AI improves medication management and aids caregivers at Emory Healthcare

It may seem obvious that patients’ medication records should be accessible to their doctors, but that’s not always the case. While technology has made it possible for hospitals to amass a mountain of patient data, hospitals can work on better making that information easily available to clinicians at the time of treatment.


“Sharing high-quality data across our health system is more than a matter of efficiency: It’s also vital to our Epic EHR’s ability to trigger critical safety checks, such as drug interactions and allergy alerts that can help reduce adverse drug events,” said Dr. Alistair Erskine, chief information and digital officer at Emory Healthcare. “Access to medication information that was documented during any encounter with our health system empowers clinicians to make evidence-based decisions and provide the best possible patient care.

“For many patients, especially those with chronic conditions, taking medications as prescribed is vital to positive health outcomes and staying out of the hospital, while non-adherence can lead to a worsening of diseases and disorders as well as increased healthcare costs,” he continued. “Unfortunately, various obstacles can keep patients from adhering to their drug therapies, and the care team often has no way of knowing when medications aren’t being taken as prescribed.”

Emory Healthcare is committed to developing strong relationships with patients beyond the physical boundaries of the hospital. By enhancing medication prescribing and management – from a visit with the clinician to picking up a prescription at the pharmacy – it can help patients access and adhere to affordable, appropriate prescription medications for their health conditions and ailments, Erskine added.

“This is crucial for the health of individual patients and vital to population health initiatives that require providers to identify patients who may need help to stay on track with their medication therapy and intervene early to prevent negative outcomes and keep them out of the hospital,” he said.


Emory was importing medication history data into its Epic EHR via an industry-standard feed. When information was missing, clinicians filled the gaps by calling pharmacies to gather and confirm a patient’s medication list, then entered that information manually, which is time-consuming and increases the likelihood of human error.

When the provider organization does not have a comprehensive prescription fill history, it can bog down medication reconciliation and impede the ability to track adherence for population health programs that support high-risk patients.

“To address the interconnected challenges of reduced staff productivity and risks to patient safety that stem from a lack of access to accurate and actionable medication history, we turned to three medication management services offered within Fuzion by DrFirst, a software platform that uses clinical-grade AI to streamline clinical workflows,” Erskine explained.

The three services, he described, include:

  • Medication history with clinical-grade AI, providing more information than the standard feed and reducing the need for manual data entry in the Epic workflows.
  • Prescription price transparency, allowing providers to see patients’ pharmacy benefits, including their out-of-pocket costs, before they leave the doctor’s office or hospital.
  • Automated messaging, engaging patients when their clinicians send electronic prescriptions to their pharmacy.


Previously at Emory, when patient medication records were imported into the Epic EHR, prescription instructions, or sigs, often arrived as unstructured free text, often with missing pieces of information and using a variety of terms for the same instructions (for example, “by mouth” versus “orally”).

The new medication history system uses a clinical-grade AI engine to convert that data into the EHR automatically, so clinicians have more complete data at the point of care and pharmacists save time on medication reconciliation.

By deciphering clinical language, interpreting missing clinical details, and standardizing with preferred terminology into the proper fields in the patient record, the AI engine helps ensure the list of medications for a patient matches what the patient is actually taking.

“For example, if a prescription for a Lisinopril 10mg tablet includes the instruction ‘tk 1xD,’ the AI interprets the clinical intent but also sees that for this medication, a route of ‘oral’ can be safely inferred,” Erskine explained. “Because it’s integrated within our native Epic workflow, the system reduces the burden on our clinicians by delivering more information and requiring less manual data entry.

“Now that we have access to this medication history data, we also can track fill rates for these patients, pinpoint gaps in adherence and identify those who may be at risk of readmission,” he continued.

The technology also gives Emory prescribers patient-specific information on drug costs while they are face-to-face with patients.

“With prescription drug prices rising at every turn, our patients need the ability to discuss options with their doctors during office visits,” Erskine noted. “Knowing these costs upfront, clinicians can have meaningful conversations with patients about options for a more affordable therapeutic equivalent and provide a better understanding of what patients can expect to pay before they arrive at the pharmacy counter.

“We hope to help minimize the financial obstacles that keep patients from filling their prescriptions by reducing the frequency of prescription abandonment due to ‘sticker shock,’ which also can help address medication adherence challenges in our most vulnerable patient populations,” he continued.

Another way Emory is engaging patients in their care is with automated text messages that offer relevant educational information with financial savings, plus reminders to pick up their prescriptions. The HIPAA-compliant functionality allows patients to review their prescription and pharmacy information, as well as get educational information and financial savings for their medication.

“Patients can opt in to receive the prescription messaging from Emory or opt out of receiving text messaging alerts,” he said. “For prescribers, the functionality is embedded in the e-prescribing workflow in our Epic EHR, so it doesn’t require additional steps.”


Emory went live with the medication history and price transparency systems in October 2023, followed by the patient engagement messaging system in mid-November.

“In the first month, we were gathering more medication history data on more patients with fewer clicks and keystrokes by our clinicians,” Erskine reported. “Actual metrics vary across hospital sites, but initial results show

“The AI is enhancing medication history data for 81% of all available home medications so our patient records contain fewer gaps,” he continued. “We also are gathering medication history data for 92% of our patient population age 65 and older, which supports our medication reconciliation initiatives regarding high-risk patients.”


Emory Healthcare believes positive staff experiences drive positive patient experiences. By bringing together the right technologies with the right people, it aims to elevate the human component of healthcare processes, using tools such as AI to give clinicians the time and information they need to do what they do best – take care of patients, Erskine said.

“Health systems interested in implementing AI into clinical workflows should assess where it can be most effective in reducing manual processes, decreasing administrative burdens and enabling clinicians to deliver the best possible care,” he continued. “Medication management is one area that presents opportunities for improvement using intelligent automation to improve the quality of patient data as well as the efficiency of workflows.”

Follow Bill’s HIT coverage on LinkedIn: Bill Siwicki
Email him:
Healthcare IT News is a HIMSS Media publication.

Source link

This post originally appeared on TechToday.