How Quiet Moves by a Publisher
Sway Billions in Drug Spending
Lawsuit Forces Hearst Unit
To Lower Prices on List
Widely Used as Benchmark
A 'Survey' of One Company
By BARBARA MARTINEZ
October 6, 2006; Page A1
For years, a little-known unit of publishing giant Hearst Corp. called First DataBank has played a powerful role in determining what Americans pay for prescription drugs. First DataBank doesn't buy or sell drugs -- it publishes lists of drug prices. Health plans and state Medicaid programs use those prices as a benchmark in determining what they pay pharmacies.
If the benchmark goes up, so do costs for these payers. That's what happened in 2002, when First DataBank suddenly made broad revisions to its key published list. The new prices had the effect of fattening the profits of pharmacies, out of the view of patients and companies who pay for the soaring cost of health care.
A 2002 email by a manager at one of the nation's largest drug wholesalers, San Francisco-based McKesson Corp., describes how pharmacies would be able to more than double their profit for dispensing the cholesterol drug Lipitor and adds, "that is awesome!!"
Now a tentative legal settlement, reached quietly this week in a Boston court, could reduce annual U.S. drug costs by billions of dollars. An economist hired by the plaintiffs puts the savings in 2007 alone at $4 billion. The actual amount could be lower if pharmacies negotiate higher fees to make up for what they are losing.
In the settlement, which is awaiting approval by a judge, First DataBank, of San Bruno, Calif., agrees to reduce many of the prices on its published list by five percentage points. It denies any wrongdoing and isn't paying any damages to the plaintiffs.
Documents made public as part of the litigation suggest that McKesson had a key part in the rise of the published benchmark prices in 2002. The documents suggest that McKesson's motive was to resolve an administrative headache. The McKesson manager's emails later noted a side benefit: The company's pharmacy customers would enjoy bigger profits.
First DataBank had long said its prices reflected a survey of national wholesalers. But a manager at the publisher said in a deposition that from 2003 only one company, McKesson, participated in the survey. In the litigation, First DataBank also said that only two of its 225 employees were trained to collect and update pricing information.
One of the most important parts of the proposed settlement in U.S. District Court involves the benchmark price at issue in the litigation, known as average wholesale price or AWP. The term is a misnomer because it no longer represents a price paid to wholesalers and is not an average of anything. An old industry joke holds that AWP stands for "ain't what's paid." First DataBank agreed that two years after the settlement is final it will stop publishing the AWP.
As AWP loses sway, employers, Medicaid programs and other drug payers may need to find new ways to figure out how much pharmacies are paying for drugs so that the pharmacies can be reimbursed at a fair, but not excessive, profit margin.
Mark Erlich, executive secretary-treasurer of the New England Regional Council of Carpenters, is one of the plaintiffs settling the case with First DataBank. He expects the settlement will save about $400,000 a year for his union's health fund, which covers 22,000 people and spent $10 million on prescription drugs last year. Mr. Erlich calls the earlier rise in First DataBank's published prices "a rip-off of consumers across the country." It affects the union, he says, because its contract with the company managing its pharmacy benefits specifies that the drug prices the union pays will be based on First DataBank's AWP benchmarks.
In a statement, First DataBank said it isn't responsible for what companies involved in drug distribution do with its data. "First DataBank does not set pharmaceutical prices. First DataBank is a reporter and publisher of information that is collected from third parties," the statement said. Hearst is a major media company whose holdings include the San Francisco Chronicle and Good Housekeeping magazine. It is a partner with Dow Jones & Co., publisher of The Wall Street Journal, in publishing SmartMoney magazine.
The changes in First DataBank's published prices are responsible for only a portion of the increase in drug prices in recent years. The prices of drugs are set, first and foremost, by drug manufacturers. From 2000 to 2005, manufacturer prices on the most popular brand drugs grew by about 40.5%, according to a study by AARP, the advocacy group for Americans over age 50.
