Picking market baskets of drugs

FINAL REPORT:

Understanding Variations in International

Drug Prices

NORC PROJECT NO. 6319 JULY 27, 2006

TASK ORDER # HHSP233000007T / CONTRACT # 100-03-0020

SUBMITTED TO:

OFFICE OF THE ASSISTANT SECRETARY FOR PLANNING AND EVALUATION

U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES

200 INDEPENDENCE AVENUE, S.W.

WASHINGTON, D.C. 20201

PRESENTED BY:

NATIONAL OPINION RESEARCH CENTER (NORC)

AT THE UNIVERSITY OF CHICAGO

1350 CONNECTICUT AVENUE, N.W., SUITE 500

ASHINGTON, D.C. 20036

WITH THE ASSISTANCE OF:

GEORGETOWN UNIVERSITY HEALTH POLICY INSTITUTE

AND

IMS

HEALTH

NORC at the University of Chicago Task Order HHSP233000007T

This report was prepared by NORC at the University of Chicago under contract to the Assistant Secretary

for Planning and Evaluation. The findings and conclusions of this report are those of the author(s) and do

not necessarily represent the views of ASPE or HHS.

This report was produced under the direction of Laina Bush,

Task Order Officer, Office of the Assistant Secretary for

Planning and Evaluation, Office of Science and Data Policy,

Jim Scanlon, Deputy Assistant Secretary for Planning and

Evaluation (Science and Data Policy), Don Young, Acting

Assistant Secretary for Planning and Evaluation. This report is

available online at: http://aspe.hhs.gov/sp/reports/2007/UVIDP/.

NORC at the University of Chicago Task Order HHSP233000007T

Project Director:

Elizabeth Hargrave, NORC at the University of Chicago

Staff contributing to this report (in alphabetical order):

Ken Copeland, IMS Health

Jérôme Caron, IMS Health

Elizabeth Eaton, Georgetown University

Ed Evaldi, IMS Health

Jyoti Gupta, NORC at the University of Chicago

Jack Hoadley, Georgetown University

Grace Yang, NORC at the University of Chicago

NORC at the University of Chicago Task Order HHSP233000007T

iii

Understanding Variations in International Drug Prices

Table of Contents

Understanding Variations in International Drug Prices...............................................................................1

Background ....................................................................................................................................................1

Methodology ..................................................................................................................................................2

Results.............................................................................................................................................................5

Differences in U.S. and Canadian Prices for Individuals with Third-Party Coverage...........5

Differences in U.S. and Canadian Prices for Individuals without Third-Party Coverage.....5

Comparing U.S. Copayments to the Full Cost of Drugs in Canada ........................................8

Price Differences by Patent and Trademark Status....................................................................9

Testing the Impact of Substituting Drugs Without a Direct Equivalent in Canada............12

Conclusion....................................................................................................................................................13

Appendix A: Methodology............................................................................................................................14

Selecting Population Groups and Their Market Baskets.......................................................................14

Identifying Potential Market Baskets ...................................................................................................14

Selecting Population Groups ................................................................................................................15

Working with Price Data............................................................................................................................18

Matching Drug Names to Prices ..........................................................................................................18

Matching with Canadian Drugs ............................................................................................................18

Weighting .................................................................................................................................................22

Currency conversion ..............................................................................................................................22

Sensitivity to Drug Substitutions..........................................................................................................22

Changes in Drug Use Since 2003 .........................................................................................................24

Prescriptions vs. Spending.....................................................................................................................25

Payer Type and Retail vs. Total Price ..................................................................................................26

Market Baskets.............................................................................................................................................27

Appendix B: Literature Review.....................................................................................................................42

The Canadian Pricing System....................................................................................................................42

Methodology of Price Comparisons.........................................................................................................44

Price Comparisons ......................................................................................................................................46

Studies of Wholesale or Manufacturer Prices.....................................................................................49

Studies of Retail Prices...........................................................................................................................52

Discussion ....................................................................................................................................................55

Appendix C. Works Cited .............................................................................................................................56

NORC at the University of Chicago Task Order HHSP233000007T

Table of Figures

Figure 1. Weighted Average Price Per Prescription, by Market Basket and Payment Type..................6

Figure 2. Average of Per-Prescription Prices for All 15 Market Baskets, by Type of Coverage ..........7

Figure 3. Copayments in the United States vs. Full Price of Drugs in Canada .......................................8

Figure 4. Average Total Price Per Prescription, by Product Type and Payment Type ..........................9

Figure 5. Variation in Relative Difference of Generic Prices...................................................................10

Figure 6. Relationship Between Makeup of Market Basket and Relative Prices ...................................11

Figure 7. Sensitivity of Average Difference Between U.S. and Canadian Price to Inclusion of Zyrtec

and Toprol XL ........................................................................................................................................12

Figure A-1. Rank Order Correlation of Drugs Used, by Age Group.....................................................15

Figure A-2. Rank Order Correlation of Drugs Used, by Race.................................................................16

Figure A-3. Rank Order Correlation of Drugs Used, by Income............................................................16

Figure A-4. Rank Order Correlation of Drugs Used, by Insurance Status ............................................17

Figure A-5. On-Patent Drugs with Same Manufacturers but Different Names ...................................19

Figure A-6. Generic Drugs with Different Names and Manufactures in the U.S. and Canada..........19

Figure A-7. Dosage Levels Substituted in U.S. and Canadian Market Baskets .....................................20

Figure A-8. Dosage Levels Substituted in Canadian Market Baskets Only ...........................................21

Figure A-9. Drugs Not Available in Canada...............................................................................................21

Figure A-13. Market Basket for Females Under 12...................................................................................27

Figure A-14. Market Basket for Females 12-24 .........................................................................................28

Figure A-15. Market Basket for Females 25-39 .........................................................................................29

Figure A-16. Market Basket for Females 40-64 .........................................................................................30

Figure A-17. Market Basket for Females 65+ ............................................................................................31

Figure A-18. Market Basket for Males Under 12.......................................................................................32

Figure A-19. Market Basket for Males 12-24..............................................................................................32

Figure A-20. Market Basket for Males 25-39..............................................................................................33

Figure A-21. Market Basket for Males 40-64..............................................................................................34

Figure A-22. Market Basket for Males 65+ ................................................................................................35

Figure A-23. Market Basket for Insured .....................................................................................................36

Figure A-24. Market Basket for Uninsured ................................................................................................37

Figure A-25. Market Basket for Non-Hispanic Whites ............................................................................38

Figure A-26. Market Basket for African Americans..................................................................................39

Figure A-27. Market Basket for Hispanics..................................................................................................40

Figure A-28. Summary of Drugs Included in Market Baskets.................................................................41

Figure B-1. Summary of U.S./Canada Prescription Drug Price Comparisons: Studies of

Manufacturer or Wholesale Prices .......................................................................................................47

Figure B-2. Summary of U.S./Canada Prescription Drug Price Comparisons: Studies of Retail

Prices.........................................................................................................................................................48

NORC at the University of Chicago Task Order HHSP233000007T

Understanding Variations in International Drug Prices

Rapidly rising prescription drug prices have caused many consumers, third-party payers, and

policymakers to look for ways to lower their drug costs. One strategy that many have seen as a

potential solution is to import drug products from countries where drug prices are lower than those

in the United States – commonly called reimportation. This report focuses on the differences

between prices in the United States and Canada, and seeks to create an “apples-to-apples”

comparison of the prices consumers might face for a specific market basket of drugs in each

country, taking into account any insurance coverage they might have. We make this comparison for

fifteen different market baskets of drugs, to test the sensitivity of results to the drugs selected.

Across the different market baskets, Canadian prices were about one-third lower on average than the

prices charged to insured Americans. Because those without coverage in the United States do not

have access to the discounts negotiated by third-party payers, these individuals would find that

typical prices in Canada are about half of those they face in the United States. Overall, differences

are larger for brand-name drugs, and much smaller for generic drugs. In fact, prices for generic

drugs are likely to be higher in Canada, at least compared to the prices charged to people in the

United States with insurance coverage.

Background

While Canada offers universal access to health insurance, that insurance covers only acute hospital

and physician services, and does not provide coverage for outpatient prescription drugs. Most

Canadians receive prescription drug coverage through their employers. The federal government

offers drug coverage to some populations, including veterans and First Nations and Inuit people,

and there are publicly funded provincial programs that pay a substantial share of drug costs for

older, disabled, and low-income Canadians (Gross, 2003). These provincial programs cover 42

percent of all national drug expenditures (Morgan, 2003).

Prices for newly-patented drugs are controlled by federal regulation in Canada. Prices for new brand

drugs are limited to the median of prices for the new drug in France, Germany, Italy, Sweden,

Switzerland, England, and the United States. Limits for “me-too” drugs are based on the

breakthrough brand’s price; most new drugs cannot cost more than the other drugs in the same

therapeutic class. Annual increases in prices for existing on-patent drugs are limited by the

consumer price index (CPI): in a single year, prices cannot rise more than 1.5 times the CPI

(PMPRB, 2006). This limit is lower than recent increases in prices for many on-patent drugs in the

United States (Gross et al, 2006).

The provinces all use formularies to keep prices low for the drugs they cover, and new drugs that are

equivalent to already listed drugs are added only if they do not increase program costs. Some

provinces take further measures. Ontario has instituted price freezes; British Columbia uses

reference pricing; and Quebec requires manufacturers to give the province the best price given to

any other province (Morgan, 2003).

NORC at the University of Chicago Task Order HHSP233000007T

Legislative and regulatory price controls in the United States apply only to drugs purchased by the

Medicaid program, the Department of Veterans Affairs, and the Federal Supply Schedule, which is

used by the health programs run by the Department of Defense and other eligible entities (e.g.,

certain community health centers and public hospitals). However, insurers and pharmacy benefit

managers (PBMs) are also able to negotiate substantial discounts and rebates from retail prices.

These discounts take two forms. First, third-party payers are able to negotiate discounts with

pharmacies on the prices they charge at the pharmacy counter. These discounts are typically taken

off the pharmacy’s markup over their acquisition cost for the drug, and the effect of the discount is

seen at the time an insured customer purchases a drug.

The second type of discount is negotiated directly between third-party payers and manufacturers.

Typically, rebates are paid by the manufacturers of on-patent drugs in exchange for giving

preferential status to a drug and demonstrating a shift of market share away from competing

products. These rebates are separate from the pharmacy transaction. A 2000 ASPE study cited

evidence that average rebates tend to be about 5 to 7 percent of drug prices, but can be as high as 35

percent on selected drugs (HHS, 2000). Research by the Congressional Budget Office, in

connection with the Medicaid drug rebate, found that price discounts including rebates generally

range as high as about 20 percent (CBO, 1996).

Uninsured individuals typically do not have access to negotiated prices of either kind. These

individuals have been at the forefront of the push to import drugs from Canada. Some third-party

purchasers, including several municipal government employee plans, have also begun to explore the

possibility of savings for their insured employees from reimportation.

Previous studies have found that prices for prescription drugs are frequently lower in Canada than in

the United States, as detailed in Appendix B. The literature also includes critiques of some of these

studies (for example, Danzon and Kim). We have attempted to address several of the concerns

raised in these critiques with the approach taken in this report. In particular, we explore two factors

that could influence these observed differences between U.S. and Canadian prices: the market

basket of drugs selected for comparison, and whether the purchaser has third-party coverage for the

prescription at the time of purchase.

Methodology

The goal of this study is to determine the difference between U.S. and Canadian prices at the retail

point of sale for the prescription drugs that are purchased by U.S. customers. Our primary question

was, what would U.S. customers pay if they were able to purchase their drugs at Canadian prices?

Throughout the design of the project, we have used this research question to help shape our

methodology.

Price comparisons can be sensitive to the market basket of drugs included in the study. In

particular, generic drugs and brand-name drugs tend to have quite different pricing patterns, which

can affect the outcome of a price comparison. Our goal in this study was to test a variety of market

baskets, but to make each of them representative of the drugs used by a subpopulation in the United

States. We selected fifteen groups based on age, gender, race, and insurance status, which were

NORC at the University of Chicago Task Order HHSP233000007T

identified from a larger set of groups as those with distinctive patterns of drug use. For each group,

we selected the most commonly used drugs accounting for a third of that population’s prescriptions

in the 2003 Medical Expenditure Panel Survey (MEPS). We then identified the most commonly

used form and strength of each drug for each group, so that we could limit our data collection to a

precisely specified set of drugs. The resulting market baskets each had from 8 to 32 drugs, with a

total of 106 drugs across the 15 groups. These drug lists are included at the end of Appendix A.

Although some drugs are included in at least half of all the market baskets, the mix of drugs varies

considerably. The market baskets for those over age 65 are dominated by treatments for the chronic

conditions that accompany older ages, whereas market baskets for children include a mix of

antibiotics, behavioral medications, and asthma drugs. Those for young women include various

contraceptive drugs.

For these commonly used drugs, we then sought to find their counterparts in Canada. We tried to

assure that the match between our U.S. market baskets and Canadian market baskets is as close as

possible, while still representing the utilization of the population groups we identified. We first

looked for an exact match by drug name, form, and strength between the United States and Canada.

When an exact match was not available in Canada, we sought to match drugs by chemical entity, and

when possible, by patent and trademark status and/or manufacturer. There were several drugs on

our list at a dosage level not available in Canada. When possible, we used another dosage level that

is available in both countries as a substitute in both the U.S. and Canada market baskets; otherwise,

we used the dosage available in Canada closest to the most common U.S. dosage. More information

on this matching process is available in Appendix A.

Once we had specified our list of drugs for the United States and Canada, we then used data

collected by IMS Health to identify the average price per pill (or, for other forms such as liquids and

inhalers, the average price for a set amount) for each drug in both countries. The IMS prescription

databases represent very large, non-probability samples of prescriptions dispensed at retail

pharmacies in the United States and in Canada. The data are representative of prescription

transactions for both large retail pharmacy chain organizations and small independent pharmacies.

In the United States, the IMS sample is geographically representative of all 50 states plus the District

of Columbia and Puerto Rico. The databases include all prescriptions dispensed at over 35,000 retail

pharmacies in the United States, representing two-thirds of all U.S. retail pharmacies and an

estimated 70 percent or more of all prescription transactions in the retail pharmacy sector.

In Canada, the IMS sample includes over 4,300 retail pharmacies, representing 60 percent of all

Canadian retail pharmacies and an estimated 70 percent of all prescription transactions. However,

we used data only from Ontario and Quebec, which make up nearly three fourths of the overall IMS

sample (3,100 pharmacies). These two provinces are the only ones for which IMS collects data on

the type of payer, a critical part of our analysis. In addition, IMS has determined that in other

provinces, pharmacies frequently report list prices rather than actual prices. Prior IMS work with

Canadian prescription data, in conjunction with understanding of pharmaceutical price controls in

NORC at the University of Chicago Task Order HHSP233000007T

effect in Canada, suggest that there is little variation in prices between these two provinces and the

rest of Canada.

Using data from MEPS about the most common prescription size, we standardized the IMS per-pill

price to represent an average per-prescription price for each drug. This method allowed us to

eliminate any price disparities that might result from practice patterns or prescription-filling behavior

that would lead to price differences caused by prescriptions covering different periods of time (e.g.,

30 vs. 90 days) in the two countries. From these standardized per-prescription prices, we calculated

an index price for each market basket, weighting based on utilization of each drug within the market

basket for each population.

While in Canada there is little variation in the prices paid by different payers, in the United States the

cost of drugs can vary widely according to who is purchasing drugs. Individuals paying cash pay the

highest price, while insurers, managed care plans, and government programs can negotiate pharmacy

discounts and manufacturer rebates because they represent large numbers of customers. One goal of

this report is to explore the effect of these intra-country price differences on cross-country

comparisons. In this report, we separate the prices paid by individuals who had a third-party

payment from the prices paid by those who paid the entire cost of a prescription themselves.

In identifying the third-party prices, we have no access to proprietary rebate amounts and have not

attempted to estimate them.

The prices reported in this paper likely reflect the discounts that

insurers have negotiated with pharmacies, but not the additional discounts they have received from

manufacturers separate from the pharmacy transaction. The effect of including these rebates, if they

were available, would be to reduce the overall amount paid by third-party insurers for brand name

drugs and to reduce the difference between the United States and Canada for the prices of these

drugs.

There are several issues that are beyond the scope of this paper. We do not address here the safety

or legality of importing drugs to the United States from Canada. Likewise, we do not estimate the

shifts in manufacturer pricing behavior that would likely occur if reimportation became widespread,

or if policy changes in the United States attempted to force manufacturers to provide drugs to

customers at Canadian prices. HHS addressed many of these issues in its 2004 report to Congress

on prescription drug importation.

