Thursday, November 17, 2011

Pharmacy Invoices Show Flaws in Drug Pricing Benchmarks

A new Office of Inspector General (OIG) report provides pretty compelling evidence that pharmaceutical reimbursement models need to change. 

In Review of Drug Costs to Medicaid Pharmacies and Their Relation to Benchmark Prices, OIG compared three drug pricing benchmarks—Average Wholesale Price (AWP), Wholesale Acquisition Cost (WAC), and Average Manufacturer Price (AMP)—to pharmacy invoice prices for Medicaid-reimbursed drugs.

So, are the benchmarks consistently related to pharmacy invoice costs? Nope, especially for generics. Here’s what OIG found:

  • From single-source drugs, the benchmarks were reasonably correlated to pharmacy invoice costs. 
  • For multi-source (generic) drugs, all three benchmarks were extremely poor proxies of pharmacy invoice cost.
  • Pharmacies receive mega-discounts vs. benchmarks on these drugs. (See my summary chart below.). 
  • The AMP benchmark had the least consistent relationship to invoice prices for both single-source and multi-source drugs. 
 The OIG’s analysis didn’t include off-invoice discounts or rebates, so it can’t tell us much about underlying pharmacy margins. But the negative results for AWP/WAC/AMP make alternate methods (hello, average acquisition cost!) look more reasonable.


OIG conducted the study to “provide information that States can use as they consider changes to their reimbursement methodologies” in the Medicaid program. Of course, the commercial market is paying close attention. See Coming Soon: Average Acquisition Costs for Pharmacies.

OIG collected invoice data from a stratified sample of 120 pharmacies (Urban/Rural, Independent/Chain). Each pharmacy submitted a single invoice from each supply source in November 2010. OIG ended up with about 20,000 invoice prices in the following four categories:

  • Single-source drugs
  • Multiple-source brand-name drugs without Federal Upper Limits
  • Multiple-source generic drugs without Federal Upper Limits
  • Multiple-source drugs with Federal Upper Limits 
 A very important limitation of these data: OIG did NOT collect any data on any discounts, rebates, or other price incentives not reflected in the invoice prices. These off-invoice amounts are greatest for multiple-source drugs, which may partially explain the lack of correlation between invoice prices and the benchmarks.


Appendix B shows invoice price as percentage of each benchmark. Here’s my summary for AWP and AMP. Click the chart to enlarge. 

Some observations: 
Big AWP discounts for generics. Multiple-source drugs are invoiced at significant discounts to the AWP list price, but at a significant premium to the AMP computed transactional price. As you can see, multiple-source drugs with Federal Upper Limits (FULs) average 13% of AWP. Average acquisition costs would be even lower by including rebates and off-invoice discounts. These discounts are precisely why payers use Maximum Allowable Cost (MAC) limits.

Independents were invoiced less than chains for brands. Invoice prices for single-source drugs—typically brand-name drugs—were 82.8% of AWP. Surprisingly, invoice prices for urban independent pharmacies were 81.97% of AWP, which was LESS THAN the 83.19% invoice price for urban chains. Rural independents also paid less than urban chains. Both differences were statistically significant. (See page 25 of the OIG report PDF file.)

Chains were invoiced less than independents for generics. The relationship was reversed for multiple-source drugs with FULs. Chains received invoice discounts that were much larger than the discounts received by independents. The difference is also statistically significant. Rebate policies may explain some of this gap, but it’s probably due primarily to the buying power of the biggest chains vs. wholesalers.


Both AWP and WAC were consistently related to pharmacy invoice prices for single-source drugs. These results are consistent with discount structures in the channel and wholesaler pricing approaches with pharmacies. The AMP results were less consistent, but still showed a rough relationship. 

This makes sense because manufacturers of brand-name drugs do not generally provide discounts from the WAC list price to retail and mail-order pharmacies. Instead, these pharmacies purchase primarily from drug wholesalers for either direct-store or warehouse delivery. 

Wholesalers receive limited discounts off the WAC price from brand-name manufacturers. Buy-side transactions from pharmaceutical manufacturers, including fee-for-service agreements and prompt payment terms, are the primary sources of a pharmaceutical drug wholesaler’s gross profit from brand-name drugs. Exhibit 18 in the 2011-12 Economic Report on Pharmaceutical Wholesalers and Specialty Distributors provides Pembroke Consulting’s estimates for the underlying buy-side and sell-side gross margins for core drug distribution at the Big Three wholesalers in 2011.

