Pricing appraisal of anti-cancer drugs in the South East Asian, Western Pacific and East Mediterranean Region

The following criteria were adopted for the inclusion of countries:

Based on the inclusion criteria, Thailand was the only eligible country from South-East Asia, Malaysia and Taiwan were included from the Western Pacific, and Oman, Pakistan, the United Arab