Between the manufacturer and the end user stand a variety of middlemen who take their cuts. These include wholesalers such as McKesson, who distribute drugs obtained from manufacturers, and pharmacies where patients go to get prescriptions filled. First DataBank plays a key but little-noticed role in drug pricing as a source of data used by middlemen to set their prices.
Another group of middlemen are pharmacy benefit managers or PBMs, which manage employers' drug benefits and often act as pharmacies themselves by filling employees' prescriptions through mail order. PBMs also frequently use AWP as a benchmark. Changes in AWP may also affect people without insurance who pay out of pocket because pharmacies sometimes use AWP in setting their cash prices.
Even as patients face higher co-payments and other out-of-pocket medical costs, many pharmacies and PBMs are prospering. The Dow Jones index of U.S. drug-retailer stocks has risen about 40% since the beginning of 2002, roughly in line with major indexes. A few big chains are doing especially well: Walgreen Co.'s net income has nearly doubled in the past five years and CVS Corp.'s has tripled. Share prices of the three major PBMs are also sharply up over the past few years.
Vestige of Old System
Average wholesale price "is a vestige of a drug-distribution system that disappeared in the early 1980s," says E.M. Kolassa of Medical Marketing Economics, an Oxford, Miss., consulting firm. In the late 1960s, the California Medicaid program needed a standardized way to reimburse pharmacies for drugs because "every claim was a paper claim based on whatever the pharmacist was charging," Dr. Kolassa says. Two consultants came up with "average wholesale price" after surmising that drug wholesalers generally charged retail pharmacies about 20% more than they paid manufacturers for drugs. California decided to pay pharmacies this new AWP, plus an additional dispensing fee.
Within a few years, Medicaid systems throughout the country had adopted AWP, and publishers such as First DataBank made a business of disseminating the pricing data. When commercial health insurers and employers began to reimburse for drugs and demand discounts from pharmacies in the 1980s and 1990s, they too adopted Medicaid's AWP system.
Gradually the 20% estimated markup became an anachronism. Wholesalers consolidated and became more efficient amid competition. They were selling drugs to pharmacies for just 2% to 3% more than what they paid. The compilers of AWPs, however, continued to report a 20% markup. States and employers adjusted by demanding discounts of 5% to 15% off the AWP.
First DataBank, founded in 1977, was bought by Hearst in 1980. Hearst bought another major publisher of AWPs, Medi-Span, for $38 million in 1998, but had to sell Medi-Span to Wolters Kluwer NV of the Netherlands in 2001 after a complaint by the Federal Trade Commission. The FTC said the acquisition gave Hearst "monopoly power" in drug data and led to "drastic price increases to customers, and reductions in product quality and customer service."
Thomson Corp.'s Red Book also publishes AWPs but First DataBank's figures are most commonly used in the industry, says Sean Brandle, a pharmacy benefits consultant to major employers and unions at Segal Co. in New York.
The trigger for litigation was a sudden rise in First DataBank's AWPs in 2002. Previously the 20% markup beyond the wholesaler's acquisition cost was common, although not universal. Suddenly First DataBank started revising its AWPs so that the markup was almost always 25%. According to internal McKesson documents, by 2004 nearly 99% of drugs carried the larger 25% markup. The cumulative effect, according to the plaintiffs, was that employers and others paid an extra $7 billion between August 2001 and March 2005 on drugs covered by the suit.
Documents from the period make clear that McKesson influenced the shift to an across-the-board 25% markup. The drug wholesaler's motivation was to simplify its system: Its computers recorded a "suggested sales price" for each drug that corresponded to the AWP, and it was easier if the markup was always the same. But McKesson managers also recognized that if the markup were to be standardized, it would be beneficial to standardize it at a high level -- that is, at 25%. The result would be higher margins for its pharmacy customers.
In an internal email on Jan. 7, 2002, McKesson's director of brand pharmaceutical product management, Robert James, wrote that "our successes recently and during this past year include raising the AWP spreads" on many drugs. As a result, he wrote, "we have an opportunity to 'market' our efforts now" to retail pharmacies who would appreciate that McKesson was working on their behalf. In a competitive market where pharmacies have a choice among wholesalers, such marketing could give McKesson an edge.
