For patients with complex chronic conditions, the handoffs — from primary care to a specialist, from hospital to home, from discharge to the pharmacy counter — are where care most often breaks down. The panel’s consensus was that the technology largely exists; what is missing is ownership, know-how and a clear way to bring in partners.
Sanofi is one of those partners. The drugmaker’s health system solutions team, led by panelist Casey Reed, PharmD, works with health systems on technology that connects primary and specialty care inside the EHR. The company sponsored the Sept. 14 featured session at Becker’s 11th Annual IT + Revenue Cycle Conference: The Future of AI & Digital Health, at which five leaders discussed where coordination fails and how to fix it.
The panelists were:
- Cheng-Kai Kao, MD, chief medical information officer at UChicago Medicine in Chicago
- Danny Lee, MD, chief medical information officer at Johns Hopkins Community Physicians in Baltimore
- Casey Reed, PharmD, head of strategic solutions at Sanofi
- Donna Roach, chief digital and information officer at University of Utah Health in Salt Lake City
- Laura Zimmermann, MD, division chief of general internal medicine and senior medical director of the clinically integrated network at Rush University System for Health in Chicago
Note: Quotes have been edited for length and clarity.
1. Fragmented records and unused capability drive the biggest breakdowns
Dr. Lee said Johns Hopkins Community Physicians, which runs more than 40 primary and specialty clinics across Maryland, Washington, D.C. and Virginia, partners with many smaller practices that are not on the same EHR, and capturing their data is often challenging.
Ms. Roach said the capability usually exists but goes unused. “The interoperability, the technology is there,” she said. “It’s sometimes the lack of knowing how to do it, just the basics.” One physician-legislator, she said, turned out never to have had the function switched on in his own system.
Dr. Zimmermann said Rush’s transitions-of-care nurses reach most discharged patients for medication reconciliation and follow-up scheduling; the sliver they cannot reach, and patients whose crucial discharge prescription fails at the pharmacy over an insurance or submission error, are the ones who come back.
2. Nobody owns the next step
Dr. Kao said care remains organized around inpatient and outpatient billing rather than around the patient, so ownership of follow-up defaults to the hospitalist or the primary care physician, who is expected to know everything. This is not just challenging, he said, but close to impossible given how fragmented data are inside and outside a system.
3. The clearest wins so far are ambient AI and smarter discharge
Dr. Lee said ambient AI in ambulatory clinics has been his fastest and most successful implementation. “I constantly get feedback from providers who tell me I was going to retire early, but this kept me in the clinic,” he said.
Ms. Roach said University of Utah Health’s Heal at Home program — built through its AI innovation process — uses clinical criteria to flag discharge-ready patients that two physicians had previously been identifying by walking units with clipboards. Identification rose from one or two patients a week to an estimated 40 to 50, she said.
4. Partnerships work when a system knows its own gaps.
Dr. Zimmermann said academic systems already have decent internal processes, so the hard part is finding where a partner adds value. “It’s not about scrapping an entire workflow or process or team that’s doing fairly well,” she said. “It’s about how do you then partner to fill the gaps.”
Dr. Reed of Sanofi said the onboarding is often “like moving the Titanic around the iceberg,” and what good looks like is a defined intake path, a clinical champion and IT resources. “There is one North Star and it’s the patient,” she said.