In particular, it is likely that customers without third-party coverage are paying prices close to the national limits set by

the Patented Medicine Prices Review Board. In this study, we find that in Canada, third-party prices are extremely close

to the prices paid by individuals without third-party coverage. Third-party payers in these provinces appear to be getting

little additional discount despite the extra limits set by the provincial governments for drugs purchased through

provincial programs.

We report prescriptions as “with third-party coverage” if a third party was involved at the time of the pharmacy

transaction, including prescriptions that were only discounted. (e.g., because the customer had a discount card). We

report prescriptions as “without third-party coverage” if a third party was not involved in the transaction at the

pharmacy counter. It is possible that some of these latter customers may be submitting receipts to an insurer for

reimbursement. However, without the involvement of an insurer at the time of the transaction, we assume the price

paid for these prescriptions is not discounted.

Other authors, including the HHS Task Force on Drug Importation (Report on Prescription Drug Importation, p. 115), have

attempted to estimate the rebates paid from manufacturers to pharmacies and wholesalers. We do not consider these

rebates here as we are studying prices paid by customers at the pharmacy counter, not pharmacy acquisition prices.

NORC at the University of Chicago Task Order HHSP233000007T

Results

For each of our fifteen market baskets of drugs, the total price paid for drugs in Canada is

consistently lower than the total price paid in the United States. This is true for individuals with and

without third-party coverage. However, the difference is greater for individuals without third-party

coverage.

Differences in U.S. and Canadian Prices for Individuals with Third-Party Coverage

For prescriptions filled by individuals with third-party coverage, the difference between Canadian

and U.S. prices ranges from about 30 percent lower for the market basket that represents the drugs

most commonly used by girls under 12, to as much as 47 percent lower for the market basket of

drugs that represents the drugs most commonly used by women ages 25 to 39 (Figure 1). Both the

average and the median of the differences for all fifteen market baskets are about 36 percent.

These observed price differences are in the general range of those found in the literature (see

Appendix B). Several of the studies described in the Appendix found differences in the range of 30

to 50 percent (with at least one considerably higher and one considerably lower). This consistency

of results provides some reassurance that the methodology used in this study is not creating

unexpected results. The consistency of results across market baskets is also significant. Although

there is variation, the price difference for most market baskets are remarkably consistent despite

differences in drug mix. As discussed below, the main difference results from the mix of brand and

generic drugs.

Differences in U.S. and Canadian Prices for Individuals without Third-Party

Coverage

For those without third party coverage, the prices reported are the total price paid by the individual

to the pharmacy. There is an even larger difference between U.S. and Canadian prices for these

customers (Figure 1). The market basket for drugs commonly taken by men aged 65 and over has

the smallest difference, with prices in Canada 38 percent lower than in the United States. The

market basket for the drugs commonly taken by women aged 25-39 again has the highest difference,

with Canadian prices 54 percent lower than U.S. prices. Both the average and the median of the

differences for all fifteen market baskets are about 45 percent.

For those with third-party coverage, the prices presented here include the total price paid at the pharmacy, including

both the insurance payment and any copayment made by the patient.

NORC at the University of Chicago Task Order HHSP233000007T

Figure 1. Weighted Average Price Per Prescription, by Market Basket and Payment Type

With Third-Party Coverage Without Third-Party Coverage

Market Basket:

Drugs Taken By...

U.S. Canada Difference U.S. Canada Difference

Females under 12 $31.44 $21.99 -30.1% $41.43 $22.70 -45.2%

Females 12-24 $45.93 $27.60 -39.9% $54.69 $28.56 -47.8%

Females 25-39 $48.47 $25.69 -47.0% $57.05 $26.44 -53.6%

Females 40-64 $51.78 $30.61 -40.9% $59.66 $31.00 -48.0%

Females 65+ $45.44 $28.48 -37.3% $50.82 $29.01 -42.9%

Males under 12 $39.07 $26.96 -31.0% $50.17 $27.64 -44.9%

Males 12-24 $70.48 $46.94 -33.4% $84.95 $47.85 -43.7%

Males 25-39 $68.04 $39.41 -42.1% $80.28 $40.06 -50.1%

Males 40-64 $57.24 $36.72 -35.9% $65.42 $37.50 -42.7%

Males 65+ $41.80 $28.99 -30.6% $47.46 $29.45 -37.9%

Insured $52.94 $31.60 -40.3% $61.24 $32.15 -47.5%

Uninsured $37.59 $25.80 -31.4% $43.47 $26.19 -39.8%

Whites $47.19 $28.80 -39.0% $54.49 $29.30 -46.2%

African Americans $47.16 $30.72 -34.9% $54.17 $31.19 -42.4%

Hispanics $36.84 $24.47 -33.6% $43.83 $24.75 -43.5%

Notes: Index price is a weighted average of the prices of the most common form and strength of the most commonly used drugs

for each population group.

Prices are in U.S. dollars, using the average exchange rate for the period of data collection.

Difference between U.S. and Canada prices calculated as (Canada price-U.S. price)/U.S. price.

Market baskets include the drugs most commonly used by individuals in each population group. Prices reflect the prices

paid by all individuals, not restricted to the population group. For example, the insured market basket includes the drugs

most commonly used by people with insurance. The price indices for that market basket use data for all people who

bought the drugs in that market basket, so it is possible to have a price index for uninsured individuals who purchased the

drugs in the insured market basket.

NORC at the University of Chicago Task Order HHSP233000007T

The higher differential between U.S. and Canadian prices for customers without third-party

insurance appears to be due largely to the difference between the prices paid by customers with and

without insurance within the United States (Figure 2). For any given market basket, the index price

in Canada is only 1 percent to 4 percent higher for customers without third-party insurance

compared to customers with insurance – when the market basket index prices are averaged, the

difference is less than a dollar. In contrast, uninsured customers in the United States pay 12 percent

to 32 percent more for a market basket of drugs than the cost of the same drugs for customers with

insurance, a difference of about $8.50 across all market baskets. The latter result is consistent with

findings from earlier work by ASPE in 2000 (HHS, 2000).

Figure 2. Average of Per-Prescription Prices for All 15 Market Baskets, by Type of Coverage

$48.09

$56.61

$30.32

$30.92

$10

$20

$30

$40

$50

$60

US, With Third Party

Coverage

US, Without Third

Party Coverage

Canada, With Third

Party Coverage

Canada, Without

Third Party Coverage

Average Price Per Prescription

(Average of All 15 Market Baskets

)

NORC at the University of Chicago Task Order HHSP233000007T

Comparing U.S. Copayments to the Full Cost of Drugs in Canada

The price comparisons for the full cost of drugs are the most relevant for many U.S. citizens without

insurance coverage for their drugs who seek savings by getting drugs in Canada. By contrast, most

individuals with coverage will not achieve savings unless their third-party agent is involved in the

transaction. Insured consumers considering drug purchases in Canada may find their insurance

coverage is not applicable for purchases from Canadian pharmacies. Thus, the relevant comparison

for these customers is between the copayment they are charged at their U.S. pharmacy and the full

cost of the drug in Canada.

In fact, individuals with insurance are unlikely to be able to find prices in Canada that are lower than

the copayments they pay for drugs that are covered by their insurance. On average across our

market baskets, the Canadian price is about 78 percent higher than the out-of-pocket cost to a

person in the United States with third party coverage (Figure 3), with a range from 31 percent to 141

percent higher. This result is not surprising given that insured consumers are paying only a share of

the total drug cost out of pocket.

There may be situations in which these individuals could achieve savings with Canadian drug prices.

For example, if their coverage includes a cap with an annual or quarterly spending limit (such as in

some Medicare Advantage plans prior to 2006), a gap in coverage (such as that faced in the new

Medicare drug benefit), or a limit on the number of monthly prescriptions, individuals may find that

it would be less expensive to buy drugs in Canada than to pay the full cost of a drug in the United

States. Those who want to take a particular drug that is not covered by their insurer’s formulary

might also benefit from the price differential by making purchases in Canada.

Figure 3. Copayments in the United States vs. Full Price of Drugs in Canada

$17.35

$48.09

$30.92

$10

$20

$30

$40

$50

US, With Third Party Coverage Canada, Without Third Party

Coverage

Average Price Per Prescription

(Average of All 15 Market Baskets

)

average

customer share:

NORC at the University of Chicago Task Order HHSP233000007T

Price Differences by Patent and Trademark Status

There is a dramatic variation in the differential between U.S. prices and Canadian prices depending

on the patent and trademark status of drugs. When generics and brand name drugs are identified

separately, brand name drugs show even larger differences between U.S. and Canadian prices.

Generics show smaller differences, or the direction of the difference changes, depending on the type

of payer.

The on-patent brand name drugs in our market baskets cost 43 percent less for individuals with

insurance and 50 percent less for individuals without insurance in Canada compared to U.S. prices

(Figure 4). The absolute price differences are also largest for brand name drugs, particularly the on-

patent brands in our market baskets.

As with the results for the overall market baskets, the variation in differences based on insurance

coverage appears to be because in the United States, transactions with third-party involvement cost

about 14 percent less than transactions without third-party coverage. Again, this is likely because of

discounts that third-party payers have negotiated with pharmacies. However, even after this third-

party discount, the difference between prices in the United States and Canada is substantial. The

remaining difference likely results from a combination of regulatory measures in Canada and the

market decisions made by manufacturers.

The availability of manufacturer rebates could further mute some of the difference between the

United States and Canada for third-party payers. Typically, rebates are only available for brand-

name drugs, and mostly for those that remain on patent. Previous studies have found that rebates

can be as high as 20 percent (CBO, 1996) to 35 percent (HHS, 2000) on selected drugs. Even if

amounts toward the higher end of these estimates were appropriate for most of the drugs in our

market baskets, the differences would fall short of the 43 percent average differential observed for

on-patent brand drugs.

Figure 4. Average Total Price Per Prescription, by Product Type and Payment Type

With Third-Party Coverage Without Third-Party Coverage

Product Type

Number

of Drugs

U.S. Canada Difference U.S. Canada Difference

Generic

27 $10.75 $14.75 37% $15.49 $14.92 -4%

Branded Generic

10 $25.79 $22.83 -11% $30.81 $23.45 -24%

Off-Patent Brand

19 $37.01 $17.10 -54% $42.65 $17.32 -59%

On-Patent Brand

38 $125.36 $71.37 -43% $145.91 $73.08 -50%

Notes: Index price is a non-weighted average of the prices for each type of drug included in any market basket for this study.

Prices are in U.S. dollars, using the average exchange rate for the period of data collection.

Difference between U.S. and Canada prices calculated as (Canada price-U.S. price)/U.S. price.

NORC at the University of Chicago Task Order HHSP233000007T

Generic drugs follow a much different pattern. For customers with insurance coverage, the full cost

of the generic drugs in our market baskets is actually higher in Canada than in the United States

(Figure 4). For customers without insurance coverage, generics are only 4 percent less expensive in

Canada than in the United States.

ASPE’s 2000 study showed relatively little price difference for generic drugs between the prices paid

by customers with a third-party payer and prices paid by those without coverage (HHS, 2000). That

does not appear to be the case for the generic drugs selected for our market baskets. The average

price of these drugs is about a third lower for customers with third-party coverage in the United

States compared to those without coverage. This third-party discount appears to be what is causing

U.S. generic prices to be lower than Canadian generic prices, for the drugs selected.

Previous studies have found more variability in the prices of generic drugs than in the prices of

generics. Possibly related to this general variation, we found variation from market basket to market

basket in the difference between U.S. and Canadian prices for generic drugs (Figure 5). For

customers with third-party coverage, we consistently find that U.S. prices are lower than Canadian

prices, but the difference ranges from 23 to 63 percent. For customers without third-party coverage,

Canadian prices are nearly 20 percent lower for some market baskets and over 20 percent higher for

other market baskets. The range in variation in the prices of brand name drugs among our market

baskets is only half as large.

Figure 5. Variation in Relative Difference of Generic Prices

Relative Difference Between Canadian and

U.S. Price

Market Basket:

Drugs Taken By...

Number of

Generic

Products in

Market

Basket

With Third-

Party Coverage

Without Third-Party

Coverage

Females under 12 2 53.6% -18.6%

Females 12-24 3 47.0% -18.1%

Females 25-39 4 42.8% -19.6%

Females 40-64 8 78.5% 24.0%

Females 65+ 7 59.8% 21.2%

Males under 12 2 47.9% -19.8%

Males 12-24 3 23.0% -10.7%

Males 25-39 5 75.7% 6.8%

Males 40-64 9 60.7% 9.3%

Males 65+ 11 49.4% 8.0%

Insured 8 53.6% 2.7%

Uninsured 11 46.0% -1.0%

Whites 10 62.5% 8.6%

African Americans 8 53.6% 6.5%

Hispanics 14 49.5% -4.8%

Notes: Index price is a non-weighted average of the prices for each type of drug included in any market basket for this study.

Difference between U.S. and Canada prices calculated as (Canada price-U.S. price)/U.S. price.

NORC at the University of Chicago Task Order HHSP233000007T

Figure 6 shows the proportion of each market basket that is made up of brand and generic drugs.

The market baskets we have selected have non-trademark generic use ranging from a low of 15

percent for females 25-39 to a high of 54 percent for males aged 65 and over. Including branded

generics, the generic rate ranges from 21 percent to 57 percent. This variation in generic use is

consistent with the wide variation in the types of drugs taken by these groups. For example, the

drugs in our market basket for young women include many on-patent birth control pills, while the

market basket for older men includes more generic heart medications.

Figure 6. Relationship Between Makeup of Market Basket and Relative Prices

Proportion of Market Basket, Weighted by

Utilization

Difference in Price Per

Prescription,

Canada vs. U.S.

Brand Generic

Market Basket:

Drugs Taken By...

Patent

Off

Patent

Trademark

(Branded

Generic)

Non-

Trademark

With

Third-

Party

Coverage

Without

Third-

Party

Coverage

Females under 12 22% 28% 8% 41% -30% -45%

Females 12-24 60% 15% 4% 21% -40% -48%

Females 25-39 52% 23% 10% 15% -47% -54%

Females 40-64 45% 27% 3% 25% -41% -48%

Females 65+ 45% 18% 4% 33% -37% -43%

Males under 12 25% 18% 15% 42% -31% -45%

Males 12-24 50% 8% 20% 23% -33% -44%

Males 25-39 57% 22% 0% 21% -42% -50%

Males 40-64 52% 13% 0% 36% -36% -43%

Males 65+ 34% 12% 0% 54% -31% -38%

Insured 50% 18% 2% 29% -40% -48%

Uninsured 32% 16% 3% 50% -31% -40%

Whites 48% 18% 3% 31% -39% -46%

African Americans 40% 16% 0% 44% -35% -42%

Hispanics 32% 15% 2% 51% -34% -44%

Correlation with Proportion of Market Basket On Patent -0.79 -0.52

Correlation with Proportion of Market Basket Non-Trademark Generic 0.82 0.84

Note: Difference between U.S. and Canada prices calculated as (Canada price-U.S. price)/U.S. price.

Nationally, both Express Scripts and the Generic Pharmaceutical Association reported that generics accounted for just

over half of all prescriptions in 2003. It may be that the bottom two-thirds of drugs used by each population group are

slightly more likely to be generics than the drugs we selected for our market baskets. (see Express Scripts, “Geographic

Variation in Generic Fill Rate”, http://www.express-scripts.com/ourcompany/news/outcomesresearch/

onlinepublications/study/regionalgenericvariation.pdf, and GPhA “Statistics,” http://www.gphaonline.org/

Content/NavigationMenu/AboutGenerics/Statistics/Statistics.htm)

NORC at the University of Chicago Task Order HHSP233000007T

Consistent with the findings in Figure 4, market baskets with a higher proportion of generic drugs

tend to show lower differences between Canadian and U.S. prices (correlation is .82 for third-party

and .84 without third-party coverage). Likewise, market baskets with a higher proportion of on-

patent brand drugs tend to show higher differences between Canadian and U.S. prices (correlation is

-.79 for third-party and -.52 without third-party coverage). For example, the drugs taken by young

adults (women 25-39 and men 25-39) have the largest differential between U.S. and Canadian prices,

and both have market baskets with a high proportion of brand-name drugs.

