GAO’s U.S.-CANADIAN Rx DRUG PRICE STUDY CRITICIZED
GAO's U.S.-CANADIAN Rx DRUG PRICE STUDY CRITICIZED by Battelle Research Scientist Fredrik Andersson, PhD, at a pharmaceutical industry-sponsored "Foresight Seminar" on Capitol Hill Nov. 19. Released Oct. 21 ("The Pink Sheet" Oct. 26, p. 3), the General Accounting Office comparison of U.S. and Canadian prescription drug prices "is a relatively good" study but raises "concerns" because it does "not compare like with like," Andersson said. His primary concern is that the study compares the lowest available prices in Canada -- those obtained from drug manufacturers by the Ontario Drug Benefit Plan -- with the highest prices manufacturers offer in the U.S. The GAO report does not account for basic differences between the countries, such as the fact that the U.S. R&D-based industry is highly productive whereas relatively few products are developed in Canada, Andersson said. The study also did not take into account the larger drug package sizes shipped in Canada and failed to allow for fluctuating exchange rates, he told the seminar. The Oct. 21 report provides results of the first of two GAO studies comparing U.S. with foreign drug prices. The second will compare prices charged by pharmaceutical manufacturers in the U.S. and European markets. Both studies were requested by House Energy & Commerce/Health Subcommittee Chairman Waxman (D-Calif.), who is expected to use them as the basis for hearings that may lead to drug price control legislation. Andersson asserted that "almost every" international drug price comparison study "has some methodological flaws" that produce "bias" in the results. The economist has found six primary types of flaws in price comparison studies: they do not account for lack of comparability of the countries' populations in terms of drug consumption or income and age distribution; drug sample selections involving top sellers constitute nonrandomized samples, and the top sellers are not the same in each country; prices are not comparable if companies are vertically integrated in one country but not the other or discounts, rebates or value-added taxes affect prices in only one country; common package sizes often vary from country to country; studies do not always account for differences in sales volume; and currency differences confuse comparisons because exchange rates fluctuate, and currencies do not reflect relative buying power. Based in Battelle's Policy Center in London, Andersson has conducted two studies commissioned by the Pharmaceutical Manufacturers Association: "International Price Comparisons of Pharmaceuticals -- A Review of Methodological Issues" and "The U.S. Pharmaceutical Expenditures in an International and National Perspective." GAO Assistant Director for Medicare and Medicaid Issues Jonathan Ratner, PhD, agreed that rigorous academic theory calls for randomizing samples in comparison studies, but argued that the lack of randomization is a less serious flaw if the sample is "dominant." "Nonrandomness" is not always a "fatal flaw," Ratner maintained. "The more dominant the sample you have" and "the more drug sales your sample represents, the less nonrandomness matters." Ratner noted that the sample examined in the GAO study includes the top 200 drugs, which "represent almost three-quarters of the market." GAO also agrees that it is "very standard" for a study to weight U.S. drug prices by their share of U.S. sales and to weight Canadian drug prices by their share of Canadian sales. However, the standard methods required by economic theory are designed to answer "a different question than what the Congress is interested in," Ratner said. Congress simply wanted to know whether the drugs that Americans buy cost less in Canada. Canadian drug-use patterns are not germane to the question, he said. Regarding the currency problem, Ratner agreed that relative purchasing power is an important consideration and that exchange rates fluctuate daily. However, GAO found the U.S.-Canadian price differentials relatively nonsensitive to exchange rate fluctuations, Ratner said. The issue will be more prominent when the U.S.-European prices are compared, he added. Ratner further agreed that package size differences from country to country "is an extremely vexing matter" for GAO, particularly in preparing the U.S.-European drug price comparison. It is an "unavoidable" problem, he said, but the study makes allowances that "give the benefit of the doubt to U.S. prices" by mitigating differentials because the presumption of the study was that U.S. prices are higher. It is possible to "take measures to deal with the problem," Ratner said, commenting that Andersson is "skeptical about our ability to do things on that score." He added that GAO could use help from economists at PMA and elsewhere in developing ways to allow for differences in package size without exaggerating price differentials. Speaking from the audience, David Schulke, an aide to Rep. Wyden (D-Ore.), commented that arguing the niceties of measurements of price differentials are not as "policy-relevant" as the fact that most countries have health insurance to help pay for drugs. Schulke noted that "people in all different countries have health insurance, usually a government program, that helps them bear the high utilization and the high cost per unit" of pharmaceuticals. "To me, that's the bottom line of all this," he said, not "endless disputes about" methodology. Ratner cautioned that the GAO study does not address the issue of drug price controls. "Whether adopting foreign price controls is a good idea is not something we can answer from a price comparison. That's a whole other issue," he said. The report simply demonstrates that prices generally are higher in the U.S., and "Dr. Andersson and I agree price differences exist," despite methodological concerns of particular studies.
Sign in to continue reading.
New to Pink Sheet?
Start a free trial today!
Register for our free email digests: