Pricing methods in outcome-based contracting: δ3: reference-based pricing

Abstract

Aims

Six Delta is a six-dimensional independent platform for outcome-based pricing/contracting. The third dimension (δ3) estimates prices on the basis of international drug price referencing methods. We describe this dimension’s methodology and present a proof-of-concept application to the treatment of non-small cell lung cancer (NSCLC) with EGFR mutation with osimertinib.

Materials and methods

The reference-based pricing dimension utilizes a six-step method: (1) selecting foreign countries based on a set of four criteria (drug is available in the foreign country, price information is available in the foreign country, foreign countries are members within the organization for Economic Co-operation and Development, pricing methods in the foreign countries involve value assessment); (2) adjusting for exchange rates; (3) generating reference price (RP) scenarios; (4) adjusting with the medical inflation rate; (5) pooling all generated RP scenarios and calculating average and standard deviation (SD); (6) and Monte Carlo Simulation (MCS) to estimate the dimension-specific DSP_Reference. A proof-of-concept exercise with osimertinib in NSCLC was performed for two hypothetical outcome-based contracts: 1-year (2019–2020) and 2-year (2019–2021).

Results

The United Kingdom and Canada met the four criteria. For the osimertinib 1-year contract price, the average of eight RP scenarios, adjusted for inflation by 0.44%, was $8,892 (SD = $2,606) for a 30-day prescription. MCS yielded a DSP_Reference estimate of $9,395 or −35.72% of the wholesale acquisition cost (WAC) of $14,616. For the 2-year contract, the average, adjusted for inflation by 0.72%, was $8,928 (SD = $2,610). MCS yielded a DSP_Reference estimate of $9,442 or −35.40% of the WAC of $14,616.

Conclusions

We demonstrated that international price referencing methods can be integrated into our proposed Six Delta platform for outcome-based pricing/contracting.

Keywords:

• Reference pricing
• outcome-based-contracting
• pricing methods
• osimertinib
• lung cancer

JEL CLASSIFICATION CODES:

• D40
• C00

View correction statement:

The six Delta platform for outcome-based contracting for pharmaceuticals

This article is part of a series including:

Pricing methods in outcome-based contracting: δ1: cost effectiveness analysis and cost-utility analysis-based pricing

Introduction

In previous papers on our platform for outcome-based pricing and contracting, we presented methods to estimate dimension-specific prices (DSPs) for two dimensions: cost-effectiveness analysis and cost-utility analysis-based pricing (δ)¹; and willingness-to-pay-based pricing (δ2)². In this paper, we describe the third of dimension six dimensions, which is based on international reference pricing and estimate the dimension specific DSP_Reference. While reference-based pricing may not be directly related to clinical outcomes, strategically, this dimension is needed to examine how drug prices in the US (home country) are aligned with prices in foreign countries.

Reference-based pricing is defined as the practice of using the price of a pharmaceutical product in one or several foreign countries to drive a benchmark or reference price (RP) for the purposes of setting or negotiating the price of the product in a home country³. Reference-based pricing practice has been endorsed by the World Health Organization (WHO) in its guidelines for pharmaceutical pricing³. Although, the WHO guidelines applied different approaches for reference-based pricing, all the methods suggested a substantial price reduction in drugs³. Reference-based pricing method has been used in countries where price negotiation is intense, such as in European countries and some developing countries with single or mixed-payer systems³.

Drug prices are higher in the US than in other developed countries [Medicare proposal]. A study using data from the Organization for Economic Cooperation and Development (OECD)⁴ showed that prescription drugs were between 5% to 198% higher in the US. This is attributed, at least in part, to the higher research and development costs in the US⁵.

