Thursday, November 15, 2007

PBMs and AMP

In my comments Monday on fictitious AWPs, I stated that alternate "list price" pharmacy reimbursement models such as Wholesale Acquisition Cost (WAC) are unlikely to become widely adopted. Given some follow-up questions from readers, I want to explain why Pharmacy Benefit Managers (PBMs) should not be materially impacted by a shift in the benchmark from AWP to Average Manufacturer Price (AMP).

At least one PBM is acknowledging that WAC/AWP pricing will fade. JoAnn Reed, CFO and SVP of Finance of Medco Health Solutions (MHS), made the following comment on Medco's November 1 earnings conference call:

"On the WAC-based [sic], what we're hearing more is that it's going to AMP not to WAC, but the consultants are just making estimates. Right now no one has come up with the new benchmark, so we don't know where it's headed but for us as you know it's really no real impact to us because of our contractual language and there has been talk that it might happen in the latter part of 2008."

There are three significant points embedded in her comments:
  • AMP is likely to become the new pricing benchmark.

  • The AWP-to-AMP switch may begin in the second half of 2008.

  • PBM contracts will protect them if the benchmark changes.
Points 1 & 2 are exactly what I told you 3 months ago in The ASP Future is Here.

Point 3 is more important because she implies that contracts will be renegotiated or adjusted to preserve the original dollar-cost economic arrangements for the PBM. Thus:

PBMs will still get paid for the services they perform even if the specific compensation model changes. You can remove an intermediary but not the services provided by that intermediary. Hence, I'm skeptical of the “PBMs add no value” critics because it's at odds with the marketplace realities. The PBM’s business success reflects many individual business decisions by payers and insurers. If PBMs really added “no value,” then sophisticated payers would simply bypass them and perform the activity themselves. There are situations where this has occurred, but there has been no rush for the exits.

As long as the market for PBM services remains competitive, then the form of compensation is essentially irrelevant. PBMs are only a mild oligopoly today, at least judging by the 4-firm concentration ratio. In a June 2007 Drug Benefit News article (sorry, no link), the top 4 PBMs had 45% of total PBM covered lives. That's relatively low compared to more familiar oligopolies.

The advantages of greater PBM transparency are overrated. Consider an analogy: Imagine you are shopping for a car. You find two dealers, each of which will sell you the car for $20,000. Do you know or care if (a) dealer #1 earned its profit by marking-up the car over their cost or (b) dealer #2 earned its profit from a rebate paid by the manufacturer after the sale is made? No, of course not. You only care about the cost of the car. In other words, healthy competition about the dealers (intermediaries) provides the customer with the benefits of "transparency." (See my comments on Monday's AWP post for more on the drugs vs. cars analogy.)

I'd welcome any comments from PBM fans or critics.


  1. In 2006, one large PBM reported that 80% of their customer contracts protected them against financial loss upon a change from AWP to another pricing basis. Since most of their contracts probably have 3-year terms or less, they will be able to incorporate this protection in the other 20% at renewal by the end of 2008. Thus, if a change from AWP occurs at the end of 2008, then I agree with JoAnn Reed that there will be no impact on Medco, at least financially (other then administrative costs).

    While these percentages vary by PBM, a change from AWP should have no financial impact on any PBM under this scenario.

  2. Since you brought up the auto analogy, I would like to expand on my thoughts about the analogy.

    While an individual consumer making a car purchase may not be a good analogy to the pharmacy reference-price-based business model, I do think fleet customers operate under a very similar model. If I am a VP of Operations for, say, Pfizer, I need to purchase around 10,000 new cars every few years for my sales reps. Now you may know more about automobile fleet contracting than I, but I think that I would make my selection based upon numerous variables such as:

    • The final unit price, including any discounts and rebates
    • The dollar amount of required regular maintenance
    • The frequency of required regular maintenance (I want my salespeople in the field, not sitting in the lobby at Jiffy Lube. And I want to keep my auto maintenance staff as small as possible.)
    • The estimated reliability of the vehicle
    • Image (Probably not a good idea to purchase a vehicle made entirely outside the US if I have customers which work in or support the labor market.)

