Discharge detection
When a member leaves an inpatient or observation stay, and from where.
Health Plans & Payers
Post-discharge intervention, visible to the health plan.
Claims tell a plan that a member was discharged and whether a service was billed. They are silent about what was found, what was resolved, and whether any of it changed what happened next.
Claims
What can claims data tell a health plan, and what can it not?
Claims establish that an event occurred and that a service was billed for it. They are an accurate record of transactions and a poor record of clinical work.
A claim for a transitional care management service tells a plan that a practice reported the service and met the conditions for reporting it. It does not say what the medication reconciliation found, whether a duplicate therapy was caught, whether the member understood the discharge instructions, or whether anything was escalated to the prescriber.
That gap matters most precisely where the spend is highest. A plan looking at a readmission after a billed post-discharge service cannot tell from claims whether the intervention was thorough and the readmission was unavoidable, or whether the service was minimally satisfied and the readmission was foreseeable.
The absence is structural rather than a data quality problem. Claims were built to adjudicate payment. No amount of cleaning turns them into a record of clinical judgment.
The Layer
What is an intervention data layer?
A structured record of what was actually done between the events that claims describe, captured as the work happens rather than reconstructed from billing afterward.
The unit is the intervention, not the claim. Each post-discharge episode carries the attempts made, the problems found, what was resolved, what was escalated and what remained open at the end of the period. Because it is captured in the workflow, it exists whether or not the episode ended in a billable service.
That last point is the one plans tend to find most useful. A member who was contacted three times and never reached generates no claim, and therefore no claims-based evidence that anything was attempted. In an intervention record, that member is visible, and so is the reason the episode failed.
Contact attempts
Every attempt made, when it was made, and whether it reached the member.
Medication reconciliation
What the member is actually taking against what was prescribed at discharge.
Problems identified
The clinical issues found during the intervention, captured as structured data.
Resolution
Whether each identified problem was closed, and what closed it.
Follow-up attendance
Whether the follow-up visit happened, and when.
Escalation
What was escalated, to whom, and what came back.
Action
What can a health plan do with intervention data?
Six things, in roughly the order a plan tends to want them. The first three change how resources are directed. The last three change what can be demonstrated to a regulator, a board or a delegated partner.
01
Identify
Which discharges carry the highest readmission risk, before the window closes rather than after it.
02
Target
Direct intervention to the members where it changes the outcome, instead of spreading it evenly.
03
Optimize
See which contracted practices are performing the work, and where the gaps sit.
04
Measure
Separate the members who received a complete intervention from those who received a billed one.
05
Prove
Connect the intervention to the utilization that followed, at the member level.
06
Scale
Extend the same structure across the network without rebuilding it per practice.
The Use Case
Where does transitional care management fit?
It is the natural starting point, because the population is defined by an event, the window is short and fixed, and the outcome of interest is measured in the same 30 days.
Every element a plan wants to evaluate is present in a compressed period. A discharge creates the cohort. The interactive contact requirement creates a measurable, time-stamped action. The follow-up visit creates a second one. The 30 day period creates a natural boundary for attributing what followed.
Transitional care management is also the program where the gap between billed and delivered is widest. Fewer than one in ten eligible Medicare discharges nationally receives a billed transitional care management service, so a plan working only from claims is reasoning about a small and unrepresentative slice of its own discharged population.
Note that the service is billed by the outpatient practice rather than by the plan or the hospital. A plan does not report these codes. What it gains is visibility into whether the work behind them happened across its contracted network, and what it produced.
Measurement
How should a plan measure whether an intervention worked?
Start by separating three things that are routinely reported as one: whether the service was billed, whether the work was complete, and whether the member's trajectory changed.
A completion rate answers the first two at best. It confirms a billable service occurred and, if the underlying record is structured, that its required elements were performed. It says nothing about the third.
Connecting intervention to outcome requires the intervention to be recorded at a granularity that supports comparison: which members received a complete intervention, which received a partial one, which received none, and what utilization followed each group. Where a plan holds that alongside its own claims, the comparison becomes possible.
Be careful about how the result is framed. Applying a national readmission rate to a specific attributed population, or presenting gross avoided utilization as a return, will not survive scrutiny from an actuary. The defensible version states the population, the comparison group, the period and the assumption behind every step.
Common Questions
Related questions
Can a health plan see whether a post-discharge call actually happened?
Not from claims. A claim confirms that a practice reported a service and attested to meeting its conditions. Whether contact was attempted three times and failed, or succeeded on the first call and surfaced a medication problem, produces the same claim. Seeing the difference requires the intervention to be recorded where it happens.
What is the difference between a completion rate and an outcome?
A completion rate says a service was finished. An outcome says something changed for the member. A program can report a high completion rate while producing no measurable change in utilization, and the two are frequently reported as though the first implies the second.
Does transitional care management billing data show what was done during the intervention?
No. The codes encode the level of medical decision making and the timing of the face-to-face visit. They do not encode what the medication reconciliation found, what was escalated, or what was left unresolved when the period closed.
How does this work across a delegated or contracted provider network?
The work that is uniform across practices, meaning discharge notification, the two business day contact and episode tracking, can be operated centrally. The face-to-face visit and the claim stay with the practice, because only one practitioner may report the service for a given discharge. That division is what allows a consistent record across a network without standardizing every practice's internal workflow.
Can this run alongside an existing care management vendor?
Yes, and it frequently has to. Augmentation is a narrower decision than replacement: the constraint is that the claim stays with one billing practitioner, so the work can be divided but the reporting cannot be duplicated.
Does a health plan need to change its claims systems?
No. The intervention record is additional to claims rather than a replacement for them. Its value comes from being joined to claims, which means the plan keeps its existing adjudication and reporting and gains a second, clinical view of the same members.
This page describes how post-discharge intervention data is captured and what a health plan can do with it. It does not state payment rates or projected savings. Any financial model should be built on the plan's own population and contract terms rather than on national averages.
Talk to us about your discharged population
Praventa operates transitional care programs with its own clinical team and licenses the same platform to organizations running them in-house. Tell us how your network is structured and we will tell you which model fits.