There are governmental efforts in the US to align drug prices with international prices, however, there are growing concerns that reference-based pricing may phase down the era of outcome-based pricing in the US⁶. In this paper, we argue instead that reference-based pricing is an important pricing dimension to inform outcome-based pricing and contracting. We describe here a method for reference-based pricing as integrated into our six-dimensional platform. In this, we assume that 1) drug payment is made by the payer directly to the manufacturer, to reduce part of the incremental cost of payment management through wholesalers; and 2) drug price is aligned with internationally and clinically valued prices. We provide a proof-of-concept exercise on osimertinib, a third-generation EGFR tyrosine kinase inhibitor (TKI), used in the treatment of NSCLC that includes two outcome-based contracts between a payer and manufacturer: a 1-year (2019–2020) and a 2-year (2019–2021).

Model

Overview

As shown in Figure 1, we aim to set different RP scenarios that can be used to estimate the DSP_Reference. Using a 6-step approach: (1) a set of four criteria for selecting foreign countries for referencing was specified; (2) for the selected countries, the exchange rate was applied to convert the price of the same drug in the foreign currency to the currency of the home country; (3) RP scenarios were generated by referencing the following three economic indicators between the US (home country) and other foreign countries: purchasing power parity (PPP), real gross domestic product (GDP) per capita, and percentage (%) of GDP spent on pharmaceuticals. A additional RP scenarios were generated on the basis of two methods, the minimum foreign price and the average foreign prices, commonly used by countries that use reference-based pricing³; (4) adjusting for inflation rates for all generated RP scenarios for a given contract duration (5) pooling all generated RP scenarios and calculating the average and standard deviation (SD); (6) and performing MCS as a confirmatory analysis to estimate DSP_Reference.

Figure 1. The 6-step method of δ3. Abbreviations. RP, reference price; PPP, purchasing power parity; GDP, gross domestic product; OECD, Organization for Economic Co-operation and Development; MCS, Monte Carlo Simulation; DSP_Reference, Dimension-specific price for δ3.

Reference-Based pricing (Figure 1)

Step 1. Selecting foreign countries based on a set of four criteria

The WHO Evidence Summary 4³ is based on a systematic review of the methods of reference-based pricing that focuses primarily on OECD countries and minimally on low-and middle-income countries³. Most countries that use reference-based pricing combine between two and five criteria to set drug prices³. Drug availability in foreign countries is the most commonly used criterion³. Other criteria include the availability of information on foreign price, time of marketing the drug of interest, and geographic and economic similarity between the home country and the foreign countries selected³.

Our proposed set of four criteria is shown in Table 1. The first criterion specifies that foreign countries should have the drug of interest on the market. This is to ensure that drug prices used for referencing are transparent and not estimated. The second criterion requires that price information is available in the foreign countries so as to ensure the accuracy of the RP in the home country. Third criterion is that the selected foreign countries should be members of the OECD so that up-to-date detailed economic indicators can be easily retrieved for referencing purposes. The fourth criterion specifies that the selected foreign countries should have value assessments of drugs after drug approval. This criterion is essential to assuming a robust reference model for drugs based on clinical value so that it can be used for outcome-based contracting.

Table 1. The set of four criteria for selecting countries.

Criteria	Rational
1. Drug is available in the foreign country	Drug prices can be transparent and not predicted
2. Price information is available in the foreign country	Referencing in home country is accurate
3. Foreign countries are members within the OECD	Data on economic indicators are available and updated
4. Pricing methods in the foreign countries involve value assessment	Value-based pricing is ensured

Abbreviations. OECD, Organization for Economic Co-operation and Development.

Step 2. Adjusting for exchange rates

Adjustment for exchange rates is performed by capturing the average exchange rates measured over the preceding year to assure a degree of stability in the currency exchange rate between the foreign countries and the home country. In the adjustment, foreign prices for the drug of interest are converted from the foreign to the home country. Importantly, the exchange rate conversion is also needed while referencing the three economic indicators used in step 3.

Step 3. Generating RP scenarios

In this step, RP scenarios are generated on the basis of three economic indicators from the OECD database: PPP, real GDP per capita, and % GDP spent on pharmaceuticals⁷. These indicators enable comparisons of the US market to other markets that differ in market size relative to the US (and are likely to be smaller in size), because of difficulty to find foreign markets at the size of the US market.

The following equations were used to estimate RP based on the three economic indicators (1) $$RP\mathrm{ }\mathit{based}\mathrm{ }on\mathrm{ }\mathit{PPP}\mathrm{ }=\mathit{price}\mathrm{ }of\mathrm{ }\mathit{the}\mathrm{ }\mathit{drug}\mathrm{ }in\mathrm{ }\mathit{the}\mathrm{ }\mathit{foreign}\mathrm{ }\mathit{country}\mathrm{ }\mathit{converted}\mathrm{ }to\mathrm{ }US\mathrm{ }\mathit{dollars}\mathrm{ }*\mathrm{ }\mathit{PPP}\mathrm{ }in\mathrm{ }\mathit{the}\mathrm{ }US/\mathit{PPP}\mathrm{ }in\mathrm{ }\mathit{the}\mathrm{ }\mathit{foreign}\mathrm{ }\mathit{country}\mathrm{ }\mathit{relative}\mathrm{ }to\mathrm{ }\mathit{the}\mathrm{ }US$$(1) (2) $$RP\mathrm{ }\mathit{price}\mathrm{ }\mathit{based}\mathrm{ }on\mathrm{ }\mathit{real}\mathrm{ }\mathit{GDP}\mathrm{ }\mathit{per}\mathrm{ }\mathit{capita}\mathrm{ }=\mathrm{ }\mathit{price}\mathrm{ }of\mathrm{ }\mathit{the}\mathrm{ }\mathit{drug}\mathrm{ }in\mathrm{ }\mathit{foreign}\mathrm{ }\mathit{country}\mathrm{ }\mathit{converted}\mathrm{ }to\mathrm{ }US\mathrm{ }\mathit{dollars}\mathrm{ }*\mathrm{ }\mathit{real}\mathrm{ }\mathit{GDP}\mathrm{ }\mathit{per}\mathrm{ }\mathit{capita}\mathrm{ }in\mathrm{ }\mathit{the}\mathrm{ }US/\mathit{real}\mathrm{ }\mathit{GDP}\mathrm{ }\mathit{per}\mathrm{ }\mathit{capita}\mathrm{ }in\mathrm{ }\mathit{the}\mathrm{ }\mathit{foreign}\mathrm{ }\mathit{country}$$(2) (3) $$\begin{array}{l}RP\mathrm{ }\mathit{based}\mathrm{ }on\mathrm{ }\%\mathrm{ }\mathit{GDP}\mathrm{ }\mathit{spent}\mathrm{ }on\mathrm{ }\mathit{pharmaceuticals}\mathrm{ }=\\ \mathrm{ }\mathit{price}\mathrm{ }of\mathrm{ }\mathit{the}\mathrm{ }\mathit{drug}\mathrm{ }in\mathrm{ }\mathit{foreign}\mathrm{ }\mathit{country}\mathrm{ }\mathit{converted}\mathrm{ }to\mathrm{ }US\mathrm{ }\mathit{dollars}\\ \mathrm{ }*\mathrm{ }\%\mathrm{ }\mathit{GDP}\mathrm{ }\mathit{spent}\mathrm{ }on\mathrm{ }\mathit{pharmaceuticals}\mathrm{ }in\mathrm{ }\mathit{the}\mathrm{ }\mathit{foreign}\mathrm{ }\mathit{country}\\ /\%\mathrm{ }\mathit{GDP}\mathrm{ }\mathit{spent}\mathrm{ }on\mathrm{ }\mathit{pharmaceuticals}\mathrm{ }in\mathrm{ }\mathit{the}\mathrm{ }US\end{array}$$(3)

Some countries may use Evidence Summary 4 provided by the WHO³ still other indicators such as the minimum foreign price⁸ and the average of foreign prices⁸. Therefore, we propose adding both metrics to generate additional RP scenarios: (4) $$RP\mathrm{ }\mathit{based}\mathrm{ }on\mathrm{ }\mathit{the}\mathrm{ }\mathit{minimum}\mathrm{ }\mathit{foreign}\mathrm{ }\mathit{price}\mathrm{ }=\mathit{min}\mathrm{ }\left[\mathit{price}\mathrm{ }in\mathrm{ }\mathit{foreign}\mathrm{ }\mathit{country}\mathrm{ }1\mathrm{ }in\mathrm{ }US\mathrm{ }\mathit{dollars},\mathrm{ }\mathit{price}\mathrm{ }in\mathrm{ }\mathit{foreign}\mathrm{ }\mathit{country}\mathrm{ }2\mathrm{ }in\mathrm{ }US\mathrm{ }\mathit{dollars},\mathrm{ }\mathit{price}\mathrm{ }in\mathrm{ }\mathit{foreign}\mathrm{ }\mathit{country}\mathrm{ }n\mathrm{ }in\mathrm{ }US\mathrm{ }\mathit{dollars}\right]$$(4) (5) $$RP\mathrm{ }\mathit{based}\mathrm{ }on\mathrm{ }\mathit{average}\mathrm{ }\mathit{foreign}\mathrm{ }\mathit{prices}\mathrm{ }=\mathrm{ }\left(\mathit{price}\mathrm{ }in\mathrm{ }\mathit{foreign}\mathrm{ }\mathit{country}\mathrm{ }1\mathrm{ }in\mathrm{ }US\mathrm{ }\mathit{dollars}\mathrm{ }+\mathrm{ }\mathit{price}\mathrm{ }in\mathrm{ }\mathit{foreign}\mathrm{ }\mathit{country}\mathrm{ }2\mathrm{ }in\mathrm{ }US\mathrm{ }\mathit{dollars}\mathrm{ }+\mathrm{ }\mathit{price}\mathrm{ }in\mathrm{ }\mathit{foreign}\mathrm{ }\mathit{country}\mathrm{ }n\mathrm{ }in\mathrm{ }US\mathrm{ }\mathit{dollars}\right)/n$$(5) where n is the number of selected foreign countries

Further, some other countries estimate the RP on the basis of the median foreign price⁹ when the number of foreign countries selected is equal or greater than two.

Step 4. Adjusting for inflation rates

The RP scenarios resulting from Equations [1–5] are not adjusted for inflation and therefore cannot be used for referencing in future outcome-based contracts. The RP scenarios must be adjusted by a projected inflation rates that can be estimated from medical consumer price index (mCPI) for prescriptions as published by the US Bureau of Labor Statistics⁹.

Step 5. Pooling all generated RP scenarios, calculating average and standard deviation

The resulting RP scenarios for a given contract are pooled and the average and standard deviation (SD) are calculated. This is an essential operational step to enable step 6.

Step 6. MCS to estimate DSP_Reference

The resulting RP scenarios could be varied, and the average of RP scenarios might need a confirmatory analysis. Thus, MCS can be applied to refer this variance in prices to a random pattern of price variations. Thus, MCS can derive a single estimate after averaging the RP scenarios and specifying a gamma distribution for price variations. MCS is a parametric method that empirically involves large numbers of iterations (e.g. 1000 or 2000) to estimate the shape of the statistical distribution of varying drug prices¹⁰. The following equations are used to specify the parameters for gamma distribution¹⁰: (6) $$ɑ\mathrm{ }=\mathrm{ }\left({\overline{X}}^2/S{D}^2\right)$$(6) (7) $$\beta =\left(S{D}^2/\overline{X}\right)$$(7) where ɑ is alpha; β is beta; X̄ is average price; SD is standard deviation

That is, the technology of MCS is used to run iterations for a given RP scenarios, and the average of the RP scenarios generated is called DSP_Reference.

