Acknowledgments

We thank our Assistant Secretary for Planning and Evaluation colleagues Kenneth Finegold, Bisma Sayed, Arielle Bosworth, Sonal Parasrampuria, and Nancy De Lew for their guidance and input on this updated report. We also thank Amber Jessup and Zeid El-Kilani at the U.S. Department of Health and Human Services Office of the Assistant Secretary for Planning and Evaluation for their contributions and comments on the initial report on which this update is based. Christopher Whaley, Mahlet Tebeka, Nathaniel Edenfield, and Alejandro Uriel Becerra-Ornelas were coauthors on the earlier report. We thank Christine Eibner and Erin Taylor of RAND and Patricia Danzon of the University of Pennsylvania for their suggestions and comments as peer reviewers of the initial report (Mulcahy et al., 2021b), and we thank Annetta Zhou, Christine Buttorff, and Erin Taylor of RAND for their review of new material in this report.

Study Approach

We used 2022 IQVIA MIDAS data to calculate price indexes comparing prescription drug prices in the United States with those in 33 OECD comparison countries.⁴ For our main results, we used presentation-level data (that is, data with separate records for each combination of active ingredient, formulation, and dosage strength) for all prescription drugs in the IQVIA MIDAS dataset.⁵ We then compared prices between the United States and individual OECD comparison countries bilaterally, using as many presentations overlapping between the United States and the other markets.⁶ Separately, we compared U.S. prices with those in all other countries in our data aggregated together as a summary measure. We used U.S. volume weights (that is, the share of total volume accounted for by each presentation) to calculate price indexes because of our interest in price differences from a U.S. policy perspective.

IQVIA MIDAS sales and volume estimates are projected from IQVIA’s audits of standardized list prices and manufacturer, wholesaler, and other invoices; they do not reflect net prices realized by the manufacturers. These data are designed to support country-level trend and pattern analyses, but they remain estimates. The MIDAS data used in this analysis were obtained under license from IQVIA. Our MIDAS extract was prepared on May 19, 2023. The main body of the report presents results in bar charts comparing U.S. prices with those in G7 (to focus the comparison on larger OECD economies), excluding the United States—plus Mexico due to its geographic proximity to the United States and its close economic connections with the United States under the United States–Mexico–Canada Agreement) (results for other countries are available in Appendix B). We separately report a comparison of U.S. prices with prices in the 33 comparison countries combined. Price indexes greater than 100 indicate that U.S. prices are higher than those in the comparison country; indexes less than 100 indicate that U.S. prices are lower than those in the comparison country.

Gross Price Comparison Results

In 2022, U.S. prices across all drugs were 278 percent of prices in the 33 OECD comparison countries.⁷ Prices in the United States were higher than those in each individual comparison country (see Figure S.1 for comparisons of U.S. prices with those in the G7 countries and Mexico). For example, U.S. prices were 229 percent of prices in Canada (or, alternatively, Canadian prices were 44 percent of U.S. prices). Across all 33 comparison countries, U.S. prices ranged from 172 percent of prices in Mexico to 1,028 percent of prices in Turkey (see Appendix B for comparisons with individual countries).

The gap between U.S. prices and prices in other countries was larger for brand-name originator drugs. U.S. prices were 422 percent of prices of non-U.S. countries for these drugs. However, prices for unbranded generic drugs were generally lower in the United States than in other countries. U.S. prices were 67 percent of prices of non-U.S. countries for unbranded generics. We found that U.S. prices were higher than most comparison countries when combining data for all non-originator drugs, including unbranded generics and brand-name non-originator drugs.⁸

France and Japan generally have the lowest prices among the comparison countries for all drugs and for brand-name originator, biologics, and nonbiologic drugs separately. Canada, Germany, and the United Kingdom tend to have higher prices across each subset of drugs.

Our main findings—that U.S. gross prices are higher than those in comparison countries for all drugs and for brand-name originator drugs but lower for unbranded generic drugs—held through several additional sensitivity analyses related to how price indexes were calculated.

Figure S.1

U.S. Manufacturer Gross Prescription Drug Prices as a Percentage of Prices in Selected Other Countries, All Drugs, 2022. NOTE: All countries refers to all 33 OECD comparison countries combined. Other countries’ prices are set to 100. Biologics (more...)

Addressing U.S. Gross-to-Net Discounts

One important limitation of our analysis is that we use manufacturer prices for the purposes of comparison, because net prices (that is, the prices realized by manufacturers after negotiated and other discounts are applied) are not released by the pharmaceutical companies. The magnitude of the difference between manufacturer gross and net prices is difficult to quantify across all drugs. Net prices reflect confidential rebates negotiated between manufacturers and plan sponsors (often through their pharmacy benefit managers) that vary depending on market conditions and negotiating leverage. Net prices also reflect Medicaid best price and rebate program provisions, discounts from the 340B prescription drug discount program that may or may not be applied as drugs are sold by manufacturers, and discounts from other sources. In cases in which net prices can be reliably estimated, the magnitude of gross-to-net reductions varies substantially across therapeutic classes.⁹

To assess how our results might change if net price information were available, we conducted a final set of sensitivity analyses in which we adjusted U.S. prices downward based on published estimates of the relative differences between manufacturer gross and net prices. The resulting U.S. prices remained substantially higher than prices in other countries—but with a smaller difference than in our main results. When adjusting prices for U.S. brand-name drugs dispensed through the retail channel downward by 37.2 percent,¹⁰ U.S. prices for brand-name drugs were 308 percent of prices in other countries (versus 422 percent in our main results).