The use of multiple market baskets allows us to see a range of possible effects while still basing

measurements on actual utilization instead of an arbitrarily selected set of drugs. It is notable that

despite this variation, the difference between Canadian and U.S. prices remains within a fairly small

range.

Testing the Impact of Substituting Drugs Without a Direct Equivalent in Canada

Two drugs included in our market baskets had unusual circumstances that resulted in higher-than-

average differences between the U.S. price and Canadian price. Zyrtec (cetirizine hydrochloride), an

on-patent brand name prescription drug in the United States, is available over-the-counter in Canada

as Reactine. We included this medication in our Canadian market baskets, utilizing data on

prescription transactions for Reactine. Toprol XL (metoprolol succinate), another on-patent brand

name drug, is not available in Canada. We included a close substitute, Lopressor SR (metoprolol

tartrate), in the relevant Canadian market baskets. In both cases, we were following the principle that

we were seeking to find the drugs that U.S. customers would find in Canada as the closest match for

the drugs they take.

In all market baskets, excluding these two brand-name drugs makes the difference between the two

countries’ prices smaller (Figure 7). In thirteen of the fifteen market baskets, the difference is 2.2

percentage points or less (regardless of payer). The exception is the market baskets for children

under 12. Zyrtec accounts for about a tenth of the prescriptions included in each of these market

baskets, and its exclusion gives generic drugs a larger share of each market basket. The result is a

more notable drop in the relative difference between U.S. and Canadian prices for these two groups.

The change is particularly notable for prescriptions purchased with third party coverage, where the

difference between United States and Canada for these two groups of drugs falls to 21.5 percent for

girls and 23.5 percent for boys. While these results show that the selection of individual drugs for a

market basket can affect the magnitude of the results, the change is not enough to affect the overall

result that prices in Canada are substantially lower than prices in the United States.

Figure 7. Sensitivity of Average Difference Between U.S. and Canadian Price to Inclusion of

Zyrtec and Toprol XL

With Third-Party

Coverage

Without Third-Party

Coverage

Including Zyrtec and Toprol XL -36.5% -46.2%

Excluding Zyrtec and Toprol XL -34.1% -44.2%

Note: Unweighted average of results for 15 market baskets.

NORC at the University of Chicago Task Order HHSP233000007T

Conclusion

This report finds differences between Canadian and U.S. drug prices that are generally consistent

with the existing literature on international drug pricing. Uninsured customers in the United States

would find Canadian prices to be about half the prices they pay in the United States. Insured

consumers seeking to buy the drugs they currently buy in Canada would find, on average, prices that

are about a third lower than the total cost of their drugs, but higher than the copays they currently

pay. These findings are somewhat sensitive to different market baskets. While the overall result that

prices are lower in Canada is fairly consistent, the inclusion of more generic drugs can decrease the

difference between Canadian prices and U.S. prices.

The differing spread in prices depending on patent and trademark status is a result worth

emphasizing. Generic drug prices are mostly equivalent or higher in Canada, whereas brand drugs

are substantially less expensive in Canada. Because on-patent brand name drugs have the highest

prices, they also are the products with the largest absolute difference between Canadian and U.S.

prices.

There are several issues related to these questions that are beyond the scope of this paper. We do

not address here the safety or legality of importing drugs to the United States from Canada.

Likewise, we do not estimate the shifts in manufacturer pricing behavior that would likely occur if

reimportation became widespread, or if policy changes in the United States attempted to force

manufacturers to provide drugs to customers at Canadian prices. HHS addressed many of these

issues in its 2004 report to Congress on prescription drug importation (HHS, 2004).

Absent these widespread market shifts and other concerns, individuals without insurance coverage

can find substantial savings in Canada for the commonly used drugs included in this study. Third-

party insurers face a more complicated arithmetic. For generic drugs, third-party payers seem to be

able to achieve enough savings through pharmacy discounts so that they would not save by steering

their enrollees to Canada. For brand-name drugs, however, there is a large difference in prices

between the United States and Canada even after these discounts. Whether third party payers could

do better by purchasing drugs at Canadian prices largely depends on whether the additional

manufacturer rebates they receive on brand-name drugs are larger than the differences reported

here.

NORC at the University of Chicago Task Order HHSP233000007T

Appendix A: Methodology

This appendix describes our methodology for selecting market baskets of drugs and identifying

prices for those market baskets of drugs in the United States and Canada. To explore possible

market baskets, we selected the top third of drugs for several potential population groups and

compared them. Age had the most differentiating effect on which drugs a population group uses;

we used age, gender, race/ethnicity, and insurance status to create market baskets. The first section

of this appendix documents this process in more detail.

For each market basket, we identified prices for the most common form and strength of each drug.

We weighted prices in each market basket index according to overall use of each drug for that

population. The second section of this appendix documents our methodology for working with

drug price data.

Selecting Population Groups and Their Market Baskets

In selecting potential population groups, we started with a list of attributes available in the Medical

Expenditure Panel Survey (MEPS) that could be reasonably expected to lead to differences in drug

use. These included age, gender, race, income, obesity, smoking, and insurance status. We identified

the drugs most commonly used by each of these groups, then compared them to identify groups

that would lead to a set of diverse market baskets. The data on utilization by these groups became

the basis for the market baskets of drugs used in this analysis.

One issue we identified in our literature review was the idea that when comparing Canadians’ drug

costs to U.S. costs, it might be more appropriate to determine drugs that are most commonly used

by Canadians and compare their prices to prices of drugs most commonly used in the United States.

The research question for this project, however, concerns the prices that United States consumers

would pay for their drugs if they bought the same drugs in Canada. For this reason, we focused on

identifying the drugs, forms, and strengths that are most commonly used in the United States.

Identifying Potential Market Baskets

As potential market baskets, we selected a group of drugs that represent a third of the prescriptions

for each potential population group. Because this list of drugs is based on prescription volume, not

total costs, inexpensive drugs will count as heavily as higher-priced drugs. Selecting a consistent

percentage of utilization across all population groups will allow for the most consistent comparisons

across population groups. We selected the cutoff of one third of prescriptions because of sample

size considerations.

MEPS collects information on prescriptions at two different levels of detail. The most aggregated

information collected is the drug name. In addition, MEPS collects for many (but not all)

prescriptions the NDC code. NDC codes specify not only the drug but also the manufacturer, the

form and the strength of a drug. There can be many NDC codes for an individual drug name,

particularly for generic drugs with multiple manufacturers.

NORC at the University of Chicago Task Order HHSP233000007T

To select the top third of drugs for each population group, we chose to rank drugs by number of

prescriptions at the drug name level. Before doing this, we cleaned drug names in MEPS to make

them more consistent, and to remove any differentiation between forms and strengths that were

included in the reported drug name (e.g., amoxil and amoxil bubblegum).

We considered but rejected the option of using NDC codes to rank-order the drugs. Because

generic drugs have multiple manufacturers as well as multiple strengths, the use of any one generic

drug is diluted across many different NDCs. Ranking by NDCs would disproportionately place

brand-name drugs in the top third of utilization. In addition, the number of observations at the

NDC level is much smaller than the number of observations at the drug name level. This makes it

more likely that drugs could make it into the top third by random chance.

Selecting Population Groups

Using the top-ranked drug names that add up to one third of total prescriptions for that group, we

ran correlations among the drugs that were common to any two subpopulations. Where correlations

are not high, it suggests that two subpopulations are using a substantially different set of drugs.

Overall, we found that age groups provided the most diverse set of drugs. In other words, the drugs

used by children are different from those used by middle-aged individuals, which are in turn

different from those used by seniors. For the other variables we tested, the differences were

generally less striking.

Age

We divided the population into six age categories: under 12, 12 to 24, 25 to 39, 40 to 54, 55 to 64,

and 65 and up. In general, there was relatively high correlation between adjoining age categories –

and lower correlations as the age separation became greater.

Figure A-1. Rank Order Correlation of Drugs Used, by Age Group

Under 12 12-24 25-39 40-54 55-64 65 and up

Under 12

12-24 0.77

25-39 0.25 0.50

40-54 -0.23 -0.34 0.09

55-64 -0.31 -0.48 -0.17 0.83

65 and up -0.38 -0.53 -0.29 0.58 0.81

Based on these correlations, we used age as a primary basis for our market baskets. We combined

two age groups (40-54 and 55-64) since their correlation is quite high (0.83). The correlation

between the 55-64 and the over 65 group is also high, but because the latter group is mostly covered

by Medicare, we kept it as a separate group. We also considered combining the two youngest

groups. However, because further examination suggests that the under 12 group often uses

different forms of drugs (e.g., liquids instead of tablets or capsules), we kept this group as a separate

market basket.

NORC at the University of Chicago Task Order HHSP233000007T

Sex

The pattern of drug use between males and females was moderately correlated (0.61). There are

biologically driven differences between some of the drugs used by males versus those used by

females. For example, the market baskets for women of childbearing age include multiple birth

control pills. As a result, we divided all the age groups described above into separate market baskets

for males and females.

Race

We established four race categories: Non-Hispanic White, African American, Hispanic, and Other.

The correlations among the patterns of drug utilization for these groups were only modest.

Figure A-2. Rank Order Correlation of Drugs Used, by Race

White African American Hispanic Other

White

African

American

0.46

Hispanic 0.55 0.46

Other 0.64 0.43 0.81

Based on these results, we created three market baskets: one based on the utilization of the African

American population, one based on the Hispanic population, and one based on the non-Hispanic

white population. The “other” race group is too small to provide reliable estimates (and is relatively

highly correlated with the Hispanic group).

Income

We divided the population into five income groups. Generally, drug use for the four lowest groups

was fairly highly correlated. The one outlier group is the high-income population. We did not use

this group as a separate market basket.

Figure A-3. Rank Order Correlation of Drugs Used, by Income

Poor Near poor Low Medium High

Poor

Near poor 0.90

Low 0.83 0.78

Medium 0.74 0.81 0.86

High 0.44 0.59 0.58 0.85

NORC at the University of Chicago Task Order HHSP233000007T

Insurance Status

We created five categories of insurance status: (1) Uninsured (not covered by insurance during the

year); (2) Private insurance (covered by private insurance during the entire year); (3) Medicare

(covered by Medicare during the entire year); (4) Medicaid (covered by Medicaid during the entire

year); and (5) Medicaid/other (covered by Medicaid during the entire year and also covered by some

other insurance for at least part of the year). The patterns of drug use are only moderately correlated

across these categories. The most distinct pattern is for those who are uninsured, presumably

because they cannot afford some of the more expensive drugs and because their health needs are

different. We created two market baskets based on these groups: one for people who had drug

insurance all year, and one for people who did not have drug insurance for any part of the year.

Figure A-4. Rank Order Correlation of Drugs Used, by Insurance Status

Uninsured Private Medicare Medicaid Medicaid/other

Uninsured

Private 0.37

Medicare 0.21 0.67

Medicaid 0.43 0.46 0.48

Medicaid/other 0.12 0.60 0.79 0.65

Health Indicators

We chose two health indicators as possible influences on drug use: obesity and smoking. For each

indicator, the sets of drugs used are similar across the two categories. Drugs used by those who are

obese are highly correlated (0.87) with drugs utilization by the non-obese. The correlation of drug

use between smokers and non-smokers is nearly as high (0.79). As a result, we did not create any

market baskets based on these groupings.

Population Groups

Based on the above analysis, we created 15 market baskets:

• Females under 12

• Females 12-24

• Females 25-39

• Females 40-64

• Females 65+

• Males under 12

• Males 12-24

• Males 25-39

• Males 40-64

• Males 65+

• Insured

• Uninsured

• Whites

• African Americans

• Hispanics

NORC at the University of Chicago Task Order HHSP233000007T

The resulting market baskets each included from 8 to 32 drugs. We have included the list of drugs

that make up the market basket for each of these groups at the end of this appendix. In addition,

we have included a consolidated list of all 106 drugs included in any market basket, with a count of

the market baskets in which each drug is included.

Working with Price Data

For each market basket, we identified prices for the most common form and strength of each drug.

Identifying prices in Canada required a matching process to ensure that we found prices for the drug

a U.S. consumer would find as the closest match for his or her drug. We then calculated a per-

prescription weighted average price for each market basket. This section provides more detail on

this process.

Matching Drug Names to Prices

Although utilization can be summarized at the drug name level, price varies by form, strength, and

manufacturer. Even when two population groups have the same drug name in their market baskets,

they will often have different usage patterns for that drug. For example, many drugs have different

dosage recommendations for children and the elderly. In addition, children are more likely than

other populations to use drugs in a liquid form.

We identified the form and strength most commonly used by each population group for each drug

in a market basket. IMS then extracted all retail prescription transactions occurring in Oct-Dec 2005

for all NDC codes associated with that form and strength of the drug for which valid price and

quantity information was reported. We used these data to calculate an average price per pill (or, for

drugs that are not in pill form, the price per unit), defined as summed price across all dispensed

prescription transactions divided by summed quantity.

We then converted these per-unit prices into per-prescription prices, in order to standardize across

different drugs that have different numbers of pills or other units per prescription. To do this, we

used MEPS to determine the median number of pills (or other units) in a prescription for the

selected form and strength of each drug for each population group. We then multiplied the price

per unit by the median units per prescription to arrive at a standardized per-prescription price.

In the case of drugs sold in inhalers, we did not use this standardization process. Instead, we

defined the per-prescription price as summed price across all dispensed prescription transactions

divided by total number of dispensed prescriptions.

Matching with Canadian Drugs

We have tried to assure that the match between our U.S. market baskets and Canadian market

baskets is as close as possible, while still representing the utilization of the population groups we

identified. In the majority of cases, we were able to find an exact match by drug name, form, and

strength between the United States and Canada. There are four general categories of cases for

which that was not possible: 1) Drugs sold under a different name in Canada; 2) Drugs with

NORC at the University of Chicago Task Order HHSP233000007T

different brand/generic status; 3) Drugs sold in different dosages or release forms; and 4) drugs not

marketed in Canada.

Drugs sold under a different name

We first looked for an exact match by drug name, form, and strength between the United States and

Canada. When an exact match by drug name was not available in Canada, we looked to match drugs

by chemical entity, brand-name status, and in the case of brand name drugs, by manufacturer.

Figure A-5 shows the seven brand name drugs that match by all factors except trade name.

Figure A-5. On-Patent Drugs with Same Manufacturers but Different Names

Chemical ingredient U.S. Product

(from MEPS)

Canadian Product

Amoxicillin and clavulanate potassium Augmentin Clavulin

Desloratadine Clarinex Aerius

Divalproex sodium Depakote Epival

Ethinyl estradiol and norelgestromin Ortho-Evra Evra

Ethinyl estradiol and norethindrone Ortho-Novum Ortho 1/35

Escitalopram Lexapro Cipralex

Certirizine Zyrtec Reactine

For six generic drugs and branded generics, we were able to find a Canadian match with the same

active ingredient but with different names and manufacturers. Figure A-6 shows the Canadian

products we used as matches for these drugs.

Figure A-6. Generic Drugs with Different Names and Manufactures in the U.S. and Canada

Chemical ingredient U.S. Product

(from MEPS)

Canadian Product

Albuterol Albuterol Salbutamol

Spironolactone Spironolactone Novo-Spiraton

Ethinyl estradiol and desogestrel Apri Ortho-cept

Ethinyl estradiol and norethindrone Necon Brevicon, Loestrin, Minestrin

Amoxicillin Trimox Amoxicillin

Levothyroxine Levoxyl Eltroxin

Reactine is available over-the-counter in Canada.

Salbutamol sulfate is the name recommended by the World Health Organization for the drug known as Albuterol

sulfate in the U.S.

These three drugs all have the same ingredients as Necon and all have very similar prices in Canada. We propose to

use a simple average of their prices.

In the United States, branded generic versions of levothyroxine include Levoxyl, Unithroid, Levo-T, Levolet, and

Novothyrox. In the United States, Levoxyl is the only one common enough to appear in our market baskets. In

NORC at the University of Chicago Task Order HHSP233000007T

We consider all of these matches to be extremely close matches that are equivalent across the two

countries’ market baskets.

Drugs with different generic status

We found one drug that is available as a generic in the United States and not in Canada: Enalapril.

For this drug, we kept the generic version in the U.S. market baskets, but substituted the brand-

name equivalent (Vasotec) in the Canadian market baskets. This is a less equivalent match than the

cases describe above, because generics and brand drugs tend to have different pricing strategies

within each country, and price comparisons between the two countries behave differently for brands

and generics. Specifically, the ratio between the Canadian price and the U.S. price tends to be higher

for generics than for brands. However, because Vasotec is the closest possible match that a U.S.

consumer would find if they looked for Enalapril in Canada, we believe using it is the best strategy

in this case. This drug is included in one market basket.