Proof-of-Concept: Application to osimertinib in NSCLC with EGFR mutation

Background

In line with two prior dimension papers^1,2 we applied our 6-step method for reference-based pricing and estimating the DSP_Reference as a proof-of-concept analysis for osimertinib in NSCLC¹¹: a 1-year (2019–2020) and a 2-year contract (2019–2021).

Model inputs

As shown in Table 2. we reviewed the literature to identify foreign countries that fit into our set of four criteria. The United Kingdom (UK) and Canada met these criteria. We used the 2018 US wholesale acquisition cost (WAC) of osimertinib for a 30-day prescription, which was $14,616¹². The corresponding price point was £5,770 in the UK and CA$8,974 in Canada^13,14. The 2018 average exchange rates coefficients were 0.78 from the USD to the GBP and 1.30 from USD to CAD conversion⁷. Economic indicators were retrieved from the OECD database⁷: the 2018 PPP relative to the US was 0.69 for the UK and 1.25 for Canada; the 2018 real GDP per capita was $62,480 for the US; $46,256 for the UK; $47,725 for Canada; and the 2018% GDP spent on pharmaceuticals was 2.04% for the US, 1.11% for the UK, and 1.86 for Canada. To assess the effects of the medical inflation rate on the DSP_Reference, we averaged the mCPI for the period 2014–2018⁹. These inflation rates were projected at 0.44% for 1-year contract (2019–2020) and 0.72% for the 2-year contract.

Table 2. Model inputs.

Parameter	Estimate	Reference
Foreign countries	UK, Canada	^12^,^13
Price of osimertinib (2018)
US (USD)	USD $14,616	^14
UK (GBP)	GBP £5,770	^12
Canada (CAD)	CAD $8,974	^13
Exchange rate (2018)
USD to GBP	0.78	^7
USD to CAD	1.30	^7
PPP (2018)
US	1 (Reference)
UK	0.69	^7
Canada	1.25	^7
Real GDP per capita (2018) US $
US	$62,480	^7
UK	$46,256	^7
Canada	$47,725	^7
Percentage (%) GDP spent on pharmaceuticals (2018)
US	2.04%	^7
UK	1.11%	^7
Canada	1.86%	^7
Inputs for the outcome-based contracts*
Medical Inflation rate projected for 2019–2020	0.44%	^9
Medical Inflation rate projected for 2019–2021	0.72%	^9

*As projected from the US Bureau of Labor Statistics by averaging the medical consumer price index (mCPI) of prescriptions estimates of 2014–2018.

Abbreviations. US, United States; UK, United Kingdom; PPP, purchasing power parity; GDP, gross domestic product.

Analysis

The analyses for the proof-of-concept exercise were performed in Microsoft Excel 365ⁱ MSO supporting visual basic coding for applications.

Results

Step 1. Selecting foreign countries based on a set of four criteria

As shown in Table 3, the UK and Canada met the four criteria. Osimertinib was available in these markets, and their prices were publicly available as shown in Table 2^14,15. Both countries are members of the OECD and the economic indicators of interest are available⁷. Value assessments for osimertinib were performed for the UK by the National Institute for Health and Care Excellence (NICE)¹⁵ and for Canada by the pan-Canadian Oncology Drug Review¹⁶.

Table 3. Selected foreign countries.

Criteria	UK	Canada
1. Drug is available in the foreign country	✓	✓
2. Price information is available in the foreign country	✓	✓
3. Foreign countries are members within the OECD	✓	✓
4. Pricing methods in the foreign countries involve value assessment	✓	✓

Abbreviations. OECD, Organization for Economic Co-operation and Development; UK, United Kingdom.

Step 2. Adjusting for the exchange rates

The 2018 price for a 30-day prescription of osimertinib in the UK, adjusted for the 2018 exchange rate, was $7,426 (Table 4). The corresponding estimate for Canada was $6,913 (Table 4).