Because of a lack of available estimates, we did not adjust prices in other countries downward to reflect increasingly common discounts on manufacturer prices.¹¹ U.S. prices would appear relatively higher—that is, more in line with our main results—if we were able to also adjust for rebates and other discounts applied to manufacturer prices in other countries.

Building on prior findings, this update provides further evidence that gross manufacturer drug prices are higher in the United States than in comparison countries. Although we apply a single, market-wide adjustment to approximate rebates and other discounts applied to U.S. brand-name prices, we recognize gross-to-net margins vary substantially across drugs and therapeutic classes and that our estimates of U.S. net prices therefore incorporate substantial measurement error. We also recognize that resulting price indexes will understate differences between prices in the United States and other countries because they adjust only U.S. prices downward even though rebates and similar discounts are increasingly common in other countries.

Report Updates

A prior RAND Corporation report systematically compared U.S. with other countries’ drug prices using 2018 data.²⁰ Like the earlier report, this update presents descriptive comparisons of drug prices between the United States and other high-income Organisation for Economic Co-operation and Development (OECD) countries using price indexes as a basis for comparison. Our updated analyses use data from 2022. Although the main analyses in the report focus on gross prices received by manufacturers, we estimate price comparisons with adjustments to reflect important gross-to-net discounts for brand-name drugs in the United States.

In addition to results from a similar array of subgroup analyses and robustness checks that were included in the earlier report, this update presents the following findings from new analyses, including price comparisons for subsets of drugs with the potential to lower drug prices and spending in the United States: biosimilars and their reference biologics (which are previously approved biologics that biosimilar applicants reference to support their biosimilar application). This update also introduces an analysis of trends in price comparison results from 2017 to 2022.

Measuring Prescription Drug Prices

There are many measurable prices in prescription drug markets, with some easier than others to measure and analyze across all drug products. Figure 1.1 is a stylized comparison of the relationships between the magnitudes of key prescription drug prices for retail-dispensed, brand-name drugs. The relevant prices are in two broad categories: manufacturer prices and prices paid by health plans and their PBMs.

The largest of the three key manufacturer prices is a manufacturer-determined catalog or list price called wholesale acquisition cost (WAC). Despite its name, WAC is not the actual transactional price between drug companies and purchasers of drugs, such as distributors and large health care delivery systems. The actual transactional price is the gross or invoice price reflecting the amount paid, including on-invoice discounts, to the manufacturer by distributors or other buyers.

Manufacturer net prices reflect the set of complex business arrangements that have evolved over time among plan sponsors, PBMs, and drug companies. U.S. health insurers often work with a PBM to develop prescription drug formularies and benefit designs and to negotiate rebates with drug manufacturers. PBMs negotiate lower prices from manufacturers by offering larger shares of volume—driven by such tools as tiered formularies and exclusive contracts—with preferred drugs tied to lower cost-sharing for patients and fewer authorization steps for prescribing. PBMs realize these lower negotiated prices by collecting rebates from drug manufacturers. Manufacturers’ net prices account for these off-invoice rebates paid after the fact.

Figure 1.1

General Drug Price Relationships for Brand-Name Drugs and Commercial PBMs or Payers.

Among prices paid by health plans, patients, and their PBMs, the initial outlay to pharmacies—reflecting the price of the drug product with markups through the supply chain, dispensing fees, and potentially other amounts—is the largest. These prices appear in prescription drug claims data and are often used in analyses of drug spending.

The net price ultimately paid by the health plans and PBMs differs from the initial gross payment to pharmacies on several dimensions. First, the net price reflects rebates paid by manufacturers. These rebates determine the net manufacturer price described above and the amount paid by the health plan or PBM for the drug product itself. However, the final net price paid by health plans and PBMs is greater than the manufacturer’s net revenue because of dispensing fees and other amounts paid by the health plan or PBM. Both of these net prices paid by health plans or PBMs are generally not known by policymakers and researchers because rebates from manufacturers are considered trade secrets.

Furthermore, the relationships between net prices to health plans and to PBMs are difficult to disentangle. Although PBMs are able to negotiate lower prices, the extent to which they pass the lower prices on to health plans is not well understood. All or a portion of the rebates might be passed on to consumers through lower premiums—a process that is also opaque. Even if negotiated rebates are completely passed on to health plans, fee-based amounts paid to PBMs must also be considered. If a large portion of the rebates and other discounts is passed on to payers, then PBMs could offer value by reducing prescription drug spending, even if they retain a share of the discounts. One 2017 study suggests that PBMs retain 7 percent of U.S. spending on prescription drugs from the plan sponsor and patient perspective.²¹

Net prices—both to manufacturers and to health plans and PBMs—are likely the most relevant prices to ongoing policy discussions in the United States. Estimates of the magnitude of gross-to-net discounts hinge on two factors: (1) whether the focus is on the net price to manufacturers, PBMs, or health plans (2) and how exactly the discount is measured. One 2018 study estimated rebates of 23 percent from the plan sponsor perspective.²² Another recent study suggests that U.S. manufacturer net prices—including rebates and other discounts—are 37 percent below invoice prices and 52 percent below WAC.²³ However, the same study finds net prices to payers are 41 percent higher than manufacturer net prices, reflecting markups and margins throughout the supply chain, including the margin retained by PBMs.²⁴

In other countries, the functions performed by PBMs in the United States are typically performed by government or quasi-governmental entities, often in conjunction with direct and indirect price controls on drugs. Discounts that would not be reflected in manufacturer sales and prices are increasingly important in other countries.²⁵ For example, German sickness funds (similar to U.S. insurers) receive statutory rebates that are not reflected in manufacturer sales. In the United Kingdom, off-invoice discounts paid to the National Health Service and formulary-based rebates are not reflected in manufacturer sales.

Although drug prices after rebates and all discounts are particularly relevant to U.S. policymakers and payers, data on prices measured in this way have historically not been available to researchers. Researchers increasingly use estimates of net prices from third-party vendors and from U.S. Securities and Exchange Commission filings from publicly traded drug companies to approximate U.S. manufacturer net prices.

Our analysis uses IQVIA MIDAS licensed data, which offer data on volume and spending at manufacturer list or invoice prices (depending on the country) for nearly all OECD countries and drugs.²⁶ For single-source brand-name drugs, these prices are likely between manufacturer net prices and WAC. In Chapter 2, we present results from some price comparisons, applying aggregate U.S. gross-to-net adjustment factors to approximate how U.S. net prices would compare with those in other countries. However, the resulting price comparisons understate differences between prices in the United States and other countries because they adjust only U.S. prices downward, even though rebates and similar discounts are increasingly common in other countries. We return to these gross-versus-net–price issues as a limitation of our price index approach in Chapter 4.

The Motivation for Price Indexes

Calculating the average price for drugs in one country by dividing total drug sales and volume yields a result that reflects both prices and the mix of drugs used. Prescription drug utilization can vary from country to country on many dimensions, such as the following:

• mix of active ingredients (specific molecules or combinations of molecules)
• mix of presentations (prices and quantities calculated narrowly for each combination of an active ingredient, formulation, and strength)
• mix of manufacturers for each active ingredient and presentation
• mix of brand-name originator, brand-name non-originator, unbranded generic, and over-the-counter drugs
• mix of drugs that reach patients through retail outlets, health care facilities, and other channels.

Price indexes are a tool used to compare differences in prices over time or across markets (such as countries). The rationale behind price indexes is that a comparison of prices is most meaningful when it focuses on prices alone without the risk of interference from differences in the mix of products. Price indexes accomplish this by holding a market basket of goods constant while allowing prices to vary across markets. The best-known price index—the Consumer Price Index (CPI)—is calculated by the U.S. Department of Labor’s Bureau of Labor Statistics and compares prices for a broad market basket of consumer goods over time. The approach that serves as the foundation for the CPI can be used to compare prices for market baskets of prescription drugs between countries.

The prior RAND report described the many methodological decisions needed to calculate and compare price indexes for prescription drugs—for example, whether to use a U.S., other-country, or blended market basket (or volume weights); the breadth of products included; and whether to adjust prices to account for differences in purchasing power parity.²⁷

Prior Research

In Mulcahy et al., 2021b, the authors found that 2018 U.S. prices across all drugs were 256 percent of prices in other OECD countries. U.S. prices for brand-name originator drugs were relatively higher at 344 percent of prices in other countries, while prices for unbranded generics were generally cheaper in the United States versus other countries. Findings from a broad set of subgroup analyses and robustness checks yielded results similar to those presented in this report. In general, as we discuss in Chapter 4, the gap between U.S. and other countries’ prices is slightly larger in the results presented here compared with those in Mulcahy et al., 2021b.

Several prior studies used similar data and price index approaches to systematically compare U.S. and other countries’ drug prices.²⁸ These studies consistently found higher U.S. prices for drugs, especially for brand-name drugs.²⁹ However, each of these studies used data that are now over a decade old.

A small number of more-recent studies directly compares U.S. to other countries’ drug prices—for example, Mattingly et al., 2021, and Kang et al., 2021.³⁰ These studies cover a relatively small number of drugs (20 brand and 20 generic drugs in Mattingly et al., 2021, and 79 top drugs by spending in Kang et al., 2021) and use prices from a relatively small set of countries (ten OECD countries in Mattingly et al., 2021, and Canada, Japan, and the United Kingdom in Kang et al., 2021). These studies both find that U.S. prices are higher than those in other countries. One recent study compared U.S. to other-country prices for top brand-name drugs by spending at net prices after adjusting for rebates and other discounts offered by manufacturers in the United States.³¹ This study found that U.S. net prices remained roughly twice as high as manufacturer prices in other countries overall and across therapeutic classes.

Report Overview

Chapter 2 describes the methodological decisions and steps underlying our main analysis, separate analyses for several subsets of drugs, and robustness checks in which we varied methodological decisions. We present the findings from our analysis in Chapter 3. Chapter 4 concludes with a discussion on the usefulness, limitations, and future directions for price index–based systematic international comparisons of drug prices.

Data

We used 2022 IQVIA MIDAS data for the United States and the following 33 countries:³²

• Australia
• Austria
• Belgium
• Canada
• Chile
• Colombia
• Czechia
• Estonia
• Finland
• France
• Germany
• Greece
• Hungary
• Ireland
• Italy
• Japan
• Latvia
• Lithuania
• Luxembourg
• Mexico
• Netherlands
• New Zealand
• Norway
• Poland
• Portugal
• Slovakia
• Slovenia
• South Korea
• Spain
• Sweden
• Switzerland
• Turkey
• United Kingdom

Our IQVIA MIDAS data extract listed sales for specific drugs in terms of quarterly 2022 manufacturer amounts paid in U.S. dollars, converted at quarterly exchange rates, and quarterly volume measured in standard units.³³ Each row in the IQVIA MIDAS extract is defined by a combination of country; manufacturer; sector (retail or hospital); active ingredient;³⁴ formulation and route of administration; strength; over-the-counter indicators; and indicators for whether the drug was a brand-name originator drug, a brand-name non-originator drug, or an unbranded generic drug. Each record at this level lists volume and sales for the four quarters of 2022 in the data used for our main results; analyses looking at trends over time use quarterly data from 2017 to 2022.

IQVIA MIDAS sales and volume estimates are projected from IQVIA’s audits of standardized list prices and manufacturer, wholesaler, and other invoices; they do not reflect net prices realized by the manufacturers. These data are designed to support country-level trend and pattern analyses, but they remain estimates. The MIDAS data used in this analysis were obtained under license from IQVIA. Our MIDAS extract was prepared on May 19, 2023.

We excluded data for specific products flagged in the IQVIA MIDAS data as “over-the-counter” in a given country.³⁵ Over-the-counter drugs were out of scope for our analyses and have markedly different prices across countries. We also excluded records that did not have positive volume or sales. We aggregated the remaining records that shared the same country and presentation (that is, the same active ingredient, formulation and route of administration, and dosage strength) into a single country-presentation-level record by summing sales and volume.³⁶ We aggregated by presentation rather than by active ingredient to address concerns that differences in the mix of presentations relating to active ingredients could be driving differences in price. We calculated U.S. and other countries’ unit prices by dividing presentation-level sales by presentation-level volume.

As in the prior RAND report, we excluded 32 country-presentation records with incorrect sales and/or volume data.³⁷ We also excluded 49 blood factor active ingredients from our analysis because of inconsistencies across country and channel markets in how volume was measured for these drugs in IQVIA MIDAS.³⁸

Overview of Drug Markets Using IQVIA MIDAS Data

Table 2.1 compares 2022 prescription drug market size for all OECD countries with data available for analysis in an extract from the IQVIA MIDAS dataset (run date May 19, 2023). We calculated total sales of $988.9 billion and total volume of 1,099.1 billion standard units across OECD countries in our data. The United States accounted for 62.4 percent of sales but only 23.8 percent of volume. The ratio of sales-to-volume weight is much higher in the United States than in any other country; without addressing issues around the mix of drugs, this is an initial sign that U.S. prices are much higher than those in other countries. Japan, by comparison, accounted for 6.6 percent of sales and 20.0 percent of volume.³⁹

There are also important differences in the mix of drugs among countries, such as the relative contributions of brand and generic drugs to sales and volume totals. Table 2.2 presents the share of drug sales and volume from brand-name originator drugs, brand-name non-originator drugs, and unbranded drugs for each country. Brand-name originator drugs are those marketed by the original developer of the drug (that is, the originator manufacturer). Brand-name non-originator drugs are uncommon in the United States and are typically competing “branded generic” versions of off-patent drugs marketed under brand names.⁴⁰ Most generic drugs in the United States are unbranded generics. Unbranded generics accounted for greater shares of U.S. volume in 2022 than in the earlier analysis using 2018 data (90 percent in 2022 versus 84 percent in 2018).⁴¹

Table 2.1

Prescription Drug Market Shares by Sales and Volume, 2022.

Table 2.2

Within-Country Shares of Brand-Name Originator, Brand-Name Non-Originator, and Unbranded Generic Drugs, by Percentage.

