What is transition of care?

Transition of care is everything that must happen when a patient moves from one setting to another: hospital to home, hospital to skilled nursing, skilled nursing to home, or emergency department back to the community.

In practice, a safe transition has five parts:

1. Notification. The people responsible for the patient know she was admitted and discharged.
2. Information. The discharge plan, medication list, and test results reach the next clinician.
3. Medication reconciliation. Someone compares what the patient was taking before admission with what she was prescribed at discharge, and resolves the differences.
4. Follow-up. The patient sees a clinician soon after discharge, in the office, at home, or virtually.
5. Engagement. The patient and caregiver understand the plan and know who to call.

Health plans already measure much of this. The HEDIS Transitions of Care measure tracks notification, receipt of discharge information, patient engagement after discharge, and medication reconciliation. Measuring a transition, however, is not the same as owning it.

What do we spend?

In 2020, 30-day readmissions of adults cost hospitals nearly $60 billion across Medicare, Medicaid, private insurance, and self-pay. Medicare alone accounted for $36.2 billion.

Source: AHRQ HCUP Statistical Brief #307, 2024. Figures are hospital costs, not payer payments, and exclude physician fees. The total of about $59.4 billion is the sum of AHRQ’s payer figures.

That number is only the readmission stay. It does not count what surrounds it:

• Emergency department returns. In Massachusetts, 26.4 percent of inpatient discharges were followed by an ED return within 30 days, and 42 percent of those returns came within the first week (CHIA, state fiscal year 2017). Nationally, 89.1 percent of unplanned Medicare readmissions came in through the ED (JAMA Network Open, 2020).
• Medications. A systematic review found about 21 percent of readmissions were medication-related, and a median of 69 percent of those were considered preventable (El Morabet et al., 2018, as summarized in a 2025 scoping review).
• Skilled nursing. The HHS Office of Inspector General found 22 percent of Medicare skilled nursing residents had an adverse event, and 59 percent of the harm events were preventable (OIG, 2014, using 2011 data).
• Missed follow-up. For half of Medicare patients readmitted within 30 days after a medical stay, there was no bill for a physician office visit in between (Jencks et al., NEJM, 2009, using 2003 to 2004 data).

The true cost of a failed transition is the readmission plus the ED visit, the duplicated test, the avoidable skilled nursing day, and the medication error. No single ledger captures it, which is exactly why no single party feels it.

Who is in control?

Everyone controls a piece of the transition. No one controls the whole of it, and the party that pays for the whole has handed most of it away.

The health plan is the only party that sees the whole member: every claim, every provider, every prior admission. Yet in most markets, the plan has delegated the transition to vendors and treated it as a quality measure to report rather than a care journey to design.

The result is a transition with many participants and no owner. Diagnosis without a next step is entertainment. So is data without a workflow.

How the health plan takes back its role

The health plan does not need to become a provider. It needs to own the intelligence and orchestrate the clinicians who deliver the care. Five moves make that real.

1. Design the transition as a care journey

Start with the member, not the measure. Define what must happen at discharge, in the first 72 hours, in the first week, and through day 30, and who is accountable at each step. A transition designed around the member produces the quality measure as a by-product. A transition designed around the measure rarely produces a healthy member.

2. Own the data

The plan already holds more of the picture than anyone else: claims history, pharmacy fills, prior admissions, network relationships, and admission and discharge notifications. The work is bringing those sources together in real time so the plan knows who was discharged today, how risky that discharge is, and what that member needs first.

3. Build the workflow, and put AI inside it

An AI orchestration layer turns data into action. It reads the admission and discharge alerts, stratifies risk, and routes each member to the right clinician in the network: the health system care team, primary care and the discharge team for medication management and follow-up, skilled nursing, a home visit when one is needed, and specialists. It then confirms the loop closed. AI does the sorting, routing, and tracking. Network clinicians do the care.

4. Aggregate the discharge plan and push it out

Today, the discharge plan lives in a hospital record and a paper folder. The plan should assemble it into one current picture of diagnoses, medications, follow-up appointments, and warning signs, and push it to three places at once:

• The patient and caregiver, in plain language, with one number to call.
• The primary care physician, before the follow-up visit, not after.
• A transition of care coordinator, who owns the member’s first 30 days and closes every open item.

5. Develop a transition of care clinic

Some members need more than a phone call and less than a hospital. A transition of care clinic, built with network clinicians, gives them a place to go. It serves four groups:

• Members seen in the emergency department and sent home, who need a next-day or next-week follow-up.
• Members leaving an inpatient stay, who need medication reconciliation and an early visit.
• Members due for wellness exams, especially those who have not seen a clinician in the past year.
• High-risk members, who need scheduled check-ups before a crisis rather than after one.

The clinic is not a new layer of cost. It is where avoidable admissions and ED visits are replaced by planned, lower-cost care.

Pulling away from avoidable admissions and unneeded ED use

The payoff of an owned transition is not a better call script. It is fewer avoidable admissions and fewer unneeded ED visits, replaced by planned care at a lower cost.

That shift depends on catching members early. When a member has a follow-up visit scheduled, a reconciled medication list, and a clinic she can reach the next day, the ED stops being the default. When the plan sees rising risk in its own data, it can route that member to a check-up before the crisis rather than paying for the admission after it.

The stakes are rising for plan partners as well. Beginning with the FY 2027 program year, CMS added Medicare Advantage patients to the Hospital Readmissions Reduction Program measures and shortened the performance period from three years to two, so penalties now reflect more recent performance (Federal Register, FY 2026 IPPS final rule). In July 2026, CMS went further, finalizing a new sepsis readmission measure that counts toward penalties beginning in FY 2030 (CMS, FY 2027 IPPS final rule). Sepsis already drives more 30-day readmissions than any other condition. Health systems now have a direct financial reason to partner with plans on transitions. Plans that arrive with data, workflow, and a clinic model will find willing partners.

Disentangling from post-acute vendors

Many plans outsourced post-acute management when skilled nursing utilization was the largest lever in the transition. Those contracts deserve a hard second look.

Plans should ask three questions of every post-acute and transitions vendor:

1. Is the return still there? Compare today’s savings against the fee, not against the baseline from the year the contract was signed.
2. Who owns the member relationship and the data? If the answer is the vendor, the plan is paying to be locked out of its own members.
3. Does the vendor’s incentive match the member’s outcome? Managing days is not the same as managing recovery.

Where the answers disappoint, the plan should bring the intelligence in-house and contract directly with network clinicians for the care. Disentangling does not mean abandoning post-acute care. It means the plan decides who goes where, based on its own data, and holds partners to outcomes.

Where to start

The health plan pays for the 30 days after discharge. It should design them, own the data behind them, and orchestrate the clinicians who deliver them.

HLTHWORKS works with health plans to make that shift in two stages:

• Review. A data, workflow, and network review that shows where members go after discharge, where the information breaks, which vendor contracts still earn their fee, and which network clinicians are ready to partner.
• Engagement. Design of the care journey and the orchestration layer, deployment with network clinicians and a transition of care clinic model, and alignment of performance targets across the plan, its partners, and its quality goals.

To start the conversation, contact HLTHWORKS at hlthworks.com.

For the physician point of view, watch Episode 9 of Redesigning Healthcare: Unfiltered, a conversation on transitions of care with Dr. Sheldon Weiss.