Table 4. Adjusting for exchange rate.

Selected foreign country	Price of osimertinib in US$ adjusted for exchange rate in 2018 (30-day prescription)
UK	$7,426
Canada	$6,913

Abbreviations. UK, United Kingdom.

Step 3. Generating RP scenarios

These two estimates from step 2 were used as the foreign price inputs in Equations [1–5] to generate RP scenarios (Table 5). For the UK, the RP scenarios with in parentheses the (% adjustment to 2018 WAC US osimertinib) yielded estimates of $10,762 (−26.37%) based on PPP, 10,030 (−31.37%) based on real GDP per capita, and $13,647 (−6.62%) based on % GDP spent on pharmaceuticals. For Canada, the corresponding estimates were $5,531 (−62.16%), $9,051 (−38.07%), and $7,582 (−48.12%), respectively. A minimum foreign price was found for Canada, and the associated reduction yielded a US osimertinib price of $6,914 (−52.70%). Using the average foreign prices, the associated US osimertinib price was $7,170 (−50.95%).

Table 5. Estimated RP scenarios (unadjusted for inflation rates) for osimertinib in US$and price adjustments to WAC 2018 price.

Economic indicators	RP scenario (% adjustment)
	Relative to the UK	Relative to Canada
Based on PPP	$10,762 (−26.37%)	$5,531 (−62.16%)
Based on real GDP per capita	$10,030 (−31.37%)	$9,051 (−38.07%)
Based on percentage (%) of GDP spent on pharmaceuticals	$13,647 (−6.62%)	$7,582 (−48.12%)
Based on minimum foreign price*	$6,914 (−52.70%)
Based on average foreign prices	$7,170 (−50.95%)

*The price in Canada was the lowest price for osimertinib.

The price adjustments are without inflation rate: +0.44% for 1-year contract; +0.72% for 2-year contract.

Abbreviations. RP, reference price; WAC, wholesale acquisition cost; UK, United Kingdom; PPP, purchasing power parity; GDP, gross domestic product.

Step 4. Adjusting for inflation rates

The above RP scenarios were subsequently adjusted for inflation (Table 6) by adding 0.44% to all price scenarios for a 1-year contract (2019–2020), and 0.72% to all price scenarios for a 2-year contract (2019–2021). This resulted in the following inflation-adjusted RP scenarios for a 1-year contract: when referenced to the UK $10,827 based on PPP, $10,095 based on real GDP per capita, and $13,712 based on % GDP spent on pharmaceutical. Referenced to Canada, the corresponding estimates were $5,595, $9,115, and $7,647. In addition, the minimum foreign price estimated for Canada was $6,944 and the average (also median) foreign price was $7,201. The resulting RP scenarios for a 2-year contract were as follows: Referenced to the UK, $10,867 based on PPP, $10,136 based on real GDP per capita, and $13,753 based on % GDP spent on pharmaceuticals. Referenced to Canada, the corresponding price estimates were $5,636, $9,156, and $7,688 based on GDP %; the minimum foreign price estimated for Canada was $$6,963 and the average (also median) foreign price was $7,221.

Table 6. Estimated RP scenarios adjusted for inflation rates as needed for 1-year contract (2019–2020) and 2-year contract (2019–2021), and the DSP_Reference.

Reference-based pricing	1-year contract (2019–2020)^⸸	2-year contract (2019–2021)^ǂ
Based on PPP
Referenced to the UK	$10,827	$10,867
Referenced to Canada	$5,595	$5,636
Based on real GDP per capita
Referenced to the UK	$10,095	$10,136
Referenced to Canada	$9,115	$9,156
Based on percentage (%) of GDP spent on pharmaceuticals^‡
Referenced to the UK	$13,712	$13,753
Referenced to Canada	$7,647	$7,688
Based on minimum foreign price	$6,944	$6,963
Based on average foreign price	$7,201	$7,221
Average RP scenarios (SD)	$8,892 ($2,606)	$8,928 ($2,610)
DSPReference (% price adjustment)*	$9,395 (−35.72%)	$9,442 (−35.40%)

*Price adjustment to 2018 wholesale acquisition cost of osimertinib for 30-day prescription.