In the United States, brand-name originator drugs account for 87 percent of sales but only 7 percent of volume. Brand-name originator drugs account for a larger share of volume in countries other than the United States (29 percent) but a smaller share of sales (74 percent). The United States has the highest share of volume for unbranded generic drugs (90 percent), and these drugs account for 8 percent of sales compared with 41 percent of volume and 13 percent of sales in the OECD comparison countries. Brand-name non-originator drugs account for larger shares of both volume (31 percent) and sales (13 percent) in the other OECD countries than in the United States, where they are only 2 percent of volume and 5 percent of sales.

Main Price Index Method

Because of our interest in price differences from a U.S. policy perspective, we used U.S. volume weights (that is, the share of total U.S. volume accounted for by each presentation) to calculate price indexes. For each bilateral comparison, we calculated a U.S. volume-weighted price equal to the sum of the products of the U.S. volume weights and U.S. prices. Similarly, we calculated an other-country volume-weighted price equal to the sum of the products of the U.S. volume weights and other countries’ prices.

We used all available IQVIA MIDAS data after applying the exclusions described above (for example, for blood products) and after further excluding country presentation records with fewer than 1,000 standard units in volume or less than $1,000 in sales in both the United States and the individual comparison country.⁴² See Appendix A for more information on these exclusions.

Our reported price indexes are the ratio of the U.S. volume-weighted price to the other countries’ volume-weighted price scaled by 100. We did not adjust price indexes by per capita gross domestic product or purchasing power parity or for other differences across markets. To compare U.S. prices with those in other countries broadly, we calculated a separate “all non-U.S. countries” price index using prices calculated by aggregating sales and volume across all non-U.S. countries in our data. The price indexes were calculated as described earlier.

Additional Analyses

Results at the Active Ingredient Level

A presentation (that is, active ingredient, dosage form, and strength)-level market basket and price comparison uses only data for which there is a specific match between products sold in the United States and a comparison country. Figure 3.4 presents results from the robustness test comparing prices calculated at the broader active ingredient level instead of at the presentation level. The differences between U.S. and other countries’ prices are typically larger at the active ingredient level than at the presentation level. U.S. prices were 320 percent of prices in other countries combined when price indexes were calculated using active ingredient–level data; they were 278 percent of other countries’ prices when using presentation-level data. This suggests the United States uses a more expensive mix of specific presentations and products within an active ingredient.

Figure 3.4

U.S. Prescription Drug Prices as a Percentage of Other Countries’ Prices, Active Ingredient Level, 2022. NOTE: All Countries refers to all 33 OECD comparison countries combined. Other countries’ prices are set to 100. Only some active (more...)

Results with Net Price Adjustments

We reduced U.S. prices based on estimates of the relative difference between U.S. manufacturer gross and net prices because of rebates and other discounts.⁴⁸ We used two alternative approaches to apply U.S. manufacturer net price adjustments. First, a constant 37.7 percent reduction across all U.S. prices yielded U.S. prices that were 173 percent of prices in other countries (Figure 3.5). Second, a 37.2 percent reduction to retail-dispensed, brand-name U.S. drug prices yielded U.S. prices that were

• 213 percent of prices in other countries when comparing across all drugs
• 308 percent of prices in other countries when comparing prices just for brand-name originator drugs
• 381 percent of prices in other countries when comparing prices for only retail-dispensed, brand-name drugs (Figure 3.6).

Figure 3.5

U.S. Prescription Drug Prices as a Percentage of Other Countries’ Prices, U.S. Net Price Adjustment Applied to All Drugs, 2022. NOTE: For the purposes of this comparison, we reduced U.S. prices for all drugs by 37.7 percent. We calculated the (more...)

The relative magnitudes of these results line up with our findings without the U.S. net price adjustment: Without adjustment, U.S. prices were 278 percent of prices in other countries across all drugs and 422 percent of prices in other countries for brand-name originator drugs. As we noted, the results from these sensitivity analyses understate differences between prices in the United States and other countries because we have not applied similar adjustments to prices in other countries to reflect rebates and other discounts on manufacturer prices, such as statutory rebates to German sickness funds or “patient access scheme” payments from manufacturers to the National Health Service in the United Kingdom.⁴⁹

Figure 3.6

U.S. Prescription Drug Prices as a Percentage of Other Countries’ Prices, U.S. Net Price Adjustment Applied to Retail-Dispensed, Brand-Name Drugs, 2022. NOTE: For the purposes of this comparison, we reduced U.S. prices for retail-dispensed, brand-name (more...)

Results for Biosimilars and Reference Biologics

U.S. prices for biosimilar products were generally closer to those in other countries compared with all brand-name drugs and with reference biologics (Figure 3.7).⁵⁰ Across all countries, U.S. biosimilar and reference biologic prices were 216 and 260 percent, respectively, of prices in other countries. When comparing prices across all active ingredients facing biosimilar competition in the United States as of the end of 2022 (that is, pooling across biosimilars, reference biologics, and any other versions), U.S. prices were 312 percent of prices in comparison countries combined. Comparisons of U.S. prices for biosimilars and reference biologics to those in other individual countries varied, with U.S. prices still higher but relatively closer to those in in Canada and Mexico.⁵¹

Figure 3.7

U.S. Prescription Drug Prices as a Percentage of Other Countries’ Prices, Biosimilar and Reference Biologic Categories, 2022. NOTE: All Countries refers to all 33 OECD comparison countries combined. Other countries’ prices are set to 100. (more...)