Mr. James wrote that in his discussions with customers, "one of the comments that was made was 'this would certainly be a good reason to renew our agreement with McKesson when it's time.' Talk about being good partners, wow!"
In an April 2002 email, Mr. James explained to colleagues that while pharmacies previously made a $6.86 profit dispensing Lipitor, with the new AWP they "will enjoy $17.18 profit...and that is awesome!!"
As Mr. James noted in his emails, many pharmacies say their profit margins have been squeezed in recent years. That is largely the result of efforts by the pharmacy-benefit managers hired by employers. PBMs have driven down reimbursements to pharmacies, passing on the savings to employers or keeping some of it for themselves.
Douglas Hoey, chief operating officer of the National Community Pharmacists Association, says any extra money pharmacies might have gained from the changes in First DataBank's prices meant little amid their woes. "We don't know where [the extra profits] went -- we just know where it did not go and that's to the community pharmacies," Mr. Hoey says. The association says many small pharmacists have closed down or are considering doing so because of slim profit margins.
In a statement, McKesson says setting the AWPs was First DataBank's job. "A full reading of McKesson documents, including e-mails, demonstrates that McKesson did not enter into any agreement with First DataBank to raise published AWPs," the statement says.
McKesson, which is named in the lawsuit, isn't a party to the settlement. "We intend to continue to press the case against McKesson," said Thomas Sobol, the plaintiffs' attorney at Hagens Berman Sobol Shapiro LLP.
The price rises published by First DataBank met with anger among some in the industry. "We were really mad," says Tim Heady, head of the pharmacy-benefits division at health insurer UnitedHealth Group Inc. UnitedHealth called First DataBank for an explanation but couldn't get a satisfactory one, Mr. Heady says.
A vice president at drug maker GlaxoSmithKline PLC wrote to the president of First Data Bank in March 2002 asking why the publisher had increased the AWP for the asthma medication Advair even though Glaxo hadn't raised its prices. He complained that "First DataBank has not been willing to share any information" about the change.
Plaintiffs' lawyers first suspected that drug manufacturers were behind the AWP increase and sued them. The drug companies denied any role, and in 2004 both sides started inundating First DataBank with subpoenas looking for answers.
Many employers and other payers for drugs didn't notice the rise in First DataBank's benchmark prices and didn't attempt to roll them back. Drug manufacturers were raising their own prices frequently during the period. The escalating prices prompted employers to shift some costs to employees through higher co-payments. Employers often have difficulty learning what they are paying for specific drugs and what factors determine prices.
For years, First DataBank described its AWPs as the results of a survey of national wholesalers. The supposed survey didn't gather the actual prices the wholesalers were charging but rather their suggested markup based on the decades-old wholesaler margins.
It emerged in the litigation that the only wholesaler in the "survey," at least in its final years, was McKesson. There are three national wholesalers. Spokesmen for two of them -- AmerisourceBergen Corp. and Cardinal Health Inc. -- say their companies didn't participate in any First DataBank surveys during the period when the price increases were occurring and still don't.
In a January 2005 deposition, Kay Morgan, who was in charge of AWPs at First DataBank, was asked: "Was First DataBank receiving any information regarding markups from any other company other than McKesson?" Ms. Morgan answered: "No, sir, we were not." However, she said she believed the McKesson-only survey started toward the end of 2003. That contradicted the accounts of AmerisourceBergen and Cardinal Health, who say they didn't provide pricing information in earlier years either.
McKesson says it never knew it was the only wholesaler being surveyed. "First DataBank has testified under oath in an earlier lawsuit involving other parties that it never told McKesson that at times McKesson was the only wholesaler being surveyed," a McKesson spokesman says.
Two months after Ms. Morgan's deposition, First DataBank sent a letter to its customers announcing an end to the survey. It said from that point forward it would "freeze" the last markup "provided through the wholesaler survey process."