Wednesday, October 20, 2010

Pharmacy Reimbursement Drops Again…or Does It?

The Pharmacy Benefit Management Institute (PBMI) just released its 2010-11 Prescription Drug Benefit Cost and Plan Design Survey, an informative and useful report that you can download for free. A hearty Drug Channels “Thank You” to Takeda Pharmaceuticals North America for sponsoring the research.

The latest survey shows that retail pharmacy payment tightened significantly in 2010, reinforcing the fact that efficient, low-cost dispensing is crucial to a retail pharmacy’s economic survival.

But here’s the unexpected conclusion from my number-crunching of the latest PBMI data: Retail pharmacy profits from brand-name prescriptions have remained remarkably stable over the past 10 years. The apparently steep drop in third-party reimbursements to retail pharmacies primarily reflects mathematical adjustments to underlying pharmaceutical price inflation.

I expect channel participants will strongly disagree, so read on and make up your own mind.


The PBMI survey data are collected from employers, not from PBMs. The 2010-11 edition includes responses from 372 employers representing 5.8 million members—a decline from the 417 employers with 7.0 million members in last year’s report. The average group size in the 2010-11 data is 9,736 members, but this is a bit misleading because 44% of the employers have plans with fewer than 2,000 members. Survey methodology wonks can see the Profile of Respondents for more details.

As far as I know, PBMI publishes the only report tracking pharmacy reimbursement rates in a consistent way over multiple years, although the sample base changes over time. I focus on retail pharmacy reimbursement for brand-name drugs in this article.


Most private and government third-party payers still use the Average Wholesale Price (AWP) benchmark to estimate a pharmacy’s ingredient costs for a brand-name drug. The PBMI survey documents that private third-party payers (employers) are reimbursing pharmacies at an ever-smaller share of AWP.

As the chart below shows, the average discount from AWP has declined by four percentage points (400 basis points) since 2000 for brand-name drugs dispensed by a retail pharmacy. (source) There's been an especially sharp drop of 100 basis points since last year.

The PBMI data also shows the dispensing fee dropping by 30% from $2.31 in 2000 to $1.62 in 2010.


AWP has a mathematical (but legally tangled) relationship to Wholesale Acquisition Cost (WAC), which represents the manufacturer’s list price as defined in Section 1847A(c)(6) of the Social Security Act.

Numerous studies have shown that manufacturer’s list prices—and therefore AWP—have been increasing more quickly than overall inflation. Therefore, deeper discounts from AWP are not necessarily bad for pharmacies given the increase in overall drug prices.

Put another way, pharmacies now receive a smaller percentage of a bigger number. How much bigger? According to data published in the 2010-11 NACDS Chain Pharmacy Industry Profile, the average retail (non-mail) prescription price has increased from $65 in 2000 to $155 in 2009—a total increase of 138% and an average annual growth rate of 10.1%.


In addition to a dispensing fee, retail pharmacies earn spreads between (a) the ingredient cost reimbursement that a pharmacy gets from a third-party payer, and (b) the pharmacy’s net acquisition cost for purchasing the product. This “spread pricing” provides retail pharmacies with more than 80% of their gross profits from prescriptions because pharmacies acquire drugs for less than the ingredient cost reimbursement amount.

To model the profit impact implied by the PBMI data, I combined the NACDS-reported prescription price data and the PMBI reimbursement data to compute a per-script AWP estimate. The table below shows my computations for the average gross profit and gross margin per brand prescription. These are averages, so YMMV. (Click the chart to enlarge it.)

The surprising result? Gross profits per brand prescription have been fairly stable, remaining in the range of about $9 to $11 during the past ten years. Estimated average gross profits per brand drug prescription have even increased over the past decade. This result makes sense given studies suggesting the average cost-of-dispensing is slightly below these figures.

Of course, gross margin—gross profit as a percentage of revenue—has declined, but this is simply a computational artifact of rising prescription prices.

Note that the gross profit dollar computations are sensitive to the particular assumptions for column [E]. The absolute dollar profit level would be higher (lower) with lower (higher) acquisition costs. However, the general conclusion doesn’t change because of the channel pricing dynamics for single-source brand-name drugs.

UPDATE: The pharmacy reimbursement data reported in the PBMI survey reflect what the employers pays, i.e., the rate paid by the employer to the PBM when a script is filled by a retail pharmacy. In my computations above, I incorrectly use the data to compute what the pharmacy gets paid, i.e., the rate paid by the PBM to the pharmacy. However, this may not affect the results. Please see my comment below the post for clarification on this technical point.

Although I had no involvement in the PBMI survey research, I want to let you know that I will be delivering the keynote address at the 16th Annual PBMI Drug Benefit Conference in February 2011. My topic: "The Future for Pharmacies." Perhaps I’ll see you there!


  1. Does the survey data collected from employers differentiate between the amount paid to pharmacy and the amount the employer paid? More specifically, does it seperate the pharmacy charges from the PBM charges or lump everything together? Just because an employer paid AWP - 15% for a prescription doesn't mean the pharmacy received AWP - 15%.

  2. I think it is the amount paid to the pharmacy, i.e., excludes PBM spread or admin fees. The benefit report website states the following:

    AWP Discount % – (AWP Minus X%)
    The negotiated amount a drug plan pays to pharmacies for the ingredient cost of a prescription and commonly expressed as a percentage off of Average Wholesale Price. (source)

    I will follow-up with PBMI and post an update if my interpretation is incorrect.


  3. The first comment is a very good one and has a direct impact on your conclusion that "gross profits per brand prescription have been fairly stable" - your data doesn't measure pharmacy gross profits, it measure employer brand expense. In addition, the impact of all this is muted because the percentage of prescriptions that are brand prescriptions has fallen since 2000. Your article also fails to mention how pharmacy costs have changed over this time. All of this is important in determining the impact of all this on pharmacies.

  4. Not true. The PBMI data measure employer expense. My computations in the second exhibit are for pharmacy gross profit (for brands) by estimating pharmacy acquisition cost.

    Pharmacy costs (operating expenses, cost of dispensing) have nothing to do with gross profits. Gross profits equal Revenues minus Cost of Goods, i.e., Earnings Before Expenses. See Shhhh! Owning a Pharmacy is Very Profitable for more on pharmacy economics.

    Despite what you may have heard, overall gross margins for pharmacies have remained in a relatively stable range over the past 10 years. See this Drug Channels post for supporting data.


  5. So if I'm understanding this right, Gross Profits have remained stable over the past 10 years.. and that's somehow OK?? I might not have an MBA, but as a business owner, I am happy to hear my gross profits have not declined, but seriously, isn't the point of running ANY business to show more and more profit every year?
    Also, since the Gross Profits have been stable, this does have an effect on Net Profits, because guess what? Rent has gone up, Utilities have gone up, Salaries have gone up, etc... Therefore, it is NOT acceptable that Gross Profits remain stable. There are only so many costs you can control running an operation.

  6. Do you know if the Average AWP Discount includes 90-day claims? AWP-17.5% seems deep for 30-day only.

  7. If this is purely an employer survey then I agree with the first two commentors - this is the amount the employer paid to the PBM, not the amount the pharmacy got paid. The spread is not taken into account - the PBMs that take spread are non-transparent, the employer would not even know what the pharmacy got paid!

  8. www.ApproRx.comOctober 20, 2010

    Bottom line expense

    As a reader, I am always amazed and have been for many years at this PBMI report and its lack of payer information within. This report is described as an employer report for the benefit of decision makers at the executive level regarding prescription benefits. This report however deals predominately with employee payments via copays and coinsurance; not employer bottom line expenses.

    This report clearly shows the cheapest way for employees to obtain prescriptions via their employers benefit package, however, it does not show the most cost effective method for the payer to provide Rx benefits to their employees.

    The bottom lines numbers are not reported and the bottom line for a payer is the real dollar cost per prescription per month they incur by the different modalities of distribution, i.e. mail for retail?

    This number should then be broken out in brands and generics so the employer can clearly see the most cost effective route of distribution medication available.

    This simple bottom line dollar number has not been made available in this report.


    This report is, and has been over the last decade, about mail order vs. retail. If all the trends, percentages, copay, and coinsurance charts were replaced with a simple bottom line cost report in real dollar amounts it would show that retail pharmacy is slightly more expensive to the employee but retail IS much LESS expensive to the employer.

    If you cost average the two (employee cost + employer cost) the report would show retail cost lower overall. This is using their numbers from their report which is highly biased by the large number of self funded respondents.

    Adam, ask PBMI for bottom line real dollar numbers and two split the numbers by fully funded vs self funded.