In MEPS, all insulin appears to be labeled as insulin, regardless of manufacturer. We have used

Humulin to represent insulin in both the U.S. and Canadian market baskets.

Drugs sold in different dosages or release forms

There were several drugs on our list at a dosage level not available in Canada. When possible, we

used another dosage level that is available in both countries as a substitute in both the U.S. and

Canada market baskets.

For the five cases shown in Figure A-7, we were able to substitute the same dosage levels for both

U.S. and Canadian market baskets.

Figure A-7. Dosage Levels Substituted in U.S. and Canadian Market Baskets

Drug Original Strength as

determined by MEPS

(not available in Canada;

not used in either market

basket)

Substitution

Strength

(both U.S. and

Canada)

Market Baskets

Affected

Allegra 180 mg 60 mg 10

Ibuprofen 800 mg 600 mg 2

Levoxyl/Eltroxin 0.125 mg 0.15 mg 1

Prednisone 10 mg 5 mg 5

Prednisone 20 mg 50 mg 1

Canada, versions of levothyroxine include Euthyrox, Eltroxin, and Levo-T. Eltroxin is the most commonly used of these

three Canadian drugs.

NORC at the University of Chicago Task Order HHSP233000007T

For four drugs, there is no dosage that is available in both countries. For these drugs, we used the

original strength in the U.S. market baskets and substituted the closest strength available in Canada

in the Canadian market baskets, as shown in Figure A-8.

Figure A-8. Dosage Levels Substituted in Canadian Market Baskets Only

Drug Original Strength as determined by

MEPS

(not available in Canada;

used in U.S. market basket)

Substitution

Strength (only in

Canada)

Market Baskets

Affected

Amoxil 400 mg/5ml 250mg/5ml 1

Flonase .05% 50 mcg 5

Flovent 44 mcg 50 mcg 1

Albuterol 90 mcg 100 mcg 14

In general, the relationship between dosage and price is not linear. That is, the price of a single 100

mg pill is rarely equal to the price of two 50 mg pills of the same drug. The differences in dosage for

these drugs should not substantially affect the price comparisons between the two countries. Again,

if a U.S. consumer were looking for these drugs, these dosages are the options they would have in

Canada.

Drugs not available in Canada

There were five drugs on our list from the United States that are not available in Canada: Toprol

XL, Glucotrol, hydrocodone with acetaminophen, Lotrel, and Vicodin. We propose dropping all of

these drugs except Toprol XL from the U.S. market baskets because no close match is available. In

addition, we propose dropping Softclix, a testing device that was listed in the MEPS drug data that is

not available in Canada. An alternative would be to include the U.S. price in the Canadian market

basket, because consumers seeking these drugs would have to buy them in the United States. This

strategy would decrease any potential differences between the two market baskets. In the case of

Toprol XL (metoprolol succinate), we have substituted Lopressor SR (metoprolol tartrate) in the

Canadian market baskets.

Figure A-9. Drugs Not Available in Canada

Drug or Device Market Baskets

Affected

Toprol XL 9

Glucotrol 1

Hydrocodone and Acetaminophen 3

Lotrel 2

Vicodin 1

Softclix 10

NORC at the University of Chicago Task Order HHSP233000007T

Weighting

The process outlined above provides a representative price per prescription for each drug name in a

population group’s market basket. We then combined these prices into a single index value for each

market basket. We calculated each population group’s index as the average price per prescription,

weighted by the volume of prescriptions for that group, as determined by MEPS data.

Because the research question for this project concerns the prices that U.S. customers would pay if

they bought their drugs in Canada, we used these U.S.-based volume weights for both U.S. prices

and Canadian prices.

Currency conversion

To convert Canadian prices into U.S. dollars, we used the average exchange rate for the fourth

quarter of 2005, which was $1.1732 Canadian for $1.00 U.S..

Because the focus of this project is

on the prices that U.S. customers would face in Canada, we did not consider using purchasing power

parity or other measures that would better represent the cost of Canadian drugs to Canadian

consumers.

Sensitivity to Drug Substitutions

To explore the sensitivity of our results to the substitution we made in the market baskets, we

calculated what the differences between U.S. and Canada prices would have been if we had excluded

two drugs: Zyrtec and Toprol XL. Zyrtec is a on on-patent prescription drug in the United States

that is available over-the-counter in Canada as Reactine, and Toprol XL is a drug for which we used

a close substitute, Lopressor SR, as the Canadian match.

In all market baskets, excluding these two drugs makes the difference between the two countries’

prices smaller. In thirteen of the fifteen market baskets, the difference is 2.2 percentage points or

less (regardless of payer). The exception is the market baskets for children under 12. Because

children take fewer drugs, these market baskets are smaller than the other market baskets – the lists

of the top third of drugs include just ten drugs for girls and eight drugs for boys. Zyrtec accounts

for about a tenth of the prescriptions included in each of these market baskets, and its exclusion

gives generic drugs a larger share of each market basket. The exclusion of Zyrtec causes the average

U.S. price per prescription for these market baskets to fall by about three dollars (6 to 10 percent),

while the average Canadian price per prescription rises slightly (1 to 4 percent). The result is a larger

drop in the relative difference between U.S. and Canadian prices for these two market baskets. This

is particularly notable for prescriptions purchased with third party coverage, where the difference

between the United States and Canada falls to 21.5 percent for girls and 23.5 percent for boys.

http://www.bankofcanada.ca/en/rates/exchange.html

NORC at the University of Chicago Task Order HHSP233000007T

Figure A-10. Sensitivity of Difference Between U.S. and Canadian Price to Inclusion of

Zyrtec and Toprol XL

With Third-Party Coverage Without Third-Party Coverage

Including

Toprol, Zyrtec

Excluding

Toprol, Zyrtec

Including

Toprol, Zyrtec

Excluding

Toprol, Zyrtec

Females under 12

-30.1% -21.5% -46.9% -41.5%

Females 12-24

-39.9% -38.6% -49.5% -48.4%

Females 25-39

-47.0% -46.0% -55.0% -54.1%

Females 40-64

-40.9% -39.5% -48.7% -47.4%

Females 65+

-37.3% -36.0% -44.0% -42.7%

Males under 12

-31.0% -23.5% -46.3% -41.1%

Males 12-24

-33.4% -31.5% -44.7% -43.1%

Males 25-39

-42.1% -40.8% -50.9% -49.7%

Males 40-64

-35.9% -34.8% -43.9% -42.9%

Males 65+

-30.6% -29.0% -38.9% -37.4%

Insured

-40.3% -38.7% -48.4% -46.9%

Uninsured

-31.4% -29.9% -40.6% -39.3%

White

-39.0% -37.1% -47.2% -45.4%

African Americans

-34.9% -32.9% -43.3% -41.5%

Hispanics

-33.6% -31.3% -44.2% -42.3%

mean

-36.5% -34.1% -46.2% -44.2%

median

-35.9% -34.8% -46.3% -42.9%

Figure A-11. Zyrtec And Toprol XL as Share of Utilitzation Included in Market Baskets

Zyrtec Toprol XL

F 12U 9.4% -

F 12-24 3.2% -

F 25-39 3.6% -

F 40-64 2.3% 2.8%

F 65+ - 4.9%

M 12U 10.7% -

M 12-24 6.0% -

M 25-39 5.3% -

M 40-64 - 4.8%

M 65+ - 4.6%

Insured 2.9% 3.1%

Uninsured - 3.9%

White 2.5% 3.7%

African Americans 2.6% 2.6%

Hispanics 2.2% 2.1%

NORC at the University of Chicago Task Order HHSP233000007T

Changes in Drug Use Since 2003

The MEPS data we used are from 2003. Because there are so many ongoing changes in the market,

and because more recent data are not available for all the specific subpopulations, it is impossible to

thoroughly update our sample to reflect current drug utilization. Rather than tinkering with some

drugs and leaving others untouched, we kept utilization as-is in the MEPS data. The exception is

drugs that are no longer on the market. We removed one off-market drug, Vioxx, from our market

baskets and adjusted weights accordingly.

Figure A-12 gives a snapshot of how rankings changed in IMS data for the most popular drugs from

2003 to 2005. Our market baskets include most of these top drugs -- 18 of the 20 top drugs in 2005,

and 19 of the top 20 drugs in 2003. Levothyroxine is one top 20 drug in 2005 that is not in any of

our market baskets. It was not on the top 20 in 2003, and in fact saw nearly 200 percent growth in

utilization between 2004 and 2005 alone. Because of this recent rapid growth, it seems reasonable

that levothyroxine was not in the top third of prescriptions for any of our groups in 2003.

Ambien is the one drug on both years’ top 20 lists that is not in any of our market baskets.

Utilization for Ambien was not high enough to be included in the top third of drugs for any single

population group in the 2003 MEPS data.

Figure A-12. Top 20 Drugs in 2003 and 2005

2005

Rank

2003

Rank Product

Market Baskets

Including this Drug

1 1 Lipitor 10

2 3 Hydrocodone and Acetaminophen (Mallinckrodt) 3

3 4 Norvasc 9

4 6 Toprol-XL 9

5 2 Synthroid 11

6 5 Zoloft 7

7 7 Hydrocodone and Acetaminophen (Watson) 3

8 10 Amoxicillin 9

9 >20 Lexapro 2

10 11 Albuterol 14

11 8 Zocor 9

12 >20 Nexium 6

13 >20 Levothyroxine 0

14 18 Ambien 0

15 >20 Singulair 7

16 9 Prevacid 8

17 >20 Plavix 2

18 12 Zithromax Z-Pak 11

19 19 Fosamax 4

20 14 Zyrtec 11

>20 13 Premarin 7

>20 15 Atenolol 10

>20 16 Levoxyl 8

>20 17 Celebrex 9

>20 20 Allegra 10

Source: IMS 2003a and IMS 2005.

NORC at the University of Chicago Task Order HHSP233000007T

Prescriptions vs. Spending

We used the number of prescriptions for each drug in each population group to determine the most

popular drugs and to weight the results. We selected this measure because we wanted our market

baskets to include both inexpensive and expensive drugs, as long as they were widely used.

Another option would have been to use total spending for each drug. Our concern was that this

would skew our sample to include more costly drugs, with disproportionately large price differences.

Figure A-13 gives an illustrative example of how spending-based market baskets might differ from

our utilization-based market baskets. Of the 20 drugs that had the most prescriptions in the United

States in 2003, over half were not on the list of the top 20 drugs with the highest spending, and vice

versa. However, many of the drugs on the highest spending list were still included in at least one

market basket. The exceptions are Epogen, Oxycontin, Procrit, Protonix, and Zyprexa.

Figure A-13. Top 20 Drugs in 2003, by Prescriptions and Spending

Rank by

Prescriptions

Rank by

Spending Product

Market

Baskets

1 1 Lipitor 10

2 >20 Synthroid 11

3 >20 Hydrocodone and Acetaminophen (Mallinckrodt) *

4 13 Norvasc 9

5 8 Zoloft 7

6 >20 Toprol-XL 9

7 >20 Hydrocodone and Acetaminophen (Watson) *

8 2 Zocor 9

9 3 Prevacid 8

10 >20 Amoxicillin 9

11 >20 Albuterol 14

12 >20 Zithromax Z-Pak 11

13 >20 Premarin 7

14 >20 Zyrtec 11

15 >20 Atenolol 10

16 >20 Levoxyl 7

17 9 Celebrex 9

18 >20 Ambien 0

19 18 Fosamax 4

20 >20 Allegra 10

>20 11 Advair Diskus 5

>20 14 Effexor XR 6

>20 6 Epogen 0

>20 10 Neurontin 2

>20 7 Nexium 6

>20 17 Oxycontin 0

>20 12 Plavix 2

>20 15 Pravachol 4

>20 4 Procrit 0

>20 19 Protonix 0

>20 16 Risperdal 1

>20 20 Vioxx N/A

>20 5 Zyprexa 0

Source: IMS 2003a and IMS 2003b.

*Hydrocodone and Acetaminophen was dropped from the study because it is not available in Canada.

NORC at the University of Chicago Task Order HHSP233000007T

Payer Type and Retail vs. Total Price

This project is more detailed than many in our literature review because of our ability to collect data

by payer type. This allows for a level of standardization that is rare in many comparable studies. It

does not allow us, however, to look at total price for cases with third party payers. There are likely

additional transactions between manufacturers and third party payers, such as rebates, that affect

total price but that are not reflected in our data.

The prescriptions identified in this paper as having third-party coverage are those for which a

customer had some sort of third party involvement at the pharmacy counter. These transactions

include customers who presented an insurance card at the time of the transaction. They also include

customers using discount cards. Although these customers do not receive insurance coverage for

their prescription drugs, they have access to negotiated discounts that make their prices similar to

the prices paid for drugs that are covered by insurance.

The prescriptions identified in this report as being without third-party coverage are those for which

a customer did not have third-party involvement at the pharmacy counter. While we expect that

most of these customers do not have insurance, it is possible that some have indemnity-type

insurance. These customers may be paying the full cost of drugs out-of-pocket at the drugstore, but

submitting receipts for reimbursement. However, without the involvement of an insurer at the time

of the transaction, the full price paid for these prescriptions is not discounted.

Market Baskets

Figure A-13. Market Basket for Females Under 12

Drug Name Status

Substitute

In Canada

Most Common

Form And

Strength

Substituted

In U.S.

Strength

Substituted

In Canada Weight

ALBUTEROL Off Patent Non-Originator Non-Trademark Salbutamol INH 90 100 0.162

AMOXICILLIN Off Patent Non-Originator Non-Trademark LIQ 250/5 0.252

AMOXIL Off Patent Originator LIQ 400/5 250/5 0.058

AUGMENTIN Off Patent Originator Clavulin LIQ 400/5 0.064

CONCERTA Off Patent Non-Originator Trademark TAB 18 0.042

FLOVENT On Patent INH 44 50 0.043

SINGULAIR On Patent ORA 4 0.080

TRIMOX Off Patent Non-Originator Trademark Amoxicillin LIQ 250/5 0.042

ZITHROMAX Off Patent Originator LIQ 200/5 0.163

ZYRTEC On Patent Reactine TAB 10 0.094

Figure A-14. Market Basket for Females 12-24

Drug Name Status

Substitute In

Canada

Most Common

Form And

Strength

Substituted

In U.S.

Strength

Substituted

In Canada Weight

ADVAIR On Patent DSK 100/50 0.038

ALBUTEROL Off Patent Non-Originator Non-Trademark Salbutamol INH 90 100 0.085

ALESSE Off Patent Originator TAB 0.02/0.1 0.035

ALLEGRA On Patent TAB 180 60 60 0.062

AMOXICILLIN Off Patent Non-Originator Non-Trademark CAP 500 0.091

APRI Off Patent Non-Originator Trademark Ortho-Cept TAB 0.15/0.03 0.035

PLAN B On Patent TAB 0.75 0.054

CELEXA Off Patent Originator TAB 20 0.043

FLONASE On Patent INH 0.05% 50 Mcg 0.032

IBUPROFEN Off Patent Non-Originator Non-Trademark TAB 600 0.031

ORTHO EVRA On Patent Evra TDM 20/150/24 0.031

ORTHO TRI-

CYCLEN

On Patent TAB -9 0.224

PAXIL On Patent TAB 20 0.038

SINGULAIR On Patent TAB 10 0.037

ZITHROMAX Off Patent Originator TAB 250 0.076

ZOLOFT On Patent TAB 50 0.057

ZYRTEC On Patent Reactine TAB 10 0.032

Figure A-15. Market Basket for Females 25-39

Drug Name Status Substitute In Canada

Most Common

Form And

Strength

Substituted

In U.S.