For multiple source drugs, the relationship between a pharmacy’s invoice price and the corresponding benchmark prices was inconsistent. For example, the relationship between invoice prices and AMPs for multiple-source drugs without FULs ranged from 5% to 8,350% percent of the related AMPs. OIG doesn’t present the results for multiple-source drugs with an FUL, but the frequency distribution in Appendix E suggests a weak relationship between invoice prices and AMPs.


OIG has identified a number of drug pricing topics in its 2012 workplan, so there will be more revelations to come in the next year.

In the meantime, expect the National Average Drug Acquisition Cost (NADAC) survey to get even more attention when it rolls out…next year?


  1. Your chart shows invoice cost to be AMP*200%. But as you have pointed out, the new FULs are AMP*175%. Doesn't this data prove that the new formula is wrong and CMS needs to fix the markup before rolling it out?

  2. Another point to consider is that the AMP that the OIG compared is not the same as the draft AMPs recently published by CMS.  I believe the OIG AMP was the raw data from mfgs.  It is not a smoothed and weighted average AMP that is being used for the new FUL calculations.  Also CMS still has not published the rules for reporting AMP.  I think the new CMS FULs will still be too low but I do think that they will correlate better with invoice prices.  Your thought?

  3. I don't think it proves anything b/c OIG looked at invoice price, not acquisition cost. But it does raise questions about whether AMP is the most appropriate benchmark. 

  4. Great points. As I note above, no definitive conclusions yet, but the data certainly make we wonder if AAC will be a better metric than AMP. I presume OIG will update when the new AMPs settle down.

  5. Doesn't this also imply that distributor margins on generics are close to 100% (using invoice price; not net of discounts)? If discounts were in the 20-30% range, it would imply that margins are 70-80%, correct?

  6. Bruce SemingsonNovember 17, 2011

    I am not surprised by the results of the OIG report.  Why on earth should our government or any other entity know our net acquisition cost?  When I purchase a new car or a new shirt I do not know their acquisition costs?

    In most cases invoice prices have never stated net acquisition cost in our industry and for good would be fully transparent and ill advised.  Until retail pharmacy is able to negotiate reimbursements for drugs propucts separate from the delivery of defined services this game must continue.  We have been discussing pharmaceutical pricing for over 40 years without worthwhile results.

    Perhaps someday our industry will take seriously the need to measure patient utilizations and outcomes versus the net price per pill and how fast we can get greater quantities into the hands of consumers for good and evil.

  7. That seems high. I showed that implied wholesaler margins weer ~50% using the AMP data and some AAC data. See the last section of Hello, Transparency: CMS Publishes its First AMP Data.

  8. Not much to add except that the profession of pharmacy is paying the price now for many missteps in the past. We sat by and said nothing as the PBMs and state Medicaid programs reduced "dispensing fees" to levels that nowhere near (sometimes zero) reflect the true cost to dispense.

    Those days were before my time, but I can only assume pharmacists did not speak up because there was sufficient margin on the buying side of the drug. Now, many years later the drug margin has been eroded to closer reflect acquisition costs, and in many cases below acquisition cost.

    I have a stack of of over 500 prescriptions dispensed at PBM MAC prices that did not even cover the cost of the drug. What I take issue with is the PBM making more margin on the computer transaction of processing a claim (which adds no value to the healthcare dollar) than the pharmacy does in the act of dispensing.

    I often wonder how a PBM or any other payer expects a pharmacy to put a
    new $500 drug on the shelf when the margin is less than 3% ($15). It
    would take almost 3 years of dispensing that drug (33 months) before the
    initial investment is recouped, let along make any money. When
    patient's asked why we don't have a drug in stock I tell them to ask
    their insurance company or HR department.

    Speaking for myself, I have no problem dispensing based on true acquisition costs as long as the fee is also corrected to reflect true dispensing costs (including inventory management). I'll go for broke a say a reasonable profit should also be included.

  9. The problem is that the fees paid to pharmacies have been traditionally well below the cost of the service provided and the spread between cost and benchmark price made up for the unrealistic fees provided. Essentially, the government wants the pharmacy to provide the drug for no mark-up and then pay the pharmacy a fee that does not cover the pharmacy's cost and the professional services provided. If this were the auto repair industry that would be like asking for a new fender at cost and then only paying have of the labor charge. Good luck getting that kind of deal but that is what the government and prescription benefits managers are demanding.