The Challenges of Accurately Repricing Pharmacy Benefits Manager Claims

By John Adler August 7th, 2018

Each year, benefit consultants review thousands of requests for proposal (RFPs) from pharmacy benefits managers (PBMs). The degree of sophistication and method of evaluation to verify each bidder’s “savings” varies widely, from a simple spreadsheet analysis to the classic historical claims repricing. The latter is the focus of this article, including how to accurately control the variables involved in verifying savings estimates from each of the responding bidders.


The challenge is to control the results in accuracy. Two processes must be controlled:

  1. The initial claims request to the PBMs
  2. The claims elements after the claims are received and loaded into the data warehouse for analysis


Not only should the PBM bidders be told what the rules are for the claims repricing, they should also be told what they cannot do. The claims repricing should not allow the bidders to do any of the following:

  • Substitute National Drug Codes (NDCs) unless the original NDC was discontinued
  • Substitute a smaller package size (with a lower average wholesale price, or AWP) for the package size of the original NDC, unless a contract provision is included that allows the smaller package size
  • Move medications from one National Association of Boards of Pharmacy (NABP) pharmacy to another
  • Price maintenance medications at retail 90 (90-day prescription) pricing if they have historically been purchased and priced at retail 30 (30-day prescription)
  • Move maintenance medications from either retail 30 or retail 90 to mail order
  • Substitute the AWP in effect on the day the claim was filled with a more recent or current AWP
  • Allow the application of the Brand/Generic Algorithm (BGA) to categorize products

In addition to telling the bidders what they cannot do, it is critical for the incumbent PBM to adhere to a key requirement: the claims must include the indicator that was in place when the claim was adjudicated. Since it is possible for a PBM to replace a claims indicator, this requirement eliminates any “flipping” of claims from a generic status to a brand status, thereby improving the accuracy of the repricing comparison between the incumbent’s repricing and prices provided by other bidders. Claims that are typically flipped are those subject to the proprietary BGA, DAW 5 (dispense as written code 5, which allows substitution of a generic for a brand drug) claims, and house generics.

Of note are specialty drug indicators, as there are typically wide variations from one PBM to another in what is defined as “specialty.” Since specialty drug discounts and rebates are substantially different from those for nonspecialty drugs, it is important to know and verify what is considered a specialty medication in each PBM’s repricing.

One last step before sending out the claims: have them sequenced by the data warehouse. In doing so, you are reassured that all the claims are accounted for in the bidders’ responses. This also provides a comparison of the claims that are typically excluded, including compounds, bulk powders, discontinued NDCs, invalid NDCs, and over-the-counter claims.


The analysis phase begins once the repriced claims are received from the bidders. The following steps should be taken in the scrubbing and analysis process:

  1. Verify that all claims sent are accounted for in the returned claims set.
  1. Perform a comparative analysis to be sure that the types of claims—single-source generics, multisource brands and so on—are consistent among all bidders. If there are any significant variations from the original claims set, ask the bidders to reconcile and explain those variations. This will help reveal any claims reclassification by the incumbent PB



One of the critical findings from the analysis is whether the incumbent vendor over- or underperformed against its contract pricing guarantees. Why is this important?

PBMs that offer a traditional pricing model typically underperform against their contract discount and pricing guarantees, especially for generics and retail generics. If the incumbent pricing model is traditional, the savings being shown by the bidders, including the incumbent, will be overstated.

Conversely, pass-through pricing models typically overperform against contract discount and pricing guarantees. If the incumbent pricing model is traditional, the savings being shown by the bidders will be understated if the bidders are also quoting a traditional pricing model.

In either case, these variations need to be considered and the bids normalized to create an accurate savings estimate. To accurately measure over- or underperformance, the historical claims must be run against the Medi-Span online database, which identifies generic medications and the discounts in all claim channels and guarantee categories determined. These data should then be compared to the contract pricing guarantees.

Finally, there is another consideration in the discount and pricing guarantee normalization: has the plan sponsor performed a contract pricing guarantee audit and recovered any discount deficiencies? If not, then the savings represented by each of the bidders stands as is unless the winning bidder (including the incumbent) underperforms as well—a strange twist but one worth understanding and considering.


Ideally, each bidder will assign actual rebates to every drug on an individual basis.

PBMs are not particularly fond of this practice, just as they aren’t particularly fond of assigning net unit cost to each drug that has a maximum allowable cost (MAC). Both are considered proprietary, as they provide insight into the PBM’s drug manufacturer rebate contracts and retail pharmacy MAC pricing.

If bidders refuse to assign actual rebates on a drug-by-drug basis, require that they assign the guaranteed rebate to each drug based on the channel in which it was purchased—retail 30, retail 90, mail order, or specialty.

Under either scenario, it is necessary to tally the number of brand claims in each channel to identify any significant brand claims count differences between bidders. If these differences exist, they should be reconciled, or the total rebate dollars could be overstated.


What drives the overall savings being estimated by each bidder, and when will the plan see them? Why is this question important?

If the overall savings estimate is 10 percent, and 8 percentage points of the savings is in improved rebate guarantees, the plan sponsor will not receive the bulk of the savings until nine months after the rebates are earned.

If the plan sponsor understands this, it can set an accurate expectation of the savings it will experience in its month-to-month drug spend. It also allows the benefits manager to budget properly and avoid potential budget misunderstandings with the chief financial officer. Based on this, it makes sense to calculate the net cost per script before and after rebates.

It also makes sense to calculate an overall net discount (ingredient cost + dispensing fees + ancillary fees + administrative fees – rebates) against the total AWP of the claims set. This gives the plan sponsor a single comparative metric by which to understand each bidder’s overall bid. It also gives the consultant a single metric to compare against market pricing. In the current market, this overall net discount should be in the 60–67 percent range, depending on the plan size and utilization patterns for brands/generics, retail/mail order and specialty pharmacy.


Understanding and accounting for all the variables in a PBM RFP claims repricing is complex and requires foreknowledge of what to consider. The necessary steps to create an accurate representation of savings from each bidder requires control of the historical claims, control of what the bidders can do and not do, normalization of the responses, and reports to the plan sponsor that simplify the complexity of the responses and set expectations for when the savings will be realized.

Although brief, it is hoped that this article will help advance the accuracy of the savings estimated from the PBM RFP process and help create a platform for analysis that is easily understood by the plan sponsor.

John R. Adler is a pharmacy benefit management RFP consultant with 24 years of experience in PBM. He can be reached at