  9. OK, I heard back from PBMI regarding the data in their report.

    As several comments correctly note above, the AWP-X% data reported in the PBMI survey reflect what the employers pays, i.e., the rate paid by the employer to the PBM when a script is filled by a retail pharmacy. In my computations above, I incorrectly assume that the data reflect what the pharmacy gets paid, i.e., the rate paid by the PBM to the pharmacy. I have added a comment to the original post to reflect my error.

    Now, the real question: Does it matter? Perhaps, but probably less than you think.

    1. The gross profit dollars estimate in column [F] could be inflated. Here's why. If the PBM gets compensated via spread pricing, then the rate received by the pharmacy will be lower that waht appears in the PBMI survey. For example, the employer might pay AWP-17.5% to the PBM, but the PBM might pay AWP-18.0% to the pharmacy. Therefore, my estimates of gross profits per prescription would be too high.

    2. My overall conclusions would not change if PBM retail network spreads have been similar over time. Personally, I believe that spreads have shrunk due to competition between PBMs. However, I am not aware of any publicly available data on spreads.

    3. Even if I accounted for PBM spreads, my analysis may still underestimate pharmacy profits. The only other public data on AWP spreads comes from state Medicaid programs, where the average ingredient cost reimbursement across states is about AWP-13%, i.e., much better for pharmacies than the PBMI data. (source). I could also have overestimated pharmacy acquisition costs since these data are also not public.

    My bottom line: The estimates above could be too high, but they also could be too low. In the absence of better information, I'll use these figures until I see a reliable report of average retail pharmacy margins on brand-name prescriptions.

    Thanks to everyone who commented on the post. We are now debating issues that most people don't even understand!


  10. WWW.APPRORX.COMOctober 20, 2010

    You are perhaps correct that the spread pricing may not make a difference regarding this blog’s premise concerning pharmacy reimbursement as a total dollar amount.

    But you must ask why such an extensive report is not made clear, valid and easy to read. Even you were confused or mislead. Now imagine how mislead HR staffs are when it comes to Rx pricing and distribution when a PBM contract is in play.

    Though not a big issue to the correctness of this article, CLARITY in all PBM issuses would makes a big difference to pharmacists and employers.

    Employers need the smoke blown away so they can make informed decisions regarding Rx benefits and pharmacy owners need to know how their PBM competition is using contract language and percentages to give mail order the appearance of reduced Rx cost and then inform local employers.

    No mail order savings exist FOR THE EMPLOYER when bottom line methodology is used.
    Jim Fields

  11. Jim,

    Despite the tricky interpretation, the PBMI report is the best data that is available. Any suggestions for an alternative?

    I don't think that your assertion about mail vs. retail is correct. I will address this topic in a future post.

    I don't understand why you assume that employers are unable to negotiate favorable agreements if they are careful. Don't assume that the marketplace is inefficient without evidence or facts. As I note in Why do pharmacy owners care about PBM transparency?, consumers should watch their wallets when a group of independent pharmacists decide to be the altruistic, self-appointed guardians of large insurance companies, big corporations, and the U.S. Federal government.


  12. I am an independent pharmacy owner currently working with 2 local self insured employers who have become frustrated with their current PBM and their drug spend. We performed an analysis of their prescription drug spending for the first 6 months of 2010, and, surprise, found that 10% of their "drug spend" was a spread between what the dispensing pharmacy was paid and what the employer paid to the PBM. The sample size of 2 is admittedly small, but in each case we found a spread of $14/prescription ($14 is more than I make on most of my third party prescriptions). I don't think that the breakdown of what went to the pharmacy and what went to the PBM can be should be ignored any longer.

  13. Adam- Once you realized your initial assumption was wrong, you began using a lot of ifs, coulds, probably's, etc in an effort to defend your initial conclusion. For a man who so often states that he is unbiased and lets facts do the talking, you are being very subjective here. Several of the posters have raised excellent points (flat growth over a decade is terrible, employers should not have to operate their own blog to understand their options, generics are not included in this report, etc) that you seem to easily discount, with zero facts. There is no reason the employee groups should have such difficulty in comparing competing PBMs, and if you cannot understand the survey, it is only reasonable to assume the PBMs are purposefully being confusing. Why?

  14. If they average pharmacy gross profit margin per brand prescription has been cut in half from 12.% to 6.35, while the PBM spread margin has remained the same and the drug cost has increased by 138%, why keep trying to take more from the pharmacy? It seems to me that there is a bigger pot of employer savings to be had from the manufacturers and PBMs.