Strength

Substituted

In Canada Weight

ALBUTEROL Off Patent Non-Originator Non-Trademark Salbutamol INH 90 100 0.062

ALLEGRA On Patent TAB 180 60 60 0.060

AMOXICILLIN Off Patent Non-Originator Non-Trademark CAP 500 0.034

APRI Off Patent Non-Originator Trademark Ortho-Cept TAB 0.15 0.041

PLAN B On Patent TAB 0.75 0.042

CELEXA Off Patent Originator TAB 20 0.027

EFFEXOR On Patent TAB 75 0.032

FLONASE On Patent INH 0.05% 50 Mcg 0.033

IBUPROFEN Off Patent Non-Originator Non-Trademark TAB 800 600 600 0.028

LEVOXYL Off Patent Non-Originator Trademark Eltroxin TAB 0.15 0.029

LEXAPRO On Patent Cipralex TAB 10 0.028

NECON Off Patent Non-Originator Trademark

Brevicon, Loestrin,

Minestrin

TAB 35/1 0.029

ORTHO EVRA On Patent Evra TDM 20/150/24 0.027

ORTHO TRI-

CYCLEN On Patent TAB -9 0.131

ORTHO-NOVUM Off Patent Originator Ortho 1/35 TAB 1/35 0.038

PAXIL On Patent TAB 20 0.051

PREDNISONE Off Patent Non-Originator Non-Trademark TAB 10 5 5 0.025

SYNTHROID Off Patent Originator TAB 0.075 0.085

WELLBUTRIN Off Patent Originator TAB 150 0.029

YASMIN On Patent TAB 3/0.03 0.039

ZITHROMAX Off Patent Originator TAB 250 0.049

ZOLOFT On Patent TAB 100 0.043

ZYRTEC On Patent Reactine TAB 10 0.036

Figure A-16. Market Basket for Females 40-64

Drug Name Status Substitute In Canada

Most

Common

Form And

Strength

Substituted

In U.S.

Strength

Substituted

In Canada Weight

ALBUTEROL Off Patent Non-Originator Non-Trademark Salbutamol INH 90 100 0.030

ALLEGRA On Patent TAB 180 60 60 0.037

ATENOLOL Off Patent Non-Originator Non-Trademark TAB 50 0.046

CELEBREX On Patent CAP 200 0.030

CELEXA Off Patent Originator TAB 20 0.026

EFFEXOR On Patent CAP 150 0.029

FLUOXETINE Off Patent Non-Originator Non-Trademark CAP 20 0.024

FOSAMAX On Patent TAB 70 0.027

FUROSEMIDE Off Patent Non-Originator Non-Trademark TAB 40 0.023

GLUCOPHAGE Off Patent Originator TAB 500 0.025

HYDROCHLOROTHIAZIDE Off Patent Non-Originator Non-Trademark TAB 25 0.044

LEVOXYL Off Patent Non-Originator Trademark Eltroxin TAB 0.1 0.033

LIPITOR On Patent TAB 10 0.082

LISINOPRIL Off Patent Non-Originator Non-Trademark TAB 10 0.032

METFORMIN Off Patent Non-Originator Non-Trademark TAB 500 0.024

NEURONTIN On Patent CAP 300 0.023

NEXIUM On Patent CAP 40 0.027

NORVASC On Patent TAB 5 0.032

PAXIL On Patent TAB 20 0.034

PREMARIN Off Patent Originator TAB 0.625 0.077

PREVACID On Patent CAP 30 0.034

SOFTCLIX On Patent Not Included

SYNTHROID Off Patent Originator TAB 0.1 0.090

TOPROL XL Off Patent Originator Lopressor SR TAB 50 0.028

TRAZODONE Off Patent Non-Originator Non-Trademark TAB 50 0.022

ZITHROMAX Off Patent Originator TAB 250 0.027

ZOCOR On Patent TAB 20 0.030

ZOLOFT On Patent TAB 100 0.042

ZYRTEC On Patent Reactine TAB 10 0.023

Figure A-17. Market Basket for Females 65+

Drug Name Status Substitute In Canada

Most

Common

Form And

Strength

Substituted

In U.S.

Strength

Substituted

In Canada Weight

ATENOLOL Off Patent Non-Originator Non-Trademark TAB 50 0.061

CELEBREX On Patent CAP 200 0.040

DIOVAN On Patent TAB 160 0.037

EVISTA On Patent TAB 60 0.024

FOSAMAX On Patent TAB 70 0.066

FUROSEMIDE Off Patent Non-Originator Non-Trademark TAB 40 0.078

HYDROCHLOROTHIAZIDE Off Patent Non-Originator Non-Trademark TAB 25 0.054

LEVOXYL Off Patent Non-Originator Trademark Eltroxin TAB 0.1 0.037

LIPITOR On Patent TAB 10 0.099

LISINOPRIL Off Patent Non-Originator Non-Trademark TAB 10 0.049

METFORMIN Off Patent Non-Originator Non-Trademark TAB 500 0.026

METOPROLOL Off Patent Non-Originator Non-Trademark TAB 50 0.038

NORVASC On Patent TAB 5 0.068

PLAVIX On Patent TAB 75 0.031

PREMARIN Off Patent Originator TAB 0.625 0.048

PREVACID On Patent CAP 30 0.026

SOFTCLIX On Patent Not included

SPIRONOLACTONE Off Patent Non-Originator Non-Trademark Novo-Spiroton TAB 25 0.025

SYNTHROID Off Patent Originator TAB 0.1 0.086

TOPROL XL Off Patent Originator Lopressor SR TAB 50 0.049

ZOCOR On Patent TAB 20 0.058

Figure A-18. Market Basket for Males Under 12

Drug Name Status

Substitute

In Canada

Most Common

Form And

Strength

Substituted

In U.S.

Strength

Substituted

In Canada Weight

ADDERALL Off Patent Non-Originator Trademark TAB 10 0.056

ALBUTEROL Off Patent Non-Originator Non-Trademark Salbutamol INH 90 100 0.214

AMOXICILLIN Off Patent Non-Originator Non-Trademark LIQ 250/5 0.203

AUGMENTIN Off Patent Originator Clavulin LIQ 400/5 0.055

CONCERTA Off Patent Non-Originator Trademark TAB 18 0.094

SINGULAIR On Patent ORA 5 0.144

ZITHROMAX Off Patent Originator LIQ 200/5 0.127

ZYRTEC On Patent Reactine TAB 10 0.107

Figure A-19. Market Basket for Males 12-24

Drug Name Status

Substitute

In Canada

Most Common

Form And

Strength

Substituted

In U.S.

Strength

Substituted

In Canada Weight

ADDERALL Off Patent Non-Originator Trademark CAP 30 0.094

ADVAIR On Patent DSK 100/50 0.041

ALBUTEROL Off Patent Non-Originator Non-Trademark Salbutamol INH 90 100 0.100

ALLEGRA On Patent TAB 180 60 60 0.089

AMOXICILLIN Off Patent Non-Originator Non-Trademark CAP 500 0.078

CLARINEX On Patent Aerius TAB 5 0.037

CONCERTA Off Patent Non-Originator Trademark TAB 54 0.101

EFFEXOR On Patent TAB 75 0.061

MINOCYCLINE Off Patent Non-Originator Non-Trademark CAP 100 0.047

NASONEX On Patent INH 50 0.047

RISPERDAL On Patent TAB 0.5 0.065

SINGULAIR On Patent TAB 10 0.052

STRATTERA On Patent CAP 40 0.043

ZITHROMAX Off Patent Originator TAB 250 0.083

ZYRTEC On Patent Reactine TAB 10 0.060

Figure A-20. Market Basket for Males 25-39

Drug Name Status

Substitute In

Canada

Most Common Form And

Strength

Substituted

In U.S.

Strength

Substituted

In Canada Weight

ACETAMINOPHEN

HYDROCODONE

Off Patent Originator Not on market

ADVAIR On Patent DSK 100/50 0.038

ALBUTEROL Off Patent Non-Originator Non-Trademark Salbutamol INH 90 100 0.047

ALLEGRA On Patent TAB 180 60 60 0.094

ALLOPURINOL On Patent Dropped due to small sample size.

AMOXICILLIN Off Patent Non-Originator Non-Trademark CAP 500 0.052

ATENOLOL Off Patent Non-Originator Non-Trademark TAB 50 0.036

CELEBREX On Patent CAP 200 0.026

CLARITIN Off Patent Originator LIQ 10/10 0.033

DEPAKOTE Off Patent Originator Epival TAB 500 0.039

EFFEXOR On Patent TAB 75 0.027

FLONASE On Patent INH 0.05% 50 Mcg 0.035

IBUPROFEN Off Patent Non-Originator Non-Trademark TAB 800 600 600 0.037

LEVOXYL Off Patent Non-Originator Trademark Dropped due to small sample size.

LEXAPRO On Patent Cipralex TAB 10 0.040

LIPITOR On Patent TAB 10 0.066

LISINOPRIL Off Patent Non-Originator Non-Trademark TAB 10 0.035

NEURONTIN On Patent TAB 600 0.033

NEXIUM On Patent CAP 40 0.063

PAXIL On Patent TAB 20 0.047

ROXICODONE On Patent Dropped due to small sample size.

SOFTCLIX On Patent Not Included

SYNTHROID Off Patent Originator TAB 0.1 0.045

VICODIN Off Patent Non-Originator Trademark Not on market

WELLBUTRIN Off Patent Originator TAB 150 0.031

ZITHROMAX Off Patent Originator TAB 250 0.069

ZOLOFT On Patent TAB 50 0.053

ZYRTEC On Patent Reactine TAB 10 0.053

Figure A-21. Market Basket for Males 40-64

Drug Name Status

Substitute In

Canada

Most

Common

Form And

Strength

Substituted

In U.S.

Strength

Substituted

In Canada Weight

ACCUPRIL Off Patent Originator TAB 20 0.026

ACETAMINOPHEN

HYDROCODONE

Off Patent Originator Not on market

ACTOS On Patent TAB 45 0.026

ALBUTEROL Off Patent Non-Originator Non-Trademark Salbutamol INH 90 100 0.028

ALLEGRA On Patent TAB 180 60 60 0.030

ALPRAZOLAM Off Patent Non-Originator Non-Trademark TAB 0.5 0.025

ATENOLOL Off Patent Non-Originator Non-Trademark TAB 50 0.061

CELEBREX On Patent CAP 200 0.024

DIOVAN On Patent TAB 80 0.027

FUROSEMIDE Off Patent Non-Originator Non-Trademark TAB 40 0.028

GLUCOPHAGE Off Patent Originator TAB 500 0.029

HYDROCHLOROTHIAZIDE Off Patent Non-Originator Non-Trademark TAB 25 0.054

LIPITOR On Patent TAB 10 0.162

LISINOPRIL Off Patent Non-Originator Non-Trademark TAB 10 0.068

METFORMIN Off Patent Non-Originator Non-Trademark TAB 500 0.037

METOPROLOL Off Patent Non-Originator Non-Trademark TAB 50 0.029

NEXIUM On Patent CAP 40 0.036

NORVASC On Patent TAB 5 0.050

PAXIL On Patent TAB 20 0.026

PRAVACHOL On Patent TAB 40 0.032

PREVACID On Patent CAP 30 0.034

RANITIDINE Off Patent Non-Originator Non-Trademark TAB 150 0.027

SOFTCLIX On Patent Not included

SYNTHROID Off Patent Originator TAB 0.112 0.023

TOPROL XL Off Patent Originator Lopressor SR TAB 50 0.048

VIOXX Off Patent Non-Originator Trademark TAB 25 0.000

ZOCOR On Patent TAB 20 0.070

Figure A-22. Market Basket for Males 65+

Drug Name Status

Substitute In

Canada

Most

Common

Form And

Strength

Substituted

In U.S.

Strength

Substituted

In Canada Weight

ALBUTEROL Off Patent Non-Originator Non-Trademark Salbutamol INH 90 100 0.029

ASPIRIN Off Patent Non-Originator Non-Trademark TAB 325 0.028

ATENOLOL Off Patent Non-Originator Non-Trademark TAB 50 0.075

COUMADIN Off Patent Originator TAB 5 0.042

DOXAZOSIN Off Patent Non-Originator Non-Trademark TAB 2 0.038

FLOMAX On Patent CAP 0.4 0.039

FUROSEMIDE Off Patent Non-Originator Non-Trademark TAB 40 0.080

HYDROCHLOROTHIAZIDE Off Patent Non-Originator Non-Trademark TAB 25 0.053

LIPITOR On Patent TAB 10 0.113

LISINOPRIL Off Patent Non-Originator Non-Trademark TAB 10 0.088

METFORMIN Off Patent Non-Originator Non-Trademark TAB 500 0.028

METOPROLOL Off Patent Non-Originator Non-Trademark TAB 50 0.047

NORVASC On Patent TAB 10 0.046

PLAVIX On Patent TAB 75 0.034

PRAVACHOL On Patent TAB 40 0.033

SOFTCLIX On Patent Not included

SYNTHROID Off Patent Originator TAB 0.1 0.032

TERAZOSIN Off Patent Non-Originator Non-Trademark CAP 5 0.033

TOPROL XL Off Patent Originator Lopressor SR TAB 50 0.046

WARFARIN Off Patent Non-Originator Non-Trademark TAB 5 0.039

ZOCOR On Patent TAB 20 0.077

Figure A-23. Market Basket for Insured

Drug Name Status Substitute In Canada

Most

Common

Form And

Strength

Substituted

In U.S.

Strength

Substituted

In Canada Weight

ADVAIR On Patent DSK 100/50 0.022

ALBUTEROL Off Patent Non-Originator Non-Trademark Salbutamol INH 90 100 0.064

ALLEGRA On Patent TAB 180 60 60 0.037

AMOXICILLIN Off Patent Non-Originator Non-Trademark CAP 500 0.031

ATENOLOL Off Patent Non-Originator Non-Trademark TAB 50 0.045

CELEBREX On Patent CAP 200 0.025

EFFEXOR On Patent CAP 150 0.021

FLONASE On Patent INH 0.05% 50 Mcg 0.020

FUROSEMIDE Off Patent Non-Originator Non-Trademark TAB 40 0.029

GLUCOPHAGE Off Patent Originator TAB 500 0.021

HYDROCHLOROTHIAZIDE Off Patent Non-Originator Non-Trademark TAB 25 0.037

LEVOXYL Off Patent Non-Originator Trademark Eltroxin TAB 0.1 0.024

LIPITOR On Patent TAB 10 0.093

LISINOPRIL Off Patent Non-Originator Non-Trademark TAB 10 0.041

METFORMIN Off Patent Non-Originator Non-Trademark TAB 500 0.023

NEXIUM On Patent CAP 40 0.025

NORVASC On Patent TAB 5 0.036

ORTHO TRI-CYCLEN On Patent TAB -9 0.021

PAXIL On Patent TAB 20 0.027

PRAVACHOL On Patent TAB 40 0.019

PREDNISONE Off Patent Non-Originator Non-Trademark TAB 10 0.019

PREMARIN Off Patent Originator TAB 0.625 0.034

PREVACID On Patent CAP 30 0.030

SINGULAIR On Patent TAB 10 0.025

SOFTCLIX On Patent Not included

SYNTHROID Off Patent Originator TAB 0.1 0.060

TOPROL XL Off Patent Originator Lopressor SR TAB 50 0.031

VIOXX Off Patent Non-Originator Trademark TAB 25 0.000

ZITHROMAX Off Patent Originator TAB 250 0.037

ZOCOR On Patent TAB 20 0.042

ZOLOFT On Patent TAB 100 0.030

ZYRTEC On Patent Reactine TAB 10 0.029

Figure A-24. Market Basket for Uninsured

Drug Name Status Substitute In Canada

Most

Common

Form And

Strength

Substituted

In U.S.

Strength

Substituted

In Canada Weight

ACCUPRIL Off Patent Originator TAB 20 0.021

ACETAMINOPHEN

HYDROCODONE

Off Patent Originator Not on market

ALBUTEROL Off Patent Non-Originator Non-Trademark Salbutamol INH 90 100 0.041

ALPRAZOLAM Off Patent Non-Originator Non-Trademark TAB 0.25 0.024

ATENOLOL Off Patent Non-Originator Non-Trademark TAB 50 0.057

CELEBREX On Patent CAP 200 0.023

DIOVAN On Patent TAB 160 0.029

FOSAMAX On Patent TAB 70 0.027

FUROSEMIDE Off Patent Non-Originator Non-Trademark TAB 40 0.064

GLYBURIDE Off Patent Non-Originator Non-Trademark TAB 5 0.027

HYDROCHLOROTHIAZIDE Off Patent Non-Originator Non-Trademark TAB 25 0.060

HUMULIN Off Patent Non-Originator Non-Trademark INJ 100 0.033

LEVOXYL Off Patent Non-Originator Trademark Eltroxin TAB 0.05 0.030

LIPITOR On Patent TAB 10 0.076

LISINOPRIL Off Patent Non-Originator Non-Trademark TAB 10 0.057

LOTREL Off Patent Non-Originator Non-Trademark Not on market

METFORMIN Off Patent Non-Originator Non-Trademark TAB 500 0.030

METOPROLOL Off Patent Non-Originator Non-Trademark TAB 50 0.051

NORVASC On Patent TAB 10 0.049

PAXIL On Patent TAB 20 0.023

PREDNISONE Off Patent Non-Originator Non-Trademark TAB 10 0.028

PREMARIN Off Patent Originator TAB 0.625 0.034

PREVACID On Patent CAP 30 0.023

RANITIDINE Off Patent Non-Originator Non-Trademark TAB 150 0.029

SOFTCLIX On Patent Not included

SYNTHROID Off Patent Originator TAB 0.1 0.061

TOPROL XL Off Patent Originator Lopressor SR TAB 50 0.039

ZOCOR On Patent TAB 20 0.043

ZOLOFT On Patent TAB 50 0.022

Figure A-25. Market Basket for Non-Hispanic Whites

Drug Name Status Substitute In Canada

Most

Common

Form And

Strength

Substituted

In U.S.