    But the point is that my final contracted price will reflect a discount from list price in some fashion. It is irrelevant if the automaker or dealer decides to author the contract with a fixed-price (with or without rebate) model, an MSRP discount model + rebate, or a cost-plus model. I will be signing the fleet contract with a contracted price at a rate far below the MSRP. Period.

    This is a nearly identical business model as the pharmacy market. Both markets have manufacturer-published prices, and average sale prices (of some sort), and rebates which are used as reference prices. Some consumers pay an amount above the reference price(s), some consumers pay an amount below the reference price(s). Very few customers, if any, actually pay an amount equal to any given reference price.

    So, again, why are automakers not being scrutinized for using a very similar business model as the pharmacy market? If you ask me, you could apply nearly every comment made in Judge Saris’s decision to the auto market as well as the pharmacy market.

    You could argue that the government and/or fleet customers “knew that the published wholesale list price was not an accurate price and was deceptive and unfair.” Or, “[automotive manufacturers] ‘unscrupulously took advantage [of the MSRP reimbursement system] by establishing secret mega-spreads between the fictitious reimbursement price they reported and the actual acquisition costs of [automotive customers].” So, why the double-standard?

    In my opinion, the analogy makes quite a bit of sense. Tell me where I am wrong.

  3. Wow, PBMGuru, you really make me work!

    No one is claiming anywhere that discounts from list are bad, wrong, evil, etc. The problem was (a) the difference between “standard” and (hidden) “mega” spreads plus (b) the separation of payer (Medicare) and buyer (physician).

    I don’t want to push the analogy too far nor do I want to comment on the legal reasoning in this particular case. I’ll just point out what Judge Saris found in her June decision, in which she writes (on page 25):

    “To recap, throughout the class period, most knowledgeable insiders understood that AWP did not reflect the average sales price to providers, but that it bore a formulaic relationship to WAC of a 20 to 25 percent markup. In addition, payors were aware there was some discounting from WAC. However, I find that in the early 1990's, payors typically did not understand that there were mega-spreads far in excess of the formulaic markup for physician-administered drugs when there was competition between therapeutic equivalents or multi-source drugs. Indeed, the named plaintiff TPPs had no knowledge or expectation as to the size of the spreads available to physicians.”

    Your analogy only makes sense if the VP of Operations buys the fleet of cars with his personal funds on behalf of his company and then asks his employer to reimburse him for the purchase. The VP might submit an expense report for MRSP minus 25%, which would look like a “standard” discount. The analogy to the AWP lawsuit would be if the car company offered a side payment to the VP that provided a “mega” windfall to him personally that was not disclosed or known to his employer (the actual payer).

    As a college professor once said, life is like an analogy…

    ‘nuff said!

  4. Don't sweat it (pun intended). I'm not directing disgust at you, but rather at the judicial, legislative, and executive branches of the state and federal government (in that order).

    "The problem was (a) the difference between 'standard' and (hidden) 'mega' spreads"

    This was exactly my point in the first post. Last I checked we don't live in a socialist economic system. Since when does the government (in this case, the judicial branch) have the right to tell the manufacturer of a product that it must keep it's spread between wholesale list price and retail list price at a certain level? This almost sounds like a "windfall profits tax" imposed on certain oil drillers/refiners.

    Last I checked a manufacturer, retailer, or service provider could set the price of their own goods and services. That is, unless you are in the oil or healthcare market. What about the rest of the commodities?

    And this pricing model benefited physicians very similarly to another profession: sales. I honestly don't know of an industry in which salespeople do not receive a commission for pushing their product. Under this model, how are physicians any different from a sales rep who is paid 100% on comission?

