You've done the comparison: formularies, copays, deductibles, mail-order discounts, rebate guarantees. The spreadsheet looks clean. But two plans with identical copay structures and the same drug list can deliver wildly different total costs and employee satisfaction.
The difference is the friction.
I'm talking about the operational microstructure behind every prescription fill: prior authorization thresholds, step therapy protocols, quantity limit algorithms, and real-time benefit verification. These friction points are rarely compared during procurement, yet they can move total cost and employee experience more than the formulary tiers do.
The Silent Spread of Prior Authorization
Every PBM claims to have a streamlined prior authorization process. The reality? PA criteria are anything but standardized, even within the same drug class.
- PBM A requires a PA for GLP-1 agonists only if the patient hasn't tried metformin and a sulfonylurea (two-step failure).
- PBM B requires a PA and documented cardiovascular history before covering any GLP-1.
The downstream effect isn't just administrative annoyance. For employees on PBM B, therapy initiation gets delayed by a week or more. When fills stall that long, abandonment climbs: one systematic review found 37 percent of prescription claims rejected for prior authorization end up abandoned. That abandoned script turns into a doctor's visit, a different (often more expensive) drug, and higher medical utilization.
What to ask for: PA approval rates by therapeutic class and median turnaround time for automated versus manual review. Two plans with identical formularies can still diverge on total drug spend purely from PA friction.
Step Therapy's Compliance Ripple
Step therapy is designed to steer patients toward lower-cost alternatives. But how it's implemented matters enormously.
- Plan X has a step therapy that allows a physician override via a simple checkbox on an e-prescribing platform. Override approval takes under two hours.
- Plan Y requires a phone-based prior authorization with clinical notes, 24 to 48 hours.
In Plan Y, the delay causes a non-trivial percentage of employees to simply not fill the step therapy drug, bounce to a different therapy without trying the preferred agent, or call HR to complain. Each override costs roughly $6 to $16 in administrative overhead. For a plan with 500 specialty drug claims monthly, that's $3,000 to $8,000 in hidden drag per month.
Key metric: Ask for the "first-pass yield" of step therapy, the percentage of steps resolved without human intervention. A PBM with over 80 percent auto-approval is fundamentally different from one with under 60 percent.
Quantity Limits: The Tail-Risk Amplifier
Most plans have quantity limits, 90 tablets for a 90-day supply, for instance. The hidden twist is how QL exceptions are handled.
- PBM Alpha grants a 30-day temporary QL override for patients starting a therapy with a titration schedule (like beta-blockers or proton pump inhibitors).
- PBM Beta requires a new PA for any QL override, even for dose adjustments.
When an employee gets a dose increase from 20mg to 40mg, they might need a second PA if the QL is rigid. That rigidity can lead to missed doses, clinical deterioration, and ER visits for conditions like asthma or diabetes. Comparing QL structures means checking the override mechanism, not the limit number, and whether that mechanism distinguishes between new starts, dose changes, and stable patients.
Real-Time Benefit Verification: The Zero-Clock Advantage
One systems-level differentiator almost never appears on a comparison spreadsheet: the depth of real-time benefit verification at the point of prescribing. Some PBMs push RTBV data to EHRs showing the exact patient cost after deductible and prior authorization logic. Others only show a generic list price.
The payoff is real, but the evidence is mixed. One study tied real-time benefit data to a higher fill rate (79.8 percent versus 71.7 percent) and fewer cancellations. A 2026 post hoc analysis of a cluster-randomized trial found the tool made no overall difference in fill rates. The value sits in how the tool surfaces cost at the point of prescribing and whether the clinician acts on it. When evaluating two plans, don't just check the feature list; ask for a demo. How many fields does the physician need to fill? Does it auto-populate patient-specific cost sharing? Does it show the cheapest therapy before a copay? Depth at the point of prescribing is what separates a useful tool from a checkbox.
The Transparency Paradox
Finally, a counterintuitive insight: PBMs that claim "full pass-through" pricing can still carry complex PA and step therapy criteria, because their revenue leans on administrative fees per transaction. A "traditional" PBM with rebate sharing may run simpler edits, because it profits from higher brand utilization.
A lower rebate guarantee doesn't always yield better outcomes. A plan with clean, simple edits can still lose on total cost once rebates and downstream medical spend are counted. Compare total medical plus drug cost impact, not just drug spend. Use "friction-adjusted drug cost": drug spend plus admin cost of PA, step therapy, and quantity limit overrides, plus adherence-related medical costs.
The Friction Worth Keeping
None of this means prior authorization and step therapy should disappear. These edits exist to block prescriptions that are duplicative, unsupported, or unsafe, and a single avoided high-cost therapy can save more than a year of administrative overhead. The "PA Friction Index" proposed in AJMC in 2026 separates avoidable workflow burden from the clinical edits that protect patients and plans. Judge a plan by the quality of its edits, not the quantity. An automated edit that resolves in minutes, follows published criteria, and offers a fast exception path is doing its job. A plan with no edits at all approves every prescription, appropriate or not, at full price. The target is low unnecessary friction: edits that need human review when they should not, that take days instead of minutes, or that lack any clear clinical rule.
What to Actually Ask For
Next time you compare prescription drug plans, go beyond the formulary. Request these data points from every PBM:
- Automated PA and step therapy rate by therapeutic category (not just overall)
- Median time to fill a new prescription with a PA requirement
- Quantity limit exception volume and turnaround time (target under one day for stable patients)
- Real-time benefit verification depth: does it include accumulator or adjuster logic? Specialty drug copay accumulators?
- Employee abandon rate at point of sale (for example, "did not pick up" as a percentage of new prescriptions)
These metrics are already in the PBM's operational data warehouse. They cost nothing to produce, and they reveal the true cost and experience differences that no formulary tier table ever will.
The best plan is the one that removes all the invisible speed bumps between your people and their medication.