01
Insights
What determines the financial return of a transitional care management program across a population?
Most published figures describe what a single practice collects per episode. For an organization at risk for total cost of care, that is the smaller of the two returns and the less important one.
The Short Answer
Why do the usual numbers not answer this question?
Search for the economics of transitional care management and the answer comes back as a per-code reimbursement rate. That figure is accurate and it answers a practice's question, not a risk-bearing organization's.
A practice asks what it collects for performing the service. An ACO or health plan asks what changes in total cost of care when the service is performed reliably across a population, and whether that change exceeds what the program costs to run.
Those are different calculations with different inputs, and the second one is barely written about. This page describes its structure. The figures themselves are population-specific and no external source can supply them credibly.
Two Returns
What are the two components, and which dominates?
There is fee-for-service revenue and there is avoided utilization. For an organization at risk, the second is larger by a wide margin, and the first is often the reason the program is affordable at all.
Fee-for-service revenue accrues to the practices that bill the service. It is predictable, it scales linearly with captured episodes, and for a practice it is frequently the whole business case.
Avoided utilization accrues to whoever holds the risk. A readmission that does not happen removes an inpatient episode from total cost of care. Because inpatient episodes cost multiples of what the care management service pays, modest changes in readmission rate move more money than large changes in capture rate.
The practical implication is that these two returns land in different places. A program can be clearly worthwhile for an ACO while being roughly neutral for an individual practice, or the reverse. Designs that ignore this produce misaligned incentives between the organization and its practices.
The Inputs
Which variables actually determine the answer?
Five inputs drive the whole model, and they should be sourced from the organization's own data rather than from benchmarks.
Eligible discharges. Not total discharges, and not the discharges you were notified about. The gap between those three numbers is itself a finding, and it is usually larger than expected.
Capture rate. The share of eligible discharges that become completed episodes. This is the variable a program most directly controls, and it is where an underperforming program has the most room.
Baseline readmission rate. This should come from the organization's own claims, not a national figure. Applying a national rate to a specific attributed population is the most common way these models go wrong, because the populations differ more than the averages suggest.
Intervention effect. The relative reduction in readmission attributable to the program. This is the input with the widest uncertainty and the one that deserves the most conservative assumption.
Program cost. Whatever the delivery model costs, however it is procured.
Where It Lands
How does avoided utilization convert to money?
This depends entirely on the contract, and the same clinical result produces very different financial outcomes under different arrangements.
Under a shared savings arrangement, avoided cost reduces spending against a benchmark, and the organization retains a share of the difference rather than the whole of it. The share depends on the track and on quality performance, and savings are typically realized only once performance clears a minimum threshold.
Under capitation or full risk, avoided cost accrues more directly.
Under fee-for-service with no risk arrangement, avoided readmissions are a cost to the organization providing the inpatient care and a benefit to the payer. This is why the same program is a straightforward investment for a health plan and a more complicated one for a hospital without a risk contract.
None of this is visible in a per-episode reimbursement figure, which is why that figure answers so little for a risk-bearing buyer.
Diligence
What should you ask before accepting any projection?
Four questions separate a model worth relying on from a sales exhibit.
Where did the baseline readmission rate come from? If the answer is a national benchmark rather than the organization's own claims, the projection describes a hypothetical population.
Is the intervention effect stated as a relative reduction against that baseline, and is it sourced? A relative reduction applied to the wrong baseline compounds the error.
Does the model separate fee-for-service revenue from avoided cost, and does it show where each lands? A single combined figure obscures the fact that they accrue to different parties.
Does it account for the share of savings the organization actually retains, and for any threshold that must be cleared first? A gross avoided-cost figure presented as a return overstates it, sometimes by half or more.
Common Questions
Related questions
Why does this page not give a dollar figure?
Because a credible one cannot be produced without the organization's own discharge volume, baseline readmission rate and contract terms. Published figures describe other populations under other arrangements, and applying them is the most common error in this analysis.
Is fee-for-service revenue not the main benefit?
For a practice, usually yes. For an organization at risk for total cost of care, avoided utilization is larger, because an inpatient episode costs a multiple of what the care management service pays.
What is the single most common modeling error?
Applying a national readmission rate to a specific attributed population. Populations vary more than averages suggest, and every downstream figure inherits the error.
How should the intervention effect be chosen?
Conservatively, and stated as a relative reduction against your own baseline rather than an absolute percentage-point change. It is the input with the widest genuine uncertainty.
Does this change under full risk versus shared savings?
Substantially. Under shared savings the organization retains a portion of avoided cost and only above a threshold. Under capitation it accrues more directly. The clinical result is identical; the financial one is not.
Related reading
Talk it through with someone who runs these programs
Praventa operates care management programs with its own clinical team and licenses the same platform to practices running them in-house. Describe your situation and we will tell you which model fits.