Trends in Results over Time

Figure 3.8 illustrates changes in the ratio of U.S. to other countries’ prices from 2017 through 2022 for prices in comparison countries combined (thick black trend line) and for prices in individual G7 countries and Mexico. Although trends from 2017 through 2021 were relatively stable, U.S. prices were proportionally higher (or other countries’ prices were proportionally lower, or both) in 2022. This may be due to the introduction of new brand-name drugs in the United States slightly before launching in some other countries, thereby changing the mix of drugs contributing to some bilateral comparisons (for example, with Japan and France where the 2021 to 2022 increases were greatest). Looking at 2021 prices, U.S. prices were 240 percent of prices in other countries, compared with 278 percent in 2022.

Changing the cohorts of matching drugs contributing to the price comparison for each comparison country and year effectively undoes the constant market basket integral to price index calculations. We also calculated changes in our price comparison results over time using a fixed market basket of drugs sold in the United States and other countries in each year from 2018 through 2022 (Figure 3.9). Trends, including the relative increase in U.S. prices from 2021 to 2022, looked relatively similar when restricting to this stable market basket. Given that Figure 3.9 presents results using a time-invariant market basket of drugs, this increase appears to be at least partially driven by factors other than new drugs and warrants further investigation.

Figure 3.8

U.S. Prescription Drug Prices as a Percentage of Other Countries’ Prices, 2017–2022. NOTE: All Countries refers to all 33 OECD comparison countries combined. Other countries’ prices are set to 100. Only some presentations sold (more...)

Figure 3.9

U.S. Prescription Drug Prices as a Percentage of Other Countries’ Prices, 2018–2022, Fixed Market Basket. NOTE: All Countries refers to all 33 OECD comparison countries combined. Other countries’ prices are set to 100. The same (more...)

Summary of Results

Figure 3.10 shows ratios of U.S. to other countries’ drug prices across a variety of subgroup analyses and robustness checks, including those described above and others that were included in Mulcahy et al. (2021b) but not separately presented above. See Appendix B for detailed findings from individual analyses not presented in the main body of the report.

Figure 3.10

Summary of Results: U.S. Prices as a Percentage of Other-Country Prices, 2022. NOTE: Other-Country Prices refers to all 33 OECD comparison countries combined. Adj. = adjustment; biosim. = biosimilar; incl. = including; ingrd. = ingredient; lvl. = level; (more...)

Comparisons with Prior Work

Our findings are broadly consistent with those described in Mulcahy et al. (2021b), which presents findings aligning with those from earlier studies.⁵² This updated report found U.S. manufacturer invoice prices were overall 278 percent of prices in other countries in 2022, which suggests a slightly wider gap in prices between U.S. and other countries’ prices compared with the analogous 256 percent finding from the analysis of 2018 data in the earlier RAND report. However, we recommend caution when comparing results from the two reports. Relative to the 2018 data analyzed in the earlier report, the 2022 data used in the current analysis adds data from Colombia (which joined the OECD in April 2020) and includes additional presentations in bilateral comparisons due to increasingly standardized IQVIA MIDAS formulation and strength categories over time.

The current and prior findings from sensitivity analyses were also broadly consistent. One notable set of sensitivity analyses involve adjustments for lower manufacturer net versus manufacturer gross prices. In the absence of comprehensive information on manufacturer net prices, we used three approaches to make this adjustment, including one applying a fixed price reduction across all drugs and the other two applying a fixed price reduction to retail-dispensed brand-name drugs only. This approach differed slightly from Mulcahy et al. (2021b), in which a reduction was applied to all brand-name originator drugs, not just those dispensed via the retail channel. We made this change because drugs dispensed through other channels—for example, physician-administered drugs—appear to be less likely to have any rebates and other discounts.

We found that U.S. prices were closer to those in other countries—at 173 and 213 percent of international prices—when applying the overall and retail-dispensed, brand-name originator drug adjustments only, provided prices were then compared across all drugs. However, when adjusting retail-dispensed, brand-name originator prices downward and then comparing prices for only brand-name drugs, U.S. prices remained 308 percent of prices in other countries. The much higher relative U.S. prices when focusing just on retail-dispensed, brand-originator drugs emphasizes the importance of unbranded generic drugs in lowering overall U.S. drug spending and average prices. The differences in these results also highlight the importance of refining methodologies to estimate manufacturer net prices and the key decision of whether to apply an invoice-to-net reduction to drugs other than retail-dispensed, brand-originator drugs (which was not discussed in the earlier report).