^⸸Inflation rate in 2019–2020 was projected at 0.4395833% but rounded to 0.44% for reporting.

^ǂInflation rate in 2019–2021 was projected at 0.71875% but rounded to 0.72% for reporting.

Abbreviations. RP, reference price; UK, United Kingdom; PPP, purchasing power parity; GDP, gross domestic product; SD, standard deviation; DSP_Reference, Dimension-specific price in δ3.

Step 5. Pooling all generated RP scenarios, calculating average and standard deviation

Table 6 shows that for a 1-year contract, the average of the eight estimates inflated by 0.44% was $8,892 (SD = $2,606). For a 2-year contract, the average of the eight estimates inflated by 0.72% was $8,928 (SD = $2,610).

Step 6. MCS to estimate the DSP_Reference

Table 6 shows that for a 1-year contract (2019–2020), the DSP_Reference using MCS (of 2000 iterations) was $9,395 (or −35.72% for the 2018 WAC US osimertinib price). For a 2-year contract (2019–2021), the DSP_Reference using MCS (of 2000 iterations) was $9,442 (or −35.40% for the 2018 WAC US osimertinib price).

Discussion

A study published in 2017 assessing the per capita pharmaceutical spending between 1980 and 2015⁵, comparing the US with nine other countries, found that spending on pharmaceutical drugs is higher by 30% to 190% in the US than in the nine countries of Norway, Sweden, the Netherlands, Australia, Switzerland, the UK, France, Germany, and Canada⁵. Five possible reasons for the high drug prices in the US have been identified: total population, volume of drugs consumed, drug utilization per person, type and mix of drugs consumed (e.g. generics versus brand-name drugs), and prices at which drugs are sold⁵. Another study has shown that drug prices are higher in the US than in other European countries¹⁷.

In October 2018, the Centers for Medicare and Medicaid Services (CMS) released an advance notice of proposed rulemaking (CMS-5528-ANPRM) to seek public input on potential options to reduce drug costs through the International Pricing Index Model (IPI) for Medicare part B drugs⁶. The IPI model is expected to be executed for five years starting in spring 2020 and ending in 2025⁶. While methods to develop new reference-based pricing for CMS are underway, such methods if applied in the US may phase down the outcome-based pricing initiatives⁶. This is because the methods of reference-based pricing are dependent on economic indicators and not on clinical values of treatments. In itself and separate from its important role in our six-dimensional platform, our proposed approach a multi-inputs and balanced method of reference-based pricing if CMS decides to rely mainly on a referencing approach to pharmaceutical pricing. In addition, integration into our platform signals that reference pricing is a relevant element in outcome-based contracting. It provides additional price information alongside other supportive price information retrieved from the five other outcome-based pricing methods in our platform.

Other cost inputs could be considered in estimating the DSP_Reference. The WHO guidelines for country pricing policies³ notes that some countries use the WAC as a price point for referencing, whereas other countries use the ex-manufacturer price. Our proposed reference-based pricing model can use WAC, the average manufacturer price (AMP), or average selling price (ASP), among others, as points for referencing. Note that the WAC does not include discounts or rebates, which are common in the US and are privately negotiated. The WAC is a preferred price point for referencing and other incentives can be added once an objective and transparent outcome-based price point has been agreed. In addition, incentives are country-specific: while common in the US, this may not be the case in many other countries.

