Does the Purchaser Effect Persist?
Review Supplement
Thomas Persichetti
Part of the Core Review
Capturing an infusion savings opportunity once does not establish that the same purchaser effect will persist.
The first supplement addressed whether a purchaser can move from an observed price differential to a projected and then measured savings result. This supplement asks the next question:
If that result is achieved, what determines whether it remains economically meaningful over time?
Persistence can fail for several different reasons. The alternative site can become more expensive. The treatment mix can change. The population exposed to the original opportunity can shrink or shift. And a reduction in claims retained by the purchaser does not necessarily translate into a comparable reduction in future stop-loss cost.
These are different mechanisms and should not be collapsed into a single claim that site-of-care optimization “continues to save.”
The lower cost comparison point can move
A site-of-care strategy depends partly on a relative price relationship.
EBRI's longitudinal analysis is useful here. Using a consistent set of physician-administered outpatient drugs, the median HOPD markup declined from 98% in 2019 to 70% in 2024. But EBRI found that the narrowing did not primarily result from lower HOPD reimbursement. Among the 20 drugs with the highest HOPD spending, physician-office reimbursement increased approximately 17% between 2019 and 2024.
This is where the behavior of the lower cost alternative begins to matter. A purchaser can successfully redirect care from a higher-priced site and still experience erosion of the original savings opportunity if the alternative site reprices upward.
Estimating future impacts often relies on historical point estimates projected forward assuming similar economic conditions. Historical relationships may persist for a time, particularly during early adoption or when a purchaser's baseline utilization differs materially from broader averages. But that persistence should not simply be assumed. When purchasers are paying for an effect, one of the important distinctions becomes:
Does the replacement economic relationship remain favorable over time?
For a purchaser, persistence should be evaluated against the same spending denominator used in the initial underwriting. If the original analysis assumed a particular expected unit-price reduction, subsequent measurement should show whether that differential was maintained, narrowed, or expanded.
A successful site change is not by itself evidence of a persistent price effect.
The underlying opportunity can also change
The purchaser's infusion opportunity is not static.
New biosimilars can change drug cost without changing the treatment site. EBRI's earlier work found that biosimilars generally reduced acquisition costs, but higher HOPD reimbursement could offset part of the potential savings.
Other treatment changes can alter the opportunity more fundamentally. A patient may shift from an infused therapy to an oral or otherwise differently administered treatment. A new formulation may change where or how treatment can be delivered. A therapy that once represented a substantial HOPD opportunity may therefore become a smaller site-of-care issue—or cease to be one.
The implication is important:
Persistence should not be measured only by following the original redirected cases. The purchaser should periodically re-underwrite the underlying opportunity.
The relevant spending base, HOPD share, movable share, and expected price differential can all change.
That creates an important distinction between two ideas:
Persistence of an intervention asks whether a redirected treatment remains in the alternative arrangement.
Persistence of purchaser value asks whether the economic advantage of that arrangement remains material relative to the purchaser's current spending opportunity.
The second is the more consequential question.
Stop-loss requires a separate persistence test
The RHV episode argues that expensive infusion claims can affect stop-loss economics and describes a direct connection between high-cost claims and future premium setting.
The immediate claim-level connection is plausible and belongs in the purchaser-savings analysis: if an infusion claim is reduced, the economic benefit depends on where that reduction falls relative to the employer's retained-risk boundary.
But the next step is different.
A purchaser that reduces a large claim in one year cannot assume that the same dollar reduction will appear as a lower stop-loss premium in the next year.
The RHV discussion itself creates reason for caution. Later in the episode, the guest states that stop-loss data are “very much lacking” and describes direct carrier involvement in cost-containment activity as an early emerging trend observed in only a small number of cases.
The evidentiary requirement for a stop-loss persistence claim is therefore higher than showing that an infusion claim was reduced.
A purchaser would need to understand how the carrier treated that change in underwriting: whether the expected claim was removed or reduced, whether a laser changed, whether the specific attachment point or premium changed, and whether the improvement was recognized as persistent rather than incidental.
Without that reconciliation, two distinct effects can be confused:
lower current-year claims and lower future risk-transfer cost.
The first may be observable in claims. The second requires evidence from the stop-loss renewal and underwriting process.
Persistence should be re-underwritten, not assumed
The same prospective/retrospective discipline used in the first supplement applies here.
At the initial purchase decision, the employer can make a prospective assumption about how long the price advantage, treatment pathway, and financial-risk relationship are expected to persist.
After implementation, those assumptions should be tested against subsequent experience.
Persistence depends on two things:
the purchaser must continue to have an addressable opportunity; and
the economic advantage captured by the intervention must remain intact.
The purchaser should be able to determine whether changes in the observed effect result from:
changes in the HOPD-versus-alternative price relationship;
changes in treatment or product mix;
changes in the population exposed to the opportunity;
changes in clinical or operational movability;
or changes in financial-risk arrangements such as stop-loss.
A declining site-of-care savings result does not necessarily mean the mechanism failed. The original opportunity itself may have changed.
Conversely, continuing to redirect patients does not establish that the economic effect remains the same.
Boundary with Market and Counterparty Response
This supplement stops short of asking why counterparties change their pricing or behavior.
If physician-office reimbursement rises, persistence asks whether that erodes the purchaser's previously captured advantage.
Market response asks whether the increase reflects a broader strategic response by hospitals, physicians, networks, or other counterparties and whether lost revenue is being recovered elsewhere.
Evidentiary ceiling
A demonstrated first-year site-of-care savings result does not establish a durable purchaser effect.
Persistence depends on whether the relative price advantage remains, whether the relevant treatment population and therapy mix remain comparable, whether the intervention continues to operate against a meaningful opportunity, and whether any claimed stop-loss benefit is reflected in subsequent underwriting rather than inferred from lower claims alone.
The purchaser should therefore treat persistence as a new underwriting question, not as an automatic extension of the initial savings result.
Publication version: v1.0
Generative AI assisted with drafting and editorial development. The author reviewed the source material and is responsible for the analytical judgments and final review.