  15. The data in my post only relate to BRAND-NAME prescriptions. The retail brand dispensing rate dropped in half during the period shown above--from 60% of scripts in 2000 to 32% in 2009 (per NACDS Industry Profile)

    Pharmacy profits from generic prescription picked up the slack, which is why TOTAL gross profit dollars per pharmacy doubled over the same period (per data published in the NCPA Digest).


  16. Man this thing is a powder keg ready to blow. It seems like the data is just starting to come together but the PBMs are still a few steps ahead. Of course Obamacare does nothing to rectify this situation, which is really too bad.

  17. This PBMI report is supposedly about the cost of drugs yet nowhere in this report is one mention of the actual cost of a drug to an employer, brand or generic?

    Employers are never told and employers never ask what the actual cost of a drug is; with and without insurance and mail vs. retail. I have done analysis’s for state unions, automotive unions, Unversities, and large private employers who are usually using one the big three PBMs and the actual cost of an individual Rx is never quoted.

    You give me any Rx benefit plan for analysis, no matter the size or assumed sophistication, and I will show the level of inefficiency employers have brought to the prescription market to date. There are many reasons for this but this is not the forum.

    The employers with fully funded plans are even less efficient consumers regarding Rx costs than the above self funded plans.

    Example: Adam, I will guess that you have a fully funded Medical Plan with an Rx benefit.
    Now try and find what is the actual price of each prescription that your PBM charged to your Health Plan; the actual cost by Rx that your plan then sent to underwriting, the actual cost that underwriting then used determine your health plans premiums for the next year’s rate.

    Welcome to Metformin priced at $154.00 for a three month supply vs. 10.00 at most pharmacies and 10% increase in premiums next year.

    These numbers are not only unavailable to you, the fully funded plan holder, but usually not available to your health plan itself. Why hide cost if they are good, answer you don’t?
    Looking forward to your blog on this subject.

  18. Tim Watson, PBMIOctober 21, 2010

    The main utility of the PBMI report is the ability of comparing prescription drug trends over time.

    We agree with the importance of employers understanding the “net cost” they are paying for prescription drugs in their plans. We also agree that the net cost analysis should be conducted by drug type (B/G) and by channel (retail, mail, specialty, etc.). A wide segment of national employers conduct net cost analysis and channel analysis on a regular basis. They also audit prescription drug pricing files from their PBM to validate whether contract pricing is being applied in accordance with the agreement.

    The net cost analysis is beyond the scope of this report, and is competently handled by a variety of qualified industry consultants.

    We are proponents of enhanced transparency in the PBM / payer marketplace, feeling its always best for purchasers of products and services to know what they are paying for in advance. Our PBM customer satisfaction report continues to demonstrate that employer satisfaction with their pharmacy program is directly correlated with how transparent the financial relationship is between themselves, and their PBM partner.

    We don’t support the notion of aggregate channel preference based on our survey data(e.g. which is better), and in fact try to draw out examples of how channel alternatives are growing (i.e. 90 day at retail). To get to the channel preference data for a specific employer, one needs more than aggregate reimbursement trend data. Relevant copays, net cost data, and other factors should be considered, and can only be accurately done on a case by case basis.

    While I think the point (i.e. importance of conducting net cost and channel analysis) may be redundant for our members (most of them understand why its important, and have been doing it for years), we may consider adding a paragraph in the pharmacy reimbursement section of future reports that highlights this point.

  19. Thanks for your thoughtful response, Tim. IMHO, the report fills a valuable need by providing data that are not available anywhere else.

    I hope some of the people who commented will take the time to read and understand your response. As you point out, an employer chooses whether and how to do business with a PBM. This basic marketplace reality appears to frustrate some Drug Channels readers.


  20. Tim Watson, PBMI said...

    The main utility of the PBMI report is the ability of comparing prescription drug trends over time.

    However the name of the PBMI report in question is "Prescription Drug Benefit COST and Plan Design"

    As stated by Takeda..... Decision makers and experts in drug benefit management rely on this report to monitor trends in retail, mail-order,and speciality pharmacy, this would be trends in Prescription Benefit COST and Plan Design as the title clearly states.

    Payers as well as PBMI members need to know cost trends but more importantly payers need know where the trend begins and the beginning is the cost of the drug. Again, try and do a Rx cost analysis on your fully funded Rx benefit.

    Thank you Mr Watson and Thank you Adam for this forum and its debates.... I am looking forward to more blogs on this subject.