Some technical comments are in order. In our proof-of-concept analysis, the 30-day prescription price of osimertinib was the lowest in Canada, followed by the UK and the US. On average, the 2018 Canadian and UK prices were 49% lower than US price. Our reference-based estimates showed that 1-year contracted pricing tended to be proportionally higher than that for 2-year contracting. This is attributable to the lower inflation rate projected during 2019–2020 for 1-year contract than during 2019–2021 for 2-year contract. Interestingly, the projected price reductions [-35.72% for 1-year contract, and −35.40% for 2-year contract] from our reference-based pricing method are close to the reduction of approximately −30% expected from the implementation of CMS IPI model⁶. Also, our suggested DSP_Reference for the two contracts, confirmed that spending on pharmaceuticals was higher by 30% to 190% in the US over spending in the UK and Canada⁵.

Our method has several strengths. First, the set of four criteria for identifying reference countries are based on guidance from the WHO³ and practices used in selected OECD countries. The set of four criteria are novel for the US. In addition, novel from a contractual perspective is the criterion that candidate foreign reference countries should have performed a formal value assessment that reconciles clinical value and monetary requirements. This process is imperative because the drug of interest in the foreign country may be assigned an accurate price associated with its clinical value. Thus, the accuracy of a value-based price is reflected in the US market when referencing occurs, thereby eliminating the concern that reference-based pricing might phase down the era of value or outcome-based pricing. Second, we aimed to provide price scenarios (i.e. variations) by using economic indicators and incorporating the most commonly used methods for price referencing—preceded by adjustment for the exchange rate.

Third, we used MCS to consider as many price scenarios as possible and suggest an DSP_Reference at random pattern of variations. This procedure has not previously been suggested in reference-based pricing models. As shown in our proof-of-concept exercise, considering the MCS, the resulting DSPs_Reference for osimertinib in the two contracts were in line with the percentages required by the CMS-IPI model to be deducted from drug prices (approximately 30%)⁶. Fourth, our method showed reproducibility and flexibility of methods in terms of integrating other variables, as additional scenarios, to provide more accurate results, if data on these variables were measured.

Our proposed model has some areas that requires further consideration. As to methodology, there is a need for empirical justification for the economic indicators used whether other indicators should be considered as well. The flexibility of our referencing method can readily accommodate additional economic or geographic indicators in which more price dispersion can be observed. We encourage contracting partners to have this discussion upfront and identify the preferred indicator (See Supplemental Material for Form of Model Inputs Selection). The proof-of-concept exercise used the WAC as the price point to be assessed. We assumed that the price points retrieved from other foreign countries for osimertinib were comparable to the WAC in the US, although these points can be subject to unique market factors that are as of yet not fully quantified. In addition, we did not apply annual growth rates for real GDP per capita, PPP, % of GDP spent on pharmaceuticals while referencing the price of osimertinib for the two contracts. This is because the heterogeneity of resources about growth projections between the US and the candidate foreign countries. However, the DSP_Reference with these growth rates could be minimally changed because the price dispersion were existing among the RP scenarios used in the MCS to generate the DSP_References.

Conclusion

The proposed reference-based pricing component of our outcome-based contracting platform is comprised of a 6-step methodology that validated in a proof-of-concept exercise that suggested a price reduction for osimertinib by approximately −35.72% for a 1-year contract (2019–2020) and −35.40% for a 2-year contract (2019–2021). The proposed DSP_Reference metric should be integrated with the DSPs from the other dimensions in our initiative.

Transparency

Declaration of funding

The work reported herein was performed without sponsorship or grant funding.

Declaration of financial/other interests

The authors have no financial relationships to declare.

JME peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Author contributions

All named authors meet the International Committee of Medical Journal Editors (ICMJE) criteria for authorship of this manuscript.

Acknowledgements

No assistance was received in the preparation of this article. Sandipan Bhattacharjee is now at the University of Texas at Austin.

Notes

i Microsoft Excel 365 is a registered trademark of Microsoft Research Lab, Redmond, Washington, DC, USA.