy, while only 22.5% are actually prescribed the recommended high-intensity statin dosing.
“That implementation gap is a disservice to our patients when lifesaving therapies are available but don’t reach patients in a timely manner,” Dr. Riello said. “How can we condense that timeframe and disseminate such therapies sooner to help improve health outcomes?”
With that question in mind, a cardiology group at Yale University School of Medicine, including Dr. Riello, explored whether EHR alerts could help guide prescriber behavior toward the practice of evidence-based medicine and launched the PRagmatic Trial Of Messaging to Providers About Treatment of Heart Failure (PROMPT-HF) clinical trial.
The PROMPT-HF trial demonstrated the efficacy of EHR alerts in improving guideline adherence
The initiative was an EHR-embedded trial designed to examine the efficacy of an EHR-based alert system that informs front-line physicians about what evidence-based medications they can prescribe for patients with heart failure with reduced ejection fraction, compared to usual care that does not incorporate such alerts. It cluster-randomized 100 internal medicine and cardiology providers, who in turn recruited 1,300 patients and split them between EHR alert exposure and usual care.
The EHR alerts, referred to as best practice advisory or BPAs, displayed patient-specific clinical data, currently prescribed heart failure medications and indicated but omitted heart failure therapies. To reduce workflow interruption and alert fatigue, they were designed to pop up only during face-to-face clinical encounters when physicians were reviewing patients’ current treatment plans. Based on feedback from cardiologists, Dr. Riello said it was “the perfect time in the clinical workflow to coach or prompt clinicians to make the right decisions for patients at the point-of-care.”
The results of the trial showed that prescribers who viewed the EHR alerts were 41% more likely to prescribe an additional guideline-directed medical therapy class or a number-needed-to-nudge of about 9. Unlike prior EHR alert studies where BPAs are frequently ignored, 25% of clinicians accepted the recommendation exactly as the prompt suggested and nearly 80% felt the alert was effective or very effective at improving heart failure care.
In focus groups with participating physicians aimed at eliciting feedback about the alerts, one practitioner said the timing of the alerts, which appear only when the provider is looking at the patient’s medication chart, made them much more inclined to agree with the recommended therapy.
A second trial, called PROMPT-LIPID, confirmed the effectiveness of EHR alerts for secondary prevention of atherosclerosis cardiovascular disease patients, where prescription of high-intensity statins and other evidence-based lipid lowering therapy can be truly lifesaving. Unfortunately, use of these agents to achieve target blood cholesterol goals among very high-risk patients are often neglected in real-world practice. The PROMPT-Lipid alert, however, increased the likelihood for prescribers to intensify lipid lowering therapy by a similar 40% margin as observed in the prior PROMPT-HF study. Moreover, prescribers who actively engaged with the lipid BPA, rather than immediately dismissing the pop up, were even more than twice as likely to prescribe the recommended lipid medications and up to five times more likely to prescribe non-statin therapies like monoclonal antibody PCSK9 inhibitors.
Dr. Riello said that while current research around EHR-based alerts is focused in the cardiorenal and metabolic space, there is no reason why this approach cannot be expanded to other diseases with significant care gaps, such as osteoporosis, musculoskeletal diseases, pain management and other chronic diseases.
IT departments can be a crucial ally in implementing an EHR alerts system
For healthcare organizations interested in implementing an EHR-based alerts system like Yale’s, the ideal approach is through an investigational pathway linked to a research grant or start-up funding. That ensures the organization can properly compensate the team members designing and implementing the system, as well as permit enough protected time internally for them to do it right. (Dr. Riello’s group developed the program thanks to research funds provided by Amgen.)
However, for institutions that may not have grant funding to design, test, pilot and deploy such systems from scratch, the IT department can be a critical partner. This is because IT departments usually have a long list of support tickets to attend to, and submitting a request for support with deploying an EHR alerts system may not take priority over other requests. Still, making a well-articulated business case requesting IT support to scale such interventions enterprise wide can occasionally help move things along.
“Through an investigational pathway, you catapult to the top of the list of pending Epic or Cerner tickets for quality improvement,” Dr. Riello said in response to a question from a session attendee whose organization had a cardiology EHR alerts program that was well-received internally, and that the department of vascular surgery wanted to replicate but did not know how to go about it. “Without such a pathway, though, and unless you know somebody in IT, it’s hard to jump that list. It takes a village to do this stuff.”