The Role of Methodological Decisions

The magnitudes of estimated price differentials varied with our measurement approach. For our main results, we opted to use presentation-level data because these more-granular price and volume weights allow comparisons that reflect the mix of drugs in each market. The differences between U.S. prices and prices in other countries was slightly wider when we used active ingredient–level data instead. This suggests that, on average, the U.S. mix of drugs within an active ingredient tended toward higher-priced presentations. While active ingredient–level analyses can help address potential mismatches between the specific dosage forms and strengths of the selected drugs sold in the United States versus other countries, they introduce new concerns by effectively averaging prices across presentations (nested within an active ingredient) with potentially important differences in characteristics and prices.

Relatedly, our use of bilateral comparisons in which we had data for both the United States and a respective comparison country in some cases led to small shares of all presentations sold in the U.S. and comparison countries contributing to each comparison. This leads to concerns regarding the generalizability of our findings. However, as discussed earlier, the presentations and active ingredients available for comparison tended to account for larger shares of volume and sales compared with presentations and active ingredients that did not contribute to our analysis.

Excluding U.S. and other countries’ presentations for which the other country’s price was very high or low compared with the U.S. price had generally modest implications on the magnitude of results. U.S. prices relative to other countries’ prices without these outlier exclusion steps were almost always within 10 percent of those from our main results. We opted to implement these outlier exclusion steps because presentations with extreme differences between U.S. and other countries’ prices have the potential to exert significant leverage over volume-weighted price calculations. Although our main concern was that inconsistencies in how sales or volumes are measured across markets could lead to outliers, there is the possibility that some of the excluded presentations with very high U.S. or other countries’ prices could have reflected an actual price difference, and, in this case, their exclusion would bias our results. For example, high prices in non-U.S. countries could reflect high private-pay amounts outside a public health care system and price controls.

We chose to use data from all of 2022 to calculate price indexes. Other studies noted the relatively quicker approval and uptake of newer, typically more-expensive drugs and presentations in the United States compared with other countries.⁵³ Our bilateral comparisons omit new drugs and presentations released in the United States in 2022 but not yet in other countries. Access to innovative treatment likely has important benefits to patients. Our study does not address the trade-offs between higher prices and earlier access to new drugs.

We opted to use U.S. volume weights because of the U.S. policy focus of the analysis. This choice had important implications for our results. Using other countries’ volume weights yielded U.S. prices that were 458 percent of prices in other countries (compared with 278 percent when using U.S. volume weights). Prescribing patterns outside the United States might be skewed toward drugs with favorable prices in individual countries as a result of price controls and/or volume purchasing. As a result, it is unsurprising that other countries’ volume weights result in findings of relatively higher U.S. drug prices.

General Limitations

There are important limitations that apply to all of the prior studies that we describe in this report and to our own analysis. First, although the drug prices net of rebates and all discounts are particularly relevant to policymakers and other stakeholder groups, there is no comprehensive source of net prices paid for drugs—either to manufacturers or by payers—in the United States. We expect a significant difference between manufacturer prices and prices to payers net of rebates in the United States and in other countries (such as Germany and the United Kingdom) where similar rebates and discounts are increasingly common. We used manufacturer prices because they are the best available comparable data for all countries.

The price differentials between the United States and other countries presented here and in other studies could be biased upward if actual negotiated discounts are larger than the factor applied to brand-name originator drugs as an ad hoc adjustment. We applied an estimated adjustment to U.S. prices to approximate rebates and other discounts applied to manufacturer prices as one of our sensitivity analyses. However, U.S. gross-to-net reductions vary across therapeutic classes and with the extent of competition for specific drug products,⁵⁴ and we recognize that the resulting net prices will almost certainly differ from the actual U.S. net prices for individual products.

We also recognize that resulting price indexes might understate differences between prices in the United States and other countries because they adjust only U.S. prices downward even though rebates and similar discounts are increasingly common in other countries. U.S. prices would appear relatively higher—that is, more in line with our main results—if we were able to also adjust for rebates and other discounts applied to manufacturer prices in other countries.

Second, manufacturer prices (that is, the prices paid to manufacturers net of discounts at the time of purchase), such as those available from the IQVIA MIDAS data, are calculated in some countries by applying a set of assumed adjustment factors on observed local-level prices. In other words, IQVIA obtains a list price or average invoice price at the local level and then calculates manufacturer prices where necessary by applying a set of country-specific average margin factors. These average margin country-specific factors are generated and updated by local industry experts. The IQVIA MIDAS dataset is the standard for use in industry to compare manufacturer prices.

Finally, we use standard units as reported by IQVIA in the IQVIA MIDAS data as our unit of volume. Although standard units are designed to improve comparability in volume measurement between different drugs, there are remaining concerns, particularly when the number of units used in practice differs from country to country. As an example, lower-dosage presentations are more common in some other countries (for example, Japan) than in the United States, and volume could be higher. We address this, in part, by using presentation-level (rather than active ingredient–level) data. We did not have data available to further adjust the volumes reported in IQVIA MIDAS by a defined daily dose or other conversion factor to improve comparability.