    "The analogy to the AWP lawsuit would be if the car company offered a side payment to the VP that provided a 'mega' windfall to him personally that was not disclosed or known to his employer"

    If the opinion is true that "payors did not understand" the pricing model it sounds like the private payors, CMS, and Medicaid all missed the boat. If this is true then what legal right do any of the payors have to claim damages?

    There isn't much I can do if I buy a car, a house, or a bottle of Lipitor as either a payor or a consumer and I decide 5 years later that I either didn't understand the pricing model or I just plain got screwed. The Romans had a phrase which I hear quite often in modern legal decisions ... caveat emptor.

    It is my own fault if I, as a payor, neglect to write my provider contracts with appropriate language prohibiting or accounting for a pricing model. I cannot go back in time and recontract if if miss a trend.

  5. Medco's analyst day was last Friday (one day after my post).

    Medco confirmed that 95% of its drug spend business is contractually protected from the economic impact of a benchmark change.

    You can listen to the analyst day webcast replay.


  6. A few thoughts on your suggestion that "the advantages of greater PBM transparency are overrated", specifically as it relates to the car-purchase analogy:

    The real reason that the car analogy does not work well is because PBMs are not providing a "product", they are providing "advice"... more specifically, they are steering members towards or away from specific branded products through their formulary strategy and their clinical interventions.

    Importantly, two branded products in the same drug class may generate very different clinical outcomes in any particular patient. In a truly transparent arrangement (one in which the PBM relies on a flat administrative fee instead of pharma rebates or pricing spread), the PBM has one primary goal/incentive: steer the member to the drug that will provide the best clinical outcome at the most efficient cost. Regardless of which branded product the member selects, the PBM earns the same amount -- the flat administrative fee.

    Introduce rebates OR pricing spread into the equation, however, and the PBM now has a very different incentive: steer the member to the branded product that will generate the largest profit for the PBM. In some cases this new incentive (maximize PBM profit) may not align with thier original incentive (maximize clinical outcomes at the most efficient cost). Take the example of the PBM that steers the member away from the lower-cost brand because they stand to earn a bigger rebate on a higher-cost product. Or, the PBM that steers a member away from a product that might be more clinically effective, because that product's manufacturer does not offer a rebate.

    It is the notion of steerage that makes the car-purchase analogy innappropriate. I've always believed that the best analogy is to compare a PBM to a stock broker (since they both make recommendations and steer their customers towards certain products).

    Let's say that you're hiring a stock broker to help you pick a mutual fund. Which arrangement would you prefer:

    a) You pay the broker a flat fee for his recommendations, and he promises to receive no other forms of revenue for your transaction.

    b) You pay the broker nothing, with the understanding that he will earn a sales credit from certain mutual fund companies for steering you to their funds.

    Option (b) is attractive, as you don't have to pay the broker an up-front fee. But consider the fact that the mutual fund company paying the sales credit to the broker is adding expenses to their fund (and therefore lowering the fund's overall ROI performance), and you'll soon realize that the broker may not be steering you to your best financial option.

    Option (a) is the financially transparent option -- and the one that makes more sense for those looking to align their stock broker's incentives with their own. Similarly, a transparent PBM arrangement will help ensure that the PBM keeps their focus on the best interests of their primary clients -- the payors and the patients.

  7. Josh,

    It has long been my opinion that a pass-through pricing arrangement is not inherently in the best interest of the client. Because the PBM does not keep any rebates (or administrative fees related to rebates) in a by-the-book pass-through arrangement, the only source of revenue is the per claim admin fee. The three inherent methods in which a PBM can increase revenue in a by-the-book pass-through pricing arrangement are:

    a.) Increase script volume,
    b.) Drive claims away from mail order and 90 days at retail (the more claims the better when the PBM gets a flat per-claim admin fee),
    c.) Do nothing (to keep overhead to a minimum).

