In a Disaggregated Model, the Processor Is the Decision That Matters Most

Key Takeaway: In a disaggregated pharmacy model, the claims processor is the most consequential decision a payer makes, because adjudication is where every formulary rule, clinical edit, and benefit provision is enforced or ignored. Processors aren't interchangeable: systems that evaluate each claim against the member's age, sex, diagnoses, and medication history catch what volume-built systems pay without question. To see the difference, payers should request a reprocessing of their historical claims, not a repricing. A repricing shows what a plan overpaid. A reprocessing shows what should never have been paid at all.

A few years ago, the president of a PBM called me with a complaint. His organization had recently moved its claims processing to our platform. His previous arrangement had ended the way too many of them do: costs climbing, questions unanswered, trust gone. Now, a few months into the new relationship, he was seeing something he didn't expect. Rejections… a lot of them.

He asked me directly: are you processing our claims incorrectly? No, I told him. His claims were being processed accurately. For the first time.

Claims stopped because the dose exceeded safe limits. Claims stopped because the drug didn't match the member's sex. These claims had been moving through his previous processor for years. Paid, invoiced, never questioned to ensure safety and accuracy. He wasn't seeing a malfunction. He was seeing what his benefit looked like when the adjudication system was actually protecting it.

Why Claims Processing Is the Overlooked Decision in Disaggregation

That call has stayed with me because his experience is repeating at scale.

His organization is a PBM, and PBMs have always contracted for processing as a distinct decision. That's exactly how he came to compare one processor against another and discover what his old one wasn't catching. Payers are now making that same decision through disaggregation, and while the model has moved quickly into the mainstream, the discipline for evaluating each component hasn't kept pace.

Rebate economics are eroding and fiduciary scrutiny is intensifying, so payers are unbundling the pharmacy benefit into whatever combination of rebates, network, formulary, clinical programs, and processing gets them the best expertise and the best value on each piece, rather than accepting one vendor's version of all of it.

The instinct's right. But in nearly every disaggregated conversation, one layer's still treated as plumbing. Payers scrutinize the rebate arrangement and debate the formulary vendor, then select the claims processor on price per claim, as if adjudication were the same everywhere. It isn't. A system built to pay claims will pay claims. A system built to question them is a different machine entirely.

What Pharmacy Claims Adjudication Actually Decides

Here's what gets missed when processing is purchased on price alone: adjudication is where the entire benefit either happens or doesn't.

Every formulary decision a P&T committee debated. Every clinical rule your team designed. Every safety edit, every benefit provision, every protection put on paper. All of it's enforced, or ignored, at the moment a claim is processed. What gets caught depends entirely on the system doing the catching.

And the difference between systems isn't subtle. Ours evaluates every claim against the member behind it: age, sex, diagnosis history, allergies, the full clinical context. Not the population average. The person. Many systems still process claims without using demographic information at all.

That difference shows up in a metric that surprises people. In legacy environments, adjudication interventions often appear in the low single digits. Across our book, 20 to 25 percent of claims require the system to apply a clinical, safety, formulary, or benefit-design control before payment is approved.

I've written before about why a paused claim is better understood as point-of-sale protection than as a rejection: duplicate therapies stopped before they compound, unsafe doses corrected before they reach a member. We didn't build a system that rejects claims. We built a filter for good claims. The filtering is the value—the protection.

Repricing vs. Reprocessing: Why the Standard Analysis Falls Short

Now for the habit I'd most like to change. When payers evaluate a new PBM, they ask for a repricing: run our historical claims through your rates and show us the spread. I understand the instinct. But a repricing asks us to process your claims exactly as your last processor did. It measures price and ignores formulary integrity, benefit design enforced and other utilization behaviors.

The better exercise is a reprocessing. Give us your claims history, your actual formulary, your actual benefit design, and let us adjudicate all of it. Not a sample. Every claim, evaluated against the member behind it.

A repricing tells you what you overpaid. A reprocessing tells you what should never have been paid at all. When we run that exercise, the inefficiencies we surface are consistently material, and they live in places a reprice never looks.

The Questions Payers Should Ask a Claims Processor

If your organization's moving toward a disaggregated model, or operates one that isn't delivering, the questions that matter rarely appear on an RFP.

Ask your PBM, current or prospective:

• How do you measure formulary accuracy at adjudication, not on paper?
• How accurate are your clinical and administrative edits?
• Will you stand behind that accuracy financially?

Because here's what more payers are about to learn as disaggregated models continue to gain traction. Operations can disaggregate. Fiduciary responsibility can't. Whoever adjudicates your claims is enforcing your benefit, every hour of every day. That makes the processor the most consequential decision in the model, and also the most fixable. Payers don't need to rebuild anything to get this right. They need to ask better questions of one vendor.

The executive who called me wasn't discovering a problem. He was having visibility, for the first time, what his benefit looked like when it was actually being enforced. Every payer stepping into this model has that same discovery ahead.

That's not a risk. That's the whole reason to unbundle.

FAQ for page

What is a disaggregated PBM model?

A disaggregated model is when a payer contracts separately for the components of a pharmacy benefit, rebates, network, formulary, clinical programs, and claims processing, instead of accepting one vendor's version of all of it. Payers often combine some pieces and separate others, choosing whatever mix delivers the best expertise and value on each component.

Why does the claims processor matter in a disaggregated model?

The claims processor is where every formulary rule, clinical edit, and safety protection is actually enforced or ignored, claim by claim. A benefit design is only as effective as the system adjudicating it. Payers often select rebate and formulary vendors carefully but choose a processor on price alone, treating adjudication as if it were the same everywhere it isn't.

What's the difference between a repricing and a reprocessing?

A repricing runs historical claims through a new processor's rates to show pricing differences. It measures price but assumes every claim was already processed correctly. A reprocessing adjudicates the full claims history against the actual formulary and benefit design, evaluating whether each claim should have been paid at all. A repricing shows what a plan overpaid; a reprocessing shows what should never have been paid in the first place.

Why do claims processors reject or pause claims?

A pharmacy claim is typically paused when it doesn't match the safety or benefit rules on file, for example, a dose outside safe limits, a therapy that doesn't match the member's clinical profile, or a drug outside the formulary. Pausing the claim is a form of point-of-sale protection, not simply a denial of care.

What should a payer ask when evaluating a claims processor?

Payers should ask how the processor measures formulary accuracy at the point of adjudication rather than on paper, how accurate its clinical and administrative edits are, and whether the processor will guarantee that accuracy financially. These questions rarely appear on a standard RFP but reveal more about processor quality than price per claim.

Is a disaggregated model right for every payer?

Disaggregation tends to benefit organizations with enough claims volume and complexity to gain from specialized expertise. Payers should evaluate whether unbundling makes sense for their size and structure before restructuring their benefit.