Strength

Substituted

In Canada Weight

ALBUTEROL Off Patent Non-Originator Non-Trademark Salbutamol INH 90 100 0.032

ALLEGRA On Patent TAB 180 60 60 0.034

AMOXICILLIN Off Patent Non-Originator Non-Trademark CAP 500 0.028

ATENOLOL Off Patent Non-Originator Non-Trademark TAB 50 0.048

CELEBREX On Patent CAP 200 0.025

DIOVAN On Patent TAB 80 0.021

EFFEXOR On Patent CAP 75 0.022

FLONASE On Patent INH 0.05% 50 Mcg 0.019

FOSAMAX On Patent TAB 70 0.025

FUROSEMIDE Off Patent Non-Originator Non-Trademark TAB 40 0.040

HYDROCHLOROTHIAZIDE Off Patent Non-Originator Non-Trademark TAB 25 0.037

LEVOXYL Off Patent Non-Originator Trademark Eltroxin TAB 0.1 0.027

LIPITOR On Patent TAB 10 0.095

LISINOPRIL Off Patent Non-Originator Non-Trademark TAB 10 0.044

METFORMIN Off Patent Non-Originator Non-Trademark TAB 500 0.020

METOPROLOL Off Patent Non-Originator Non-Trademark TAB 50 0.025

NEXIUM On Patent CAP 40 0.024

NORVASC On Patent TAB 5 0.035

ORTHO TRI-CYCLEN On Patent TAB -9 0.021

PAXIL On Patent TAB 20 0.029

PREDNISONE Off Patent Non-Originator Non-Trademark TAB 10 0.020

PREMARIN Off Patent Originator TAB 0.625 0.035

PREVACID On Patent CAP 30 0.028

RANITIDINE Off Patent Non-Originator Non-Trademark TAB 150 0.019

SOFTCLIX On Patent Not included

SYNTHROID Off Patent Originator TAB 0.1 0.071

TOPROL XL Off Patent Originator Lopressor SR TAB 50 0.037

VIOXX Off Patent Non-Originator Trademark TAB 25 0.000

ZITHROMAX Off Patent Originator TAB 250 0.033

ZOCOR On Patent TAB 20 0.047

ZOLOFT On Patent TAB 100 0.031

ZYRTEC On Patent Reactine TAB 10 0.025

Figure A-26. Market Basket for African Americans

Drug Name Status

Substitute In

Canada

Most

Common

Form And

Strength

Substituted

In U.S.

Strength

Substituted

In Canada Weight

ADVAIR On Patent DSK 100/50 0.026

ALBUTEROL Off Patent Non-Originator Non-Trademark Salbutamol INH 90 100 0.075

ALLEGRA On Patent TAB 180 60 60 0.024

ATENOLOL Off Patent Non-Originator Non-Trademark TAB 50 0.053

CELEBREX On Patent CAP 200 0.023

CLONIDINE Off Patent Non-Originator Non-Trademark TAB 0.2 0.027

DIOVAN On Patent TAB 160 0.031

FUROSEMIDE Off Patent Non-Originator Non-Trademark TAB 40 0.063

GLUCOPHAGE Off Patent Originator TAB 500 0.035

GLUCOTROL Off Patent Originator Not on market

HUMULIN Off Patent Originator LIQ 100 0.022

HUMULIN Off Patent Non-Originator Non-Trademark INJ 100 0.032

HYDROCHLOROTHIAZIDE Off Patent Non-Originator Non-Trademark TAB 25 0.087

LIPITOR On Patent TAB 10 0.069

LISINOPRIL Off Patent Non-Originator Non-Trademark TAB 10 0.050

LOTREL Off Patent Non-Originator Non-Trademark Not on market

METFORMIN Off Patent Non-Originator Non-Trademark TAB 500 0.030

NORVASC On Patent TAB 10 0.081

PRAVACHOL On Patent TAB 40 0.030

PREDNISONE Off Patent Non-Originator Non-Trademark TAB 10 0.024

PREMARIN Off Patent Originator TAB 0.625 0.024

PREVACID On Patent CAP 30 0.028

SINGULAIR On Patent TAB 10 0.032

SOFTCLIX On Patent Not included

SYNTHROID Off Patent Originator TAB 0.1 0.025

TOPROL XL Off Patent Originator Lopressor SR TAB 50 0.026

ZITHROMAX Off Patent Originator TAB 250 0.024

ZOCOR On Patent TAB 20 0.033

ZYRTEC On Patent Reactine TAB 10 0.026

Figure A-27. Market Basket for Hispanics

Drug Name Status Substitute In Canada

Most

Common

Form And

Strength

Substituted

In U.S.

Strength

Substituted

In Canada Weight

ALBUTEROL Off Patent Non-Originator Non-Trademark Salbutamol INH 90 100 0.075

ALLEGRA On Patent TAB 180 60 60 0.030

AMOXICILLIN Off Patent Non-Originator Non-Trademark LIQ 250/5 0.046

ASPIRIN Off Patent Non-Originator Non-Trademark TAB 325 0.026

ATENOLOL Off Patent Non-Originator Non-Trademark TAB 50 0.043

CELEBREX On Patent CAP 200 0.036

ENALAPRIL Off Patent Non-Originator Non-Trademark Vasotec TAB 10 0.020

FUROSEMIDE Off Patent Non-Originator Non-Trademark TAB 40 0.020

GLUCOPHAGE Off Patent Originator TAB 500 0.031

GLYBURIDE Off Patent Non-Originator Non-Trademark TAB 5 0.029

HYDROCHLOROTHIAZIDE Off Patent Non-Originator Non-Trademark TAB 25 0.035

IBUPROFEN Off Patent Non-Originator Non-Trademark TAB 600 0.037

HUMULIN Off Patent Non-Originator Non-Trademark INJ 100 0.021

LEVOXYL Off Patent Non-Originator Trademark Eltroxin TAB 0.125 0.15 0.15 0.019

LIPITOR On Patent TAB 10 0.064

LISINOPRIL Off Patent Non-Originator Non-Trademark TAB 10 0.043

METFORMIN Off Patent Non-Originator Non-Trademark TAB 500 0.045

METOPROLOL Off Patent Non-Originator Non-Trademark TAB 50 0.022

NAPROXEN Off Patent Non-Originator Non-Trademark TAB 500 0.019

NEXIUM On Patent CAP 40 0.022

NORVASC On Patent TAB 10 0.029

PAXIL On Patent TAB 20 0.030

PREDNISONE Off Patent Non-Originator Non-Trademark TAB 20 50 50 0.028

PREMARIN Off Patent Originator TAB 0.625 0.022

PREVACID On Patent CAP 30 0.023

SINGULAIR On Patent TAB 10 0.035

SOFTCLIX On Patent Not included

SYNTHROID Off Patent Originator TAB 0.1 0.037

TOPROL XL Off Patent Originator Lopressor SR TAB 50 0.021

ZITHROMAX Off Patent Originator LIQ 200/5 0.044

ZOCOR On Patent TAB 20 0.027

ZYRTEC On Patent Reactine TAB 10 0.022

Figure A-28. Summary of Drugs Included in Market Baskets

Drug

/Canadian Drug Form

Strength

(Canadian)

Market

Baskets

Accupril TAB 20 2

Actos TAB 45 1

Adderall CAP 30 1

Adderall TAB 10 1

Advair DSK 100/50 5

Albuterol/Salbutamol INH 90 (100) 14

Alesse TAB 0.02/0.1 1

Allegra TAB 60 10

Alprazolam TAB 0.25 1

Alprazolam TAB 0.5 1

Amoxicillin CAP 500 6

Amoxicillin LIQ 250/5 3

Amoxil LIQ

400/5

(250/5)

Apri/Ortho-cept TAB 0.15 1

Apri/Ortho-cept TAB 0.15/0.03 1

Aspirin TAB 325 2

Atenolol TAB 50 10

Augmentin/

Clavulin

LIQ 400/5 2

Plan B TAB 0.75 2

Celebrex CAP 200 9

Celexa TAB 20 3

Clarinex/Aerius TAB 5 1

Claritin LIQ 10/10 1

Clonidine TAB 0.2 1

Concerta TAB 18 2

Concerta TAB 54 1

Coumadin TAB 5 1

Depakote/Epival TAB 500 1

Diovan TAB 160 3

Diovan TAB 80 2

Doxazosin TAB 2 1

Effexor CAP 150 2

Effexor CAP 75 1

Effexor TAB 75 3

Drug

/Canadian Drug Form

Strength

(Canadian)

Market

Baskets

Enalapril/Vasotec TAB 10 1

Evista TAB 60 1

Flomax CAP 0.4 1

Flonase INH

0.05%

(50 mcg)

Flovent INH 44 (50) 1

Fluoxetine CAP 20 1

Fosamax TAB 70 4

Furosemide TAB 40 9

Glucophage TAB 500 5

Glyburide TAB 5 2

Humulin Insulin LIQ 100 1

Humulin Insulin INJ 100 3

Hydrochlorothiazide TAB 25 9

Ibuprofen TAB 600 2

Levoxyl/Eltroxin TAB 0.05 1

Levoxyl/Eltroxin TAB 0.1 4

Levoxyl/Eltroxin TAB 0.15 2

Lexapro/Cipralex TAB 10 2

Lipitor TAB 10 10

Lisinopril TAB 10 10

Metformin TAB 500 9

Metoprolol TAB 50 6

Minocycline CAP 100 1

Naproxen TAB 500 1

Nasonex INH 50 1

Necon/Brevicon,

Loestrin, Minestrin

TAB 35/1 1

Neurontin CAP 300 1

Neurontin TAB 600 1

Nexium CAP 40 6

Norvasc TAB 10 4

Norvasc TAB 5 5

Ortho Evra/Evra TDM 20/150/ 2

Ortho Tri-cyclen TAB -9 4

Drug

/Canadian Drug Form

Strength

(Canadian)

Market

Baskets

Ortho-Novum/

Ortho 1/35

TAB 12785 1

Paxil TAB 20 9

Plavix TAB 75 2

Pravachol TAB 40 4

Prednisone TAB 5 5

Prednisone TAB 50 1

Premarin TAB 0.625 7

Prevacid CAP 30 8

Ranitidine TAB 150 3

Risperdal TAB 0.5 1

Singulair ORA 4 1

Singulair ORA 5 1

Singulair TAB 10 5

Spironolactone/

Novo-Spiraton

TAB 25 1

Strattera CAP 40 1

Synthroid TAB 0.075 1

Synthroid TAB 0.1 9

Synthroid TAB 0.112 1

Terazosin CAP 5 1

Toprol TAB 50 9

Trazodone TAB 50 1

Trimox/Amoxicillin LIQ 250/5 1

Vioxx TAB 25 3

Warfarin TAB 5 1

Wellbutrin TAB 150 2

Yasmin TAB 3/0.03 1

Zithromax LIQ 200/5 3

Zithromax TAB 250 8

Zocor TAB 20 9

Zoloft TAB 100 4

Zoloft TAB 50 3

Zyrtec/Reactine TAB 10 11

Appendix B: Literature Review

Rapidly rising prescription drug prices have caused many consumers, third-party payers, and

policymakers to look for ways to lower their drug costs. One strategy that many have seen as a

potential solution is to import drug products from countries where drug prices are lower than those

in the United States—commonly called reimportation.

Prices for on-patent drugs are controlled by federal regulation in Canada. New brand drugs are

limited to the median of prices for the new drug in France, Germany, Italy, Sweden, Switzerland, the

UK, and the United States. Limits for drugs that are modifications of existing drugs (usually known

as “me-too” drugs) are based on the breakthrough brand’s price; most new drugs cannot cost more

than the other drugs in the same therapeutic class. Increases in prices for existing on-patent drugs

are linked to a multiple of the consumer price index, often resulting in a lower rate of increase than

the drug inflation rate in the United States.

Other price controls for a significant portion of the Canadian market take place at the provincial

level. The Canadian provinces cover 42 percent of all national drug expenditures through their

coverage of prescription drugs for the elderly. The provinces all use formularies to keep prices low

for the drugs they cover, and some take further measures. Ontario has instituted price freezes;

British Columbia uses reference pricing; and Quebec requires companies to give the province the

best price given to any other province.

Legislative and regulatory price controls in the United States apply to drugs purchased through the

federal supply schedule (for sales to certain federal purchasers and other entities, such as community

health centers, that are by law authorized to use the federal supply schedule) and drugs purchased by

the Medicaid program, the Veterans Administration, and the Department of Defense. However,

insurers and pharmacy benefit managers (PBMs) are also able to negotiate substantial discounts and

rebates from retail prices. Uninsured individuals in the United States do not have access to these

lower prices, and they typically pay retail prices that are about 15 percent above those paid by people

with insurance. As a result, those without insurance have been at the forefront of the push to import

drugs from Canada. Some purchasers, including several municipal government employee plans, have

also begun to explore the possibility of savings for their insured employees from reimportation,

seeking to get better prices than those normally obtained by insurers and PBMs.

The Canadian Pricing System

While Canada does offer universal access to health insurance, that insurance covers only acute

hospital and physician services, and does not provide coverage for outpatient prescription drugs.

Most Canadians receive prescription drug coverage through their employers, although the federal

government offers coverage to some populations, including veterans and First Nations and Inuit

people, and there are publicly funded provincial programs that pay a substantial share of the costs

for older, disabled, and low-income Canadians (Gross, 2003).

Graham (2000) provides an explanation of how prescription drug prices are regulated in Canada.

Canada’s Patented Medicine Prices Review Board (PMPRB), established in 1987, regulates the prices

of newly patented drugs. The PMPRB controls the price at which drug manufacturers sell to

wholesalers. Drugs are divided into three categories by the PMPRB when they are introduced:

Category 1, or “line extension,” drugs are drugs that are a new strength of an existing drug. Prices of

new drugs in this category must bear a “reasonable relationship” to the average price of the same

medicine in the same or comparable dosage forms.

Category 2 is composed of “breakthrough” drugs, which produce a substantial improvement over

older drugs. These drugs may not exceed the higher of the cost of therapy with medicines in the

same therapeutic class and the median of prices of the same drugs in the United States, United

Kingdom, Switzerland, Sweden, Italy, Germany, and France.

Category 3 drugs, or “me-too” drugs provide moderate, little, or no improvement over existing

medications and introductory prices is presumed excessive if the cost of therapy with the new drug

is higher than the cost of therapy with comparable drugs in the same therapeutic class.

After drugs are first introduced, price increases are limited by the Consumer Price Index: single year

increases are limited to 1.5 times the forecast change in the annual index, and in periods when

inflation is greater than 10 percent, price increases are limited to five percentage points more than

the forecast change in the CPI.

The PMPRB has the power to order price reductions, if it considers that a price is excessive, and it

can also order rebates to customers, a payment to the Crown, a temporary reduction below the

assessed fair price, or a temporary price reduction of another patented drug manufactured by the

same company (Graham, 2000).

The PMPRB only has the power to control manufacturers’ “gate prices,” or the price at which

manufacturers sell to wholesalers. Wholesale price, retail margin, pharmacists’ dispensing fees, and

other distribution costs are not subject to federal regulations. Some provincial governments,

however, set policies that affect the prices paid by consumers.

Because provincial governments finance 42 percent of national prescription drug expenditures, the

provincial drug benefit plans have the ability to influence the prescription drug marketplace, but

these plans do not generally negotiate prices with suppliers. Instead, some provincial governments

set policies for the overall prices they will pay for drugs. In Quebec, the government requires

manufacturers to charge it no more than the best available price in the rest of Canada. The

government of Ontario has frozen the retail price it will pay for all drugs listed on its formulary

(Morgan et al., 2003). British Columbia uses “reference-based pricing,” in which the province’s drug

benefits program only reimburses for the cost of the drug in a therapeutic class with the lowest price

(Menon, 2001).

