When Penny Patient visited her primary care physician diagnosed her with heart failure and diabetes. Grim news for sure. But there were treatment plans and specialists who could mitigate her health situation. Over an eight week period, Penny will travel from her primary care office to emergency rooms to specialists, but her medical records will take a more complicated journey, sometimes spending time in places Penny could never go.
Penny’s primary care office
Let’s start at the beginning. Penny’s initial visit to her primary care doctor birthed her EHR. The Electronic Health Record (EHR) began circulating in the early 1990s. It wasn’t until 2004, at the creation of the Office of the National Coordinator (ONC) of Health Information Technology (IT) that EHRs became regulated.
Penny’s record, that day at her doctor’s office, was created inside her physician’s network. Later, when the record moved outside of that office, it would do so under the safe and trusted oversight of TEFCA. Her records included her list of complaints, her medications, and a faxed copy of her 2023 echo report. And this information was sufficient for her primary doctor to diagnose and treat her. Penny’s EHR remained completely at home within its home system.
Penny’s 2 a.m. emergency room visit
One night, Penny had trouble breathing. She tried to calm herself down, considering the hassle of getting up and dressed and driving herself to the hospital. Would she end up spending the two hours and 42 minutes most Americans average in an ER waiting room?
Would she wait all that time only to be only to be patted on the head and sent home? Penny overcame these objections because the shortness of breath was alarming enough to urge her to drive to the ER.
Her records may not be readable. In the most recent national hospital survey, 71% of hospitals said needed clinical information from outside providers was routinely available electronically at the point of care, while 42% said clinicians often used it. The gap between those two numbers is where Penny’s record spends most of the night.
Penny is admitted, adding volumes to her EHR
The ER doctor decided Penny's condition warranted admission. The doctor records her decision as an event in Penny’s EHR and an ADT (admission, discharge, transfer). This hospital adventure has changed things for Penny’s EHR.
Naturally, Penny’s hospital stay brings a barrage of medication changes. The doctor ordered three new drugs started and two previous medications withheld. Additionally, Penny’s diuretic dose was adjusted twice during her stay. The following day, around 4:00 pm, the doctor ordered her discharge summary which was then added to her EHR.
The skilled nursing facility
Penny leaves the hospital on Friday evening for a Skilled Nursing Facility bed across town. Her records are supposed to follow her. To be fair, some of them do.
But this handoff is where records go todie. Medication discrepancies are common at every transition and worst here: studies put them at 31% to 67% on hospital admission, 36% to 63% at hospital discharge, and roughly 75% at SNF admission. Penny's medication list left the hospital with three new drugs, two held, and a diuretic adjusted twice. The nurse admitting her at the SNF is working from whatever arrived, which may be a printed packet, a faxed summary, or a phone call with the covering hospitalist on a Friday night.
Part of the problem is structural. Even among hospitals that routinely exchange data across all four domains, only about a quarter sent electronic summary of care records to most or all of their external long-term post-acute care and behavioral health partners.
The EHR road is paved right up to the settings where the sickest patients go, and then it turns to gravel. Penny’s records are now traveling on gravel, and so is everything the hospital learned about her over the past four days.
Back home to her primary care
Penny’s primary care physician should learn about the admission without being told by Penny. Since 2021, hospitals using conformant electronic systems have been required under Medicare Conditions of Participation at 42 CFR 482.24(d) to send ADT notifications to post-acute and community providers, including the patient's primary care practitioner.
Whether that notification lands somewhere a human can catch it is a separate question. The same hit or miss pattern shows up in where patients end up. A study of 275,189 Medicare admission-readmission pairs showed that patients readmitted to a hospital other than the one that discharged them were more likely to go home with home health rather than to a nursing facility when the two hospitals shared a health information exchange, with 9% to 15% higher odds. The clinical picture was the same in both cases. What differed was whether the second hospital could read the record the first one wrote.
Stop six: the organization accountable for the outcome
Penny is back home now, happily recovering. But her record is still on the move. Separately from TEFCA, under the CMS Interoperability Framework, payers and value-based care organizations can query CMS Aligned Networks for specific quality data elements needed for payment and health care operations, along with clinical data tied to recent claims. Twenty-one networks pledged to meet the framework criteria when the initiative launched in July 2025. By the first wave launch in April 2026, more than 700 organizations had pledged support.
Seven hundred organizations pledging is a supply-side answer. It settles whether Penny's record can leave the buildings that hold it. What it does not settle is whether the next clinician opens it, reads it, and changes what happens next. Penny's record can now travel farther than it ever has. Whether anyone is waiting, or whether the data that arrives is useful, that’s the part that’s still being built. Not all networks are created equally. Penny’s data must travel along trusted networks that de-dupe, normalize and identity match in order to fulfill the possibilities that interoperability presents.
What did Penny’s record journey teach us?
Penny's record left the hospital. That part worked. An ADT fired when she was admitted, a summary of care was generated when she left, and both traveled the paths that fifteen years of policy built for them. If the measure of interoperability is whether a record can get out of the building, Penny's case is a success.
But what didn’t happen? The SNF nurse admitting her on Friday night had a medication list she could not reconcile against anything. The hospital that readmitted her three weeks later had a record it could open, which is why she went home with home health instead of to a bed. The doctor who eventually saw her had a discharge summary somewhere in a chart, written at 4 p.m. on the Friday of her discharge, one document among the hundreds that arrived that week.
Penny’s data was on the move but was it moving to where it was needed and did arrive with usable information?
Absorption is whether the person on the receiving end can act on what shows up. That is a demand-side problem, regulations currently don’t have that kind of reach. A record delivered in eleven seconds into a workflow with no time to open it has the same clinical value as a fax nobody pulled off the machine.
Penny's record arrived translated, complete, and on time. They were available. The last five years solved the translation. Translation is everything that has to happen for a record to arrive in a form the receiving system can hold so it can be read.
Reading is what a human has to do once it is there. It’s the absorption piece. Open the document. Find the relevant part of it. Compare it against what is already in the chart. The next five are about improving the reading.