Potential Further Analyses

Future analyses should address the core limitation of our work: the lack of systematic data on U.S. and other countries’ net prices from either the manufacturer or payer perspectives. Although there are promising avenues of research focusing on estimates of U.S. manufacturer net prices, other prices remain elusive. The Section 204 provisions introduced by the Consolidated Appropriations Act, 2021, will, for the first time, require plan sponsors to report net spending overall, for certain specific drugs, and by therapeutic class.⁵⁵ These data will be useful to the U.S. Department of Health and Human Services and other policymakers as an input into analyses on net prices to U.S. commercial payers, particularly when combined with other data available to the U.S. government, such as Medicare Part D rebates and Medicaid best price information.

Further research is also needed to measure net manufacturer prices in countries outside the United States. Off-invoice discounts through rebates, tendering, and other arrangements are increasingly common outside the United States. As a result, the magnitude of our price comparisons adjusting for lower U.S. net prices are likely too small because we did not have data available to apply similar adjustments to lower net prices in other countries. Although there are growing calls among OECD countries for transparency in drug prices, including net prices,⁵⁶ there have been few concrete steps toward net price publication or comparison among countries.

Several prior studies have noted that adjusting for differences in per capita income explains a portion of the difference in prescription drug prices across countries, particularly for drugs sold to middle- and low-income countries.⁵⁷ However, the differences in income are not large enough to explain the entire difference in prices between the United States and other OECD countries, and four OECD countries (Luxembourg, Ireland, Norway, and Switzerland) have both higher per capita gross domestic product and lower drug prices than the United States.⁵⁸ Other studies assess whether variations among countries in health care systems and regulatory characteristics explain variation in prices.⁵⁹ Because an improved understanding of the drivers of drug price differences among countries is an important input into U.S. policy discussions on drug prices, we recommend further study on this topic.

Biologics and Nonbiologics

Figures B.1 and B.2 illustrate compared prices for biologics and nonbiologics, respectively.⁶² U.S. prices are higher than prices in all comparison countries for both biologics and nonbiologics, at 359 percent and 226 percent, respectively, of prices in all other countries combined; these higher prices are likely driven by brand-name originator drugs.

Figure B.1U.S. Biologic Prices as a Percentage of Other Countries’ Prices, 2022

NOTE: All Countries refers to all 33 OECD comparison countries combined. Other countries’ prices are set to 100. Only some presentations sold in each country contribute to bilateral comparisons.

SOURCE: Authors’ analysis of 2022 sales and volume data from IQVIA, undated (run date May 19, 2023).

Figure B.2U.S. Nonbiologic Drug Prices as a Percentage of Other Countries’ Prices, 2022

NOTE: All Countries refers to all 33 OECD comparison countries combined. Other countries’ prices are set to 100. Only some presentations sold in each country contribute to bilateral comparisons.

SOURCE: Authors’ analysis of 2022 sales and volume data from IQVIA, undated (run date May 19, 2023).

Results from Robustness Checks

Figures B.3 and B.4 present results from a sensitivity analysis using other countries’ volume weights (Figure B.3) and blended volume weights (Figure B.4) rather than U.S. volume weights. Compared with the main results, U.S. prices are even higher than other countries’ prices in these sensitivity analyses. U.S. prices were 458 percent of prices in other countries when using other countries’ weights and 357 percent of prices in other countries when using a blended rate (the geometric mean to calculate what is called the Fisher index), compared with 278 percent in our main results. When using other countries’ weights, higher U.S. prices are expected if prescribing patterns outside the United States are skewed toward drugs with favorable prices in individual countries because of price controls and/or volume purchasing.

Figure B.3U.S. Prescription Drug Prices as a Percentage of Other Countries’ Prices, Other Countries’ Volume Weights, 2022

NOTE: All Countries refers to all 33 OECD comparison countries combined. Other countries’ prices are set to 100. Only some presentations sold in each country contribute to bilateral comparisons.

SOURCE: Authors’ analysis of 2022 sales and volume data from IQVIA, undated (run date May 19, 2023).

Figure B.4U.S. Prescription Drug Prices as a Percentage of Other Countries’ Prices, Blended Volume Weights (Fisher Index), 2022

NOTE: All Countries refers to all 33 OECD comparison countries combined. Other countries’ prices are set to 100. Only some presentations sold in each country contribute to bilateral comparisons.

SOURCE: Authors’ analysis of 2022 sales and volume data from IQVIA, undated (run date May 19, 2023).

Figure B.5 presents price comparisons using retail prices rather than manufacturer prices. U.S. prices are still notably higher than prices in other countries (271 percent, compared with 278 percent in our main results using manufacturer prices). Other countries’ prices increase when using retail prices rather than manufacturer prices in some bilateral comparisons (for example, for Germany, Italy, Mexico, and the United Kingdom), likely because of higher wholesale and retail markups in these countries.

Figure B.5U.S. Prescription Drug Retail Prices as a Percentage of Other Countries’ Prices, 2022

NOTE: All Countries refers to all 33 OECD comparison countries combined. Other countries’ prices are set to 100. Only some presentations sold in each country contribute to bilateral comparisons.

SOURCE: Authors’ analysis of 2022 sales and volume data from IQVIA, undated (run date May 19, 2023).