In addition to these direct controls on prices, the formularies of provincial prescription drug benefit

plans also play an important role in keeping drug prices low. The provincial plans only cover a set

list of drugs, and manufacturers must submit new drugs to provincial programs to be reviewed for

inclusion on the plan formulary. Drugs are reviewed for their effectiveness in relation to their costs,

and drugs that are equivalent to already listed drugs are added only if they do not increase program

costs (Menon, 2001).

The prices paid by individuals, private insurers, and health care institutions are not regulated in

Canada. However, the prices that these groups pay are influenced by the prices set by the provincial

governments. Individuals paying retail prices likely face the highest drug costs, while third-party

payers and hospitals may have the buying power to negotiate discounts, as they do in the United

States.

Methodology of Price Comparisons

While there is a general agreement that prescription drugs cost less in Canada than in the United

States, there is less agreement on the magnitude of the difference. Several important methodological

issues are raised by international price comparisons, for example, which drugs are included in the

study sample and at which point in the chain of distribution prices are measured. These issues can

affect the results of such comparisons.

Level of distribution

Comparisons of international drug prices must select a point in the chain of distribution at which to

measure the price of drugs. Previous studies have looked at the price charged by drug

manufacturers, the prices charged by drug wholesalers, and the prices charged by retail pharmacies

(Gross, 2000). While studies that compare any of these points in the chain of distribution may have

valid results, studies must use apples-to-apples comparisons, comparing the prices at the same level

of the distribution chain for all countries in the analysis (Danzon, 1999). In the following analysis,

we have separated studies by whether they work with manufacturer or wholesale prices or with retail

prices.

Data sources

Measuring prices accurately at a single of level of distribution also proves difficult. Reliance on list

prices published by the manufacturer can be particularly problematic (Wagner and McCarthy, 2004).

Wagner and McCarthy argue the upward bias in list prices is a particular issue in the U.S., where

reimbursements by the Medicaid program and other payers create incentives for high list prices with

discounts to pharmacies, so that prices for generic multi-source and brand single-source drugs are,

on average, 78 percent and 22 percent lower, respectively, than list prices. The list prices used in

some of the studies cited here include Medi-Span and Redbook prices.

More realistic prices are available from transaction data collected by organizations like IMS Health.

The IMS MIDAS database includes invoice data on transactions between manufacturers and

wholesalers, pharmacies, hospitals, and other large purchasers in over 60 countries. IMS also

collects data from pharmacy transactions in several countries, including the U.S. and Canada.

Other studies have directly contacted pharmacies to determine prices, either by phone or internet.

The prices obtained in this manner likely reflect prices charged to individuals without insurance, and

may be higher than the prices that third-party insurers pay. Studies using this method also tend to

have far smaller sample sizes than IMS databases, which collect data from thousands of pharmacies.

Drug sample

Studies comparing different countries’ pharmaceutical prices vary in which drugs they select for

analysis. Some studies may select only a few drugs. Several studies attempt to make broad

comparisons among all commonly used drugs (Gross, 2003). Many international price comparisons

look only at brand-name prescription drugs, while several other comparisons include generic drugs

as well. Danzon (1999) argues that international drug price comparisons must include a

representative random sample of drugs, and that in order to be representative, the market basket

must include generic drugs. In addition, she states that studies that consider drugs with high sales

volumes are biased if the products are chosen by those with the highest volume by dollar value, not

by number of prescriptions, as products with relatively high prices will be more likely to be selected.

Danzon and Kim (1998) argue that valid measures of average price levels can only be obtained from

comparisons based on a comprehensive or representative sample of products, appropriately

weighted, following standard index number methods. Researchers must accept trade-offs between

the ability to compare identical products and the need to examine a representative sample of a

country’s pharmaceutical market. There is wide variation between countries not only in the range of

compounds available, but also in the dosage forms, strengths, and pack sizes available. If

comparisons are made only among products that are the same molecule, manufacturers, strength,

and pack size, only a very small and unrepresentative sample will available. Danzon and Kim

maintain that requirements that products be identical should be dropped in favor of broad

representation for each country’s market basket of products, which should include generics and

over-the-counter products that are good substitutes for branded prescription drugs, with all forms,

strengths, and package sizes.

Price indices and weights

Researchers conducting international comparisons of prescription drug prices should choose

appropriate weights for each drug’s price difference in the process of calculating an average price

difference for a market basket of products (Gross, 2003). However, many published comparisons do

not weight their samples. Danzon (1999) argues that studies that use unweighted averages of price

ratios violate standard principles of price comparisons and lead to results that are extremely sensitive

to the sample used. Wagner and McCarthy (2004) support weighting by the quantity of each drug

dispensed to develop a comparative index of price levels, although the country whose quantity

weights used as the basis for comparison can influence the resulting index of relative prices.

In an analysis summarized in more detail below, Danzon and Kim (1998) report the results of three

different price indices to demonstrate the effects that weighting has on comparison results: the

Laspeyres index, which uses the base country quantity weights, the Paasche index, which uses the

comparison country quantity weights, and the Fisher index, which uses an average of the observed

quantities as weights. Danzon and Kim contend that the Laspeyres index has advantages for use in

price comparisons, because accurate data for the country selected as the base are more likely to be

available if it is the country undertaking the study. In addition, the Laspeyres index does not

assume, as the Paasche and Fisher indices do, that consumption patterns would be identical in two

countries if they both faced the same prices. This can be important when drug consumption has

low price elasticity due to the role of medical norms.

Whose price

The results of international pharmaceutical price comparisons will also differ with regard to which

payer’s price is being measured. While in Canada there is little variation in the prices paid by

different payers, in the United States the cost of drugs can vary widely according to who is

purchasing drugs: individuals paying cash pay the highest price, while insurers, managed care plans,

and government programs can negotiate discounts and manufacturer rebates by buying drugs in

large quantities (CBO, 2004; Hollis, 2004; Gross, 2003; Morgan et al., 2003). A 2000 study by the

Department of Health and Human Services found that U.S. senior citizens paying for their own

prescription drugs pay about 15 percent more for those drugs (excluding the effect of rebates) than

individuals with health insurance. In addition, this 2000 study and a 1999 analysis by the House of

Representatives Committee on Government Reform found that favored customers, such as health

maintenance organizations, and the federal government obtain even lower prices than individuals

with insurance.

Danzon (1999) argues that these differences in prices are not a matter of discrimination, but an issue

of different market segments. She maintains that the difference in prices does not imply cost

shifting, but simply that a manufacturer serving two separate markets rationally determines the price

to charge in each market independently. As with other methodological issues, it is important that

price comparisons use the prices of similar payers in both countries under consideration.

Currency conversion

In addition to methodological issues in selecting a sample and choosing which prices to use,

international pharmaceutical price comparisons must choose an exchange rate to use in analyses.

The exchange rate should not be sensitive to day-to-day currency fluctuations, but should be able to

capture the cost to citizens in one country of buying drugs in another country (Gross, 2003).

Danzon and Furukawa (2003) found that currency conversion can contribute greatly to measured

price differences: converting currencies using purchasing power parities (PPPs) based on an

economy-wide market basket of goods (GDP) or on a basket of medical goods or services, rather

than on the current exchange rate, can drastically change the results of a price comparison.

Price Comparisons

As a result of these methodological challenges, the results of previous studies that compare the

prices of prescription drugs in Canada and the United States vary. Figure B-1 summarizes

comparisons of manufacturer or wholesale prices; Figure B-2 summarizes comparisons of retail

prices. More details about the analyses follow.

Figure B-1. Summary of U.S./Canada Prescription Drug Price Comparisons: Studies of Manufacturer or Wholesale Prices

Study Data source, year Drug sample Matching Key findings

HHS Task

Force on

Drug

Importation,

2004

IMS MIDAS database

for 10 countries, 2003

54 top-selling brands and

29 top-selling generics

Excluded drugs not available in all 10

countries and drugs that could not

find a match in the National Sales

Perspective database.

Brand prices are typically 60% lower in other

countries; generic prices are lower in the U.S.

Prices were adjusted to reflect rebates at the

wholesale level (rebated prices on generic drugs

were 24 percent lower than IMS prices).

Hollis, 2004 Ontario Drug Benefit

Plan (Canada) and

Federal Supply

Schedule (U.S.), 2003

16 of the top 50

prescribed brand drugs

measured by expenditures

Not defined by author The prices paid by the United States

government for drugs are 2.5% higher than the

prices paid by the Ontario provincial

government.

Anderson et

al., 2004

IMS Health MIDAS

database, 2003

30 drugs (brand and

generic) with highest total

spending in the US that

were available in study

countries

Matched by manufacturer,

compound, and form. Included all

available dosage strengths. Excluded

drugs not available in Canada, France,

or the UK.

Canadian prices were 52% lower than prices in

the United States.

Danzon and

Furukawa,

2003

IMS Health MIDAS

database, 1992 and

1999

Brand, generic, and OTC Matched by molecule, and including

all products with that active

ingredient and all doses

Canada’s prices were 33% lower than prices in

the U.S. net of discounts, and 40% lower

ignoring discounts.

Graham and

Robson, 2000

Red Book (U.S.) and

Ontario Drug Benefit

Program, 1998

45 brand and generic

drugs

Identical in chemical name, dosage,

strength, and form

Wholesale prices are 42% lower in Canada than

in the U.S. (72% higher in the United States

than in Canada). Two generic drugs were

cheaper in the U.S.

Danzon and

Chao, 2000

IMS, 1992 171 brand and generic

drugs available in all seven

countries; up to 438 drugs

for bilateral comparisons

Matched by active ingredient and

therapeutic category.

Canada and Germany have prices 2% and 25%

higher than the U.S., respectively, while prices

for drugs are lower in Japan (12%), Italy (13%),

the U.K. (17%), and France (67%).

Danzon and

Kim, 1998

IMS Health MIDAS

database, 1992

Brand, generic, and OTC Matched first by chemical

composition and brand name or

manufacturer. Second analysis

matching by active ingredient and

therapeutic class (MOL/ATC).

Canadian prices were 6.2% higher than U.S.

prices when matched by brand name or

manufacturer, but with the larger sample

obtained by matching by MOL/ATC, U.S.

prices were 16.6% higher than in Canada.

GAO, 1992 Medi-Span Master

Drug Database-Select,

1991

121 of the 200 brand and

genericdrugs most

commonly dispensed in

U.S.

Matched by brand name,

manufacturer, dosage strength, dosage

form, and whether available only

prescription in both the United States

and Canada.

The entire sample of drugs would cost 32

percent more in the United States than in

Canada. Drug-by-drug, the median price

differential per package between the United

States and Canada was 43 percent.

Figure B-2. Summary of U.S./Canada Prescription Drug Price Comparisons: Studies of Retail Prices

Author(s) Data source, year Drug sample Matching Key findings

Quon et al.,

2005

Prices obtained by

authors from three

U.S. pharmacies 12

Canadian internet

pharmacies, 2004

Brand

44 of the 50 most

common drugs by sales

volume

Matched by brand name The authors found that Americans could save

about 24% by purchasing drugs from Canadian

internet pharmacies. Only three drugs were

cheaper in the United States than in Canada.

Minority

Staff,

Committee

Government

Reform, 2005

Prices from the top 10

Medicare drug plans

(U.S.), 2005

Canadian price source

not defined by author,

2004

Brand

10 drugs with highest sales

to United States Medicare

beneficiaries

Not defined by author The prices offered by the Medicare drug plans

on average were 61% higher than the Canadian

prices, and one plan’s price was 73% higher

than the Canadian prices. Prices for individual

drugs in the United States approached or even

exceeded 100% more than the Canadian prices

in some cases.

Task Force

on Drug

Importation,

2004

Prices from

PharmacyChecker.com

(30 Canadian and 6

U.S. online

pharmacies), 2004

Brand and generic

22 top-selling brands and

5 top-selling generics

Matched by ingredient, dosage form,

strength, and package size. Excluded

drugs not available in Canada or

available OTC in Canada.

Branded drugs are about 37% less at Canadian

internet pharmacies, while generics are about

32% less at U.S. internet pharmacies, net of

shipping costs.

Graham and

Tabler, 2001

50 pharmacies

randomly selected in

six areas (3 in Canada,

3 in the United States),

2001

Brand

3 drugs that many patients

use for a long period of

time

Not defined by author For all three drugs, the prices at Canadian

pharmacies were significantly lower than the

prices at pharmacies in the United States, but

there was also substantial variation in prices

across areas within the same country and even

within an individual region.

Graham and

Robson, 2000

Costco pharmacies in

both countries, 2000

Brand and generic

45 drugs

Identical in chemical name, dosage,

strength, and form

Retail prices are 28% lower in Canada than in

the United States (39% higher in the United

States than in Canada).

Minority

Staff,

Committee

Government

Reform, 1998

Prices obtained by

authors phone calls to

pharmacies, 1998

Brand

10 drugs with the highest

dollar sales to the elderly

in the United States

Matched by dosage, form, and

package size used in GAO report, or

most common dosage, form, and

package size according to the Drug

Topics Red Book, for drugs not

included in GAO report

On average, retail prices in United States were

72 percent more than the average prices in

Canada.

Studies of Wholesale or Manufacturer Prices

In 1992, the GAO examined manufacturers’ prices to wholesalers for a market basket of drugs most

frequently dispensed by U.S. drug stores.

Drug selection and matching: Using a list of the 200 drugs most frequently dispensed in U.S. drug

stores, the authors selected the most commonly used dosage form, dosage strength, and

package size for the United States, and were able to match 121 by brand name,

manufacturer, dosage strength, dosage form, and whether available only by prescription in

both the United States and Canada.

Data source: U.S. prices were obtained from the Wholesale Acquisition Cost in the Medi-Span

Master Drug Data Base-Select. Canadian prices were obtained from the Best Available Price

listed in the Ontario Drug Benefit (ODB) Formulary when available, or directly from the

manufacturers of the drugs or from a major Canadian wholesaler for the 20 studied drugs

that were not on the ODB formulary. The authors compared the prices for specific package

sizes, as well as the aggregate cost of purchasing a common prescription of all 121 products

in both countries.

Currency conversion: Canadian prices to U.S. dollars using the May 1, 1991, exchange rate.

Results: For the drugs studied, drug manufacturers charged wholesalers significantly more in

the United States than the same manufacturers did in Canada. The entire market basket of

drugs would cost 32 percent more in the United States than in Canada. When the price

differentials were computed drug-by-drug, the median price differential per package between

the United States and Canada was 43 percent.

Danzon and Kim (1998) undertook an international price comparison study to demonstrate the

sensitivity of such studies to methodological decisions.

Drug selection and matching: including brand name, generic, and over-the-counter products, in

nine countries (United States, Canada, Germany, France, Italy, Japan, United Kingdom,

Switzerland, and Sweden). Products were matched first based on the international product

name (IPN, in which products must have the same chemical composition, and the same

brand name or same manufacturer), and then by active ingredient (molecule) and anatomical

therapeutic class (MOL/ATC), and found that significantly more products could be included

when matches were based on MOL/ATC. Rather than pick one dosage and pack size, the

authors converted all products to grams of active ingredient and the number of standard

units.

Data source: The authors used IMS data for sales of cardiovascular products for a one-year

period between 1991 and 1992.

Price index and weights: Price differences were determined using three different price indices

(Laspeyres, Paasche, and Fisher).

Results: Using their preferred method of matching drugs by MOL/ATC and weighting using

a Laspeyres price index, the authors found that prices in the United States were 16.6 percent

higher than prices in Canada, but if the same price index was used with drugs that matched

according to IPN, Canadian prices were 6.2 percent higher than U.S. prices.

Danzon and Chao (2000) used bilateral drug price and quantity indices for seven countries to

demonstrate differences compare manufacturer prices of drugs in the U.S. to manufacturer prices in

other nations.

Drug selection and matching: The authors matched drugs based on active ingredient and

anatomic therapeutic category. The weighted average price for the molecule incorporates all

forms and packs of that molecule.

Data source: The authors used IMS data from 1992.

Price index: The authors used the Laspeyres index, which may be most relevant to the U.S.

perspective because it uses U.S. weights, and may be interpreted as a lower-bound of what

the U.S. could save by adopting other countries’ pricing.