    None of the three above options are in the client’s best interest. Quite the contrary. The PBM has no inherent interest in generic switching, therapeutic interchange, case management, or any other cost saving measures unless the client is paying for additional programs (in which compensation is based on performance). After all, why would a PBM invest the resources to administer a program which provides no revenue or profit incentive?

    I would argue that there is a much greater incentive (for both the PBM and the client) built into a traditional PBM pricing model. Current market network rates will dictate that the PBM takes a loss on at least half of retail brand claims and make little or nothing on the other half. Most pharmacies are not contracted with any given PBM at AWP-16%+, which is where the SMB market has been for the last 18-24 months. In a standard pricing arrangement there are a few opportunities for revenue:

    a.) Rebates. There is money to be made, anywhere from a few pennies per claim to $4 or $5 per claim, whether the PBM keeps all rebates, keeps a flat rebate fee, or keeps a percentage of rebates.
    b.) Spread on generics. Careful MAC management can generate anywhere from a few pennies per claim to upwards of $100 per claim.
    c.) Spread on specialty pharmacy. Careful specialty pharmacy management (on the retail side) can generate from 1% to 8% spread. Considering an average per claim cost of around $1500, even 1% can add up quickly.
    d.) Mail dispensing (if the PBM owns its own mail pharmacy). The same situation exists with brands at mail as with brands at retail: market conditions will prohibit the pharmacy from making a measurable profit. Current mail rates are from AWP-23% to AWP-24% for the SMB market, with no dispensing fees, and I don’t know of a pharmacy that can buy brands for better than AWP-25%. There may be a couple percent for the largest mail order pharmacies, but that isn’t enough to keep your doors open. Generics, on the other hand, are tremendously profitable. A typical PBM agreement will bill the client a flat AWP-50% to AWP-55% (no MAC) for generic claims. Most mail pharmacies I have seen are purchasing generics in the effective range of AWP-75% to AWP-85%.
    e.) Specialty pharmacy dispensing (if the PBM owns its own specialty pharmacy). Brand specialty products do not have the same margin problem as non-specialty products. Purchasing rates for specialty products are all over the map (depending on the product mix, demographics, etc.). Many specialty products can be had for the same discount as their non-specialty counterparts. A few are at or better than AWP-23%, a few are as low as AWP-0%. SMB market rates currently put specialty products from AWP-13% to AWP-16% (depending on the number of included services, such as case management, clinical PA, shipping, etc.).

    While, at the surface, it appears there is a tremendous conflict of interest, you must look deeper. If a PBM keeps 100% of rebates, for example, the largest per claim rebate it could expect to earn is about $5. In contrast, the AWP for an average retail generic claim will put you around $55 or $60. If the PBM is contracted with the retail pharmacy at AWP-65%, the pharmacy gets reimbursed $21 ($60 average AWP). If the PBM is contracted with the client at AWP-55%, the client pays $27 ($60 average AWP). This puts the spread for an average generic at about $6. High price generics, such as simvastatin or pravastatin, have an AWP of around $130 per Rx. Using the same pricing methodology, these retail generics will give the PBM $13 per claim in spread. At mail, with an AWP of about $375 per claim (purchasing at AWP-80% and selling at AWP-55%), the PBM stands to make over $131 per claim.

    The PBM can push it’s formulary all day long, but it is never going to get anywhere near that much in spread plus rebates on a brand drug. And, even though the PBM is making a margin of $131 per claim, it could be saving a tremendous amount over the brand equivalent. Mail AWPs for brand statins are between $260 and $500 per claim. Even with a mail discount of AWP-24% the client will pay between $197 and $380 per claim. The generic equivalent saves the client between -$9.25 and $174 per claim (depending on the drug).

    Above is a perfect example of the inherent win-win incentives built into a standard pricing arrangement which do not exist in a pass-through arrangement. This supports my original point that a pass-through pricing arrangement is not inherently in the best interest of the client.

    And, by the way, the auto analogy was not for a Client / PBM relationship, but for a manufacturer / wholesaler / dispenser relationship.