Results: The authors found that, while drug prices are higher in the U.S. than in most other

countries, their methods show smaller differences than previous studies. Calculations using

the Laspeyres index show that Canada and Germany have prices 2.1 percent and 24.7

percent higher than the U.S., respectively, while prices for drugs are lower in Japan (11.6%),

Italy (12.9%), the U.K. (16.6), and France (67%).

Graham and Robson (2000) compared both wholesale and retail prices for drugs in the United

States and Canada.

Drug selection and matching: The researchers started with the 60 drugs with the highest volume

of prescriptions written in the United States from January through October 1998, using

those that were identical in their chemical name, dosage and strengths, and form, and

compared brand-name drugs only to other brand-name drugs and generic drugs only to

other generic drugs. Fifteen drugs had to be excluded from each part of the study because

they were not identical in every way, or because price information was not available in each

data source, although not the same 15 drugs in each case.

Data source: For wholesale prices, Graham and Robson used the prices from the Red Book for

the United States and the Ontario Drug Benefit Program formulary for Canada.

Weights: The authors weighted by number of prescriptions.

Results: The authors found that, while drugs are typically 42% cheaper in Canada, there is a

wide range of differences between the prices of drugs in the two countries. Two generic

drugs had higher wholesale prices.

Danzon and Furukawa (2003) compared average price levels for pharmaceuticals in nine countries,

including the United States and Canada in 1992 and 1999. Using the manufacturer price levels in

order to eliminate differences attributable to wholesaler and retailer margins and taxes, and adjusting

for manufacturer discounts to managed care plans and Medicaid, the authors compared price

indexes for a sample of products from each country.

Drug selection and matching: The authors began with the 350 leading molecules in the U.S. (by

sales volume), and from these selected 249 that were available in at least four of the study

countries in 1992. For each molecule, they included all products with that active ingredient,

including both brand-name, generic, and over-the-counter drugs. At least 90 percent of the

drugs were available in all the study countries, except Japan, which only had 76 percent.

Data source: The authors used data from the IMS Health Midas data set.

Price index: The authors calculated four price indices for each country: all products for each

molecule; all products for each molecule, adjusted for discounts; drugs that match only on

indication, form, and strength; and drugs that match only on indication, form, and strength,

adjusted for discounts.

Results: Except for Japan, which had higher prices, all of the study countries had lower prices

than the United States. Canada’s prices were 33 percent lower than U.S. prices net of

discounts, and 40 percent lower ignoring discounts.

Anderson and colleagues (2004) used methods similar to those used by Danzon and Furukawa

(2003), but opted for greater standardization, rather than greater representativeness, in order to

simulate the prices that would paid by the United States’ Medicare program if usage were fixed but

prices matched those in other countries.

Drug selection and matching: Prices for all available dosage strengths in each country were

considered for a market basket of the thirty drugs, both brand and generic, with the highest

total spending in the United States that were available in the other three countries.

Data source: The authors used IMS health data to obtain the prices of drugs in Canada,

France, the United Kingdom, and the United States.

Price index: The authors calculated a Laspeyres price index, with the quantity sold in the

United States as the base, using average wholesale prices (AWP).

Results: Analysis revealed the average prices for the market basket of drugs were 52 percent

lower in Canada, 59 percent lower in France, and 47 percent lower in the United Kingdom,

than in the United States. When the authors assumed that Medicare could purchase drugs at

a 20 percent discount on AWP, the difference was reduced somewhat, with prices 40 percent

lower in Canada, 48 percent lower in France, and 34 percent lower in the United Kingdom

than in the United States.

To show that it matters who is the purchaser of drugs, Hollis (2004) compared government prices

for drugs in the United States and Canada.

Drug selection and matching: Hollis used a sample of 16 drugs without generic competition in

either country from the list of the top 50 drugs in Canada, as measured by expenditure.

Generic drugs were excluded due to the fact that competitive market pressures, including

government price controls and negotiating tactics drive the pricing of generic drugs.

Data source: The comparison used the Ontario Drug Benefit Plan price and the Federal

Supply Schedule price (used by the Veterans Administration and the Department of Defense

in the United States) for the most popular strength of each medication.

Currency conversion: U.S. prices were converted to Canadian dollars using the approximate

exchange rate at the time that Speaker of the U.S. House of Representatives Dennis Hastert

complained about Canadian price controls on prescription drugs.

Results: The analysis showed that the prices paid by the U.S. government for drugs are only

2.5 percent higher than the price paid by the Ontario plan.

The HHS Task Force on Drug Importation (2004) estimated differences between the U.S. and nine

other countries: Australia, Canada, France, Germany, Greece, Japan, Poland, Switzerland, and the

United Kingdom.

Drug selection and matching: The authors began with a list of 60 best-selling brand name drugs

and 50 best-selling generic drugs, and narrowed it down to 54 brand drugs and 29 generics

after excluding drugs not available in all 10 countries.

Data source: The comparison used invoice data from the 2003 IMS MIDAS database for 10

countries.

Weighting: The authors weighted by number of U.S. prescriptions, and also adjusted for

wholesale rebates by comparing IMS data with CMS data on Average Manufacturer Price.

Results: Foreign prices for top-selling brand name products were approximately 60 percent

of the U.S. price. Generic prices were lower in the United States than every country except

Poland. Canadian prices were just over half U.S. prices for brand name drugs, and nearly

twice as much as U.S. prices for generics.

Studies of Retail Prices

In 1998, the minority staff of the House Committee on Government Reform issued a report

comparing retail prescription drug prices in the 1

Congressional District of Maine with drug prices

in Canada and Mexico.

Drug selection and matching: The study examines the pricing of twelve drugs: ten brand name

prescription drugs with the highest dollar sales to the elderly in the United States and two

drugs that in a previous study were found to experience substantial discrimination in price

according to whether the drug was purchased by an individual or by a larger organization.

The study used the prices paid by consumers without prescription drug coverage in both

countries. The study used the drug dosage, form, and package size used by the GAO its 1992

report. For drugs that were not included in the GAO report, the study used the most

common dosage, form, and package size listed in the Drug Topics Red Book.

Data Source: Prices in the United States were obtained from a survey conducted by the

minority staff and the staff of Representative Allen of nine drug stores in Allen’s

congressional district—six independent pharmacies and three chain stores. Prices for drugs

in Canada and Mexico were obtained from a survey of four pharmacies in Canada (in three

different provinces) and three in Mexico (in Ciudad Juarez).

Currency conversion: Canadian and Mexican prices were converted to U.S. dollars using

exchange rates effective on October 5, 1998.

Results: For all the drugs in the study, U.S. prices were significantly higher than Canadian and

Mexican prices. On average, senior citizens in the United States paid 72 percent more than

the average prices in Canada and 102 percent more than the prices in Mexico.

Graham and Robson (2000) compared both wholesale and retail prices for drugs in the United

States and Canada.

Drug selection and matching: The researchers started with the 60 drugs with the highest volume

of prescriptions written in the United States from January through October 1998, using

those that were identical in their chemical name, dosage and strengths, and form, and

compared brand-name drugs only to other brand-name drugs and generic drugs only to

other generic drugs. Fifteen drugs had to be excluded from each part of the study because

they were not identical in every way, or because price information was not available in each

data source, although not the same 15 drugs in each case.

Data source: For the retail part of their study, the authors compared the prices at Costco, the

only pharmacy with locations in both countries.

Weights: The authors weighted by number of prescriptions.

Results: The authors found that, while drugs are typically 28% cheaper in Canada at the retail

level, there is a wide range of differences between the prices of drugs in the two countries.

Seven generic drugs had higher retail prices.

Graham and Tabler (2001) studied the retail prices of three patented prescription drugs (Celebrex,

Lipitor, and Paxil) in three areas in the United States and in three Canadian areas along the border

between the two countries.

Data source: The authors randomly selected 50 pharmacies in each of the six areas, and called

each of the pharmacies to obtain the price of a 30-day supply of each of the drugs, as well as

information about any dispensing fees and the availability and price of delivery of

prescriptions.

Currency conversion: The authors calculated an exchange rate based on a 12-day period.

Results: For all three drugs, the prices at Canadian pharmacies were significantly lower than

the prices at pharmacies in the United States, but there was also substantial variation in

prices across areas within the same country and even within an individual region.

The HHS Task Force on Drug Importation (2004) compared internet prices in the United States

and Canada, adjusted for shipping costs for U.S. consumers (typically $10-$15 from Canada and

zero to $2 from the U.S.).

Drug selection and matching: The study started with a list of 30 top-selling brand drugs and

seven top-selling chronic use generics. The list was narrowed to 22 brand drugs and 5

generic drugs after excluding drugs that were not available via the internet, were not available

in Canada, or were OTC in one or both countries.

Products were matched by strength and package size when possible, or packages were

adjusted by the number of tablets when computing unit costs. This led to a list of 118

matching brand products and 14 generic products.

Data source: Prices were collected from PharmacyChecker.com, a website that collects prices

from 30 Canadian and six U.S. internet pharmacies. For each drug, the authors collected the

lowest U.S. price and the lowest Canadian price.

Results: Brand drugs are about 37 percent less at Canadian internet pharmacies, while

generics are about 32% less at U.S. internet pharmacies, net of shipping costs.

Quon and colleagues (2005) compared the prices of prescription drugs from three U.S. pharmacies

and 12 Canadian Internet pharmacies that sell drugs to both Canadian and American customers.

Drug selection and matching: Only brand-name drugs were used; the authors purposely excluded

generic drugs because other studies have shown that Canadian prices for generic drugs are

similar to or higher than U.S. prices. Forty-four of the 50 most popular (by sales volume)

drugs were brand-name drugs and available in both countries and were used for the study.

Unit prices were based on a lot size of 100 units when available or on the next size smaller in

the 25 percent of cases where that lot size was not available. Dosages for each medication

were based on recommended daily dose ranges, daily frequencies of administrations, and

dose availabilities.

Data source: The authors collected their data from the pharmacies’ websites, which all listed

their prices in U.S. dollars.

Results: Calculating the cost per year for a single patient by multiplying the mean unit prices

by the number of units per year, the authors found that Americans can save about 24

percent on brand-name medications if purchased from Canadian internet pharmacies rather

than major U.S. drug chain pharmacies. Only three drugs, all for erectile dysfunction, were

more expensive in Canada.

In a 2005 report the minority staff of the U.S. House of Representatives Committee on Government

Reform studied whether the new Medicare Part D drug benefit is effective in lowering drug prices

for seniors.

Drug selection and matching: The authors examined the prices of ten drugs with the highest sales

to Medicare beneficiaries in 2004, comparing the prices negotiated by the ten leading

Medicare drug plans to the prices that consumers in Canada pay for the same drugs.

Data source: The authors used the Medicare website to obtain the prices the ten leading

Medicare drug plans negotiated with manufacturers and pharmacies, prices available for the

drugs at drugstore.com at the time of the study, prices available for the drugs at Costco

drugstores, and prices paid by consumers in Canada. (The source of Canadian prices was not

defined).

Results: The analysis revealed that the prices offered by the drug plans were substantially

higher than the prices paid by consumers in Canada: on average, the prices offered by the

Medicare drug plans were 61 percent higher than the Canadian prices, with one plan’s prices

coming in at 73 percent higher than the Canadian prices, and prices for individual drugs

approaching or even exceeding 100 percent more than the Canadian prices.

Discussion

While most studies have found that Canadian drug prices are lower than prices in the United States,

estimates of the magnitude of the difference vary widely, as results are very sensitive to

methodological choices. It is not possible to define one perfect methodology for these comparisons,

as many decisions with regard to methodology involve tradeoffs. However, it is clear that certain

methodological choices will strengthen the results of an analysis. For example, while simple

comparisons that select a few popular drugs that are available in the same form in both countries

provide interesting results and may reveal whether an individual in the United States could save

money by purchasing a particular drug in Canada, more complex analyses of larger market baskets of

drugs available in each country may offer a better understanding of systematic differences in the

markets. In such a larger study, careful methodological decisions can help ensure an “apples-to-

apples” comparison and increase confidence in the results.

Appendix C. Works Cited

Anderson, Gerard F., Dennis G. Shea, Peter S. Hussey, Salomeh Keyhani, and Laurie Zephyrin.

2004. "Doughnut holes and price controls." Health Affairs Web Exclusives.

Canada Patented Medicine Prices Review Board (PMPRB). “Compendium of Guidelines, Policies,

and Procedures.” http://www.pmprb-cepmb.gc.ca/english. Updated June 9, 2006.

Danzon, Patricia M. 1999. "Price comparisons for pharmaceuticals: a review of U.S. and cross-

national studies." American Enterprise Institute, Washington, DC.

Danzon, Patricia M. 2000. "Making sense of drug prices." Regulation 23:56-63.

Danzon, Patricia M. and Li-Wei Chao. 2000. “Cross-national price differences for pharmaceuticals:

how large, and why?” Journal of Health Economics 19:159-195.

Danzon, Patricia M. and Michael F. Furukawa. 2003. "Prices and availability of pharmaceuticals:

Evidence from nine countries." Health Affairs Web Exclusives.

Danzon, Patricia M. and Jeong D. Kim. 1998. "International price comparisons for pharmaceuticals:

Measurement and policy issues." Pharmacoeconomics 14:115-128.

Graham, John R. 2000. "Prescription drug prices in Canda and the United States - Part 2: Why the

difference?" The Fraser Institute, Vancouver, BC.

Graham, John R. and Beverley A. Robson. 2000. "Prescription drug prices in Canada and the United

States - Part 1: A comparative survey." The Fraser Institute, Vancouver, BC.

Graham, John R. and Tanya Tabler. 2001. "Prescription drug prices in Canada and the United States

- Part 3: Retail price distribution." The Fraser Institute, Vancouver, BC.

Gross, David, Leigh Gross, Stephen W. Schondelmeyer, and Susan Raetzman. 2006. “Trends in

Manufacturer Prices of Prescription Drugs Used by Older Americans.” AARP Public Policy

Institute, Washington, DC.

Gross, David. 2003. "Prescription drug prices in Canada." AARP Public Policy Institute,

Washington, DC.

Hollis, Aidan. 2004. "How cheap are Canada's drugs really?" Journal of Pharmacy and Pharmaceutical

Science 7.

IMS. 2003a. “Leading 20 Products by Total U.S. Dispensed Prescriptions, 2003.” Accessed at

http://www.imshealth.com/ims/portal/front/indexC/0,2773,6599_5264_0,00.html.

IMS. 2003b. “Leading 20 Products by Total U.S. Sales, 2003.” Accessed at

http://www.imshealth.com/ims/portal/front/indexC/0,2773,6599_5264_0,00.html

IMS. 2005. “Leading 20 Products by Total U.S. Dispensed Prescriptions, 2005.” Accessed at

http://www.imshealth.com/ims/portal/front/indexC/0,2773,6599_5264_0,00.html

Menon, Devidas. 2001. "Pharmaceutical cost control in Canada: Does it work?" Health Affairs 20.

Minority Staff, U.S. House of Representatives Committee on Government Reform and Oversight.

1998. "Prescription drug pricing in the 1st Congressional District in Maine: an international

price comparison." U.S. House of Representatives, Washington, DC.

Minority Staff, U.S. House of Representatives Committee on Government Reform and Oversight.

2005. "New Medicare drug plans fail to provide meaningful drug price discounts." U.S.

House of Representatives, Washington, DC.

Morgan, Steven G., Morris L. Barer, and Jonathan D. Agnew. 2003. "Whither seniors' pharmacare:

Lessons from (and for) Canada." Health Affairs 23:49-59.

Quon, Bradley S., Rafael Firszt, and Mark J. Eisenberg. 2005. "A comparison of brand-name drug

prcies between Canadian-based internet pharmacies and major U.S. drug chain pharmacies."

Annals of Internal Medicine 143:397-403.

U.S. Congressional Budget Office. 1996. How the Medicaid Rebate on Prescription Drugs Affects Pricing in

the Pharmaceutical Industry. U.S. Congressional Budget Office, Washington, D.C.

U.S. Department of Health and Human Services. 2000. “Report to the President: Prescription

Drug Coverage, Spending, Utilization, and Prices.” U.S. Department of Health and Human

Services, Washington, DC.

U.S. Department of Health and Human Services Task Force on Drug Importation. 2004. “Report

on Prescription Drug Importation.” U.S. Department of Health and Human Services,

Washington, DC.

U.S. General Accounting Office. 1992. "Prescription drugs: Companies typically charge more in the

United States than in Canada." U.S. General Accounting Office, Washington, DC.

Wagner, Judith L. and Elizabeth McCarthy. 2004. “International differences in drug prices.” Annual

Review of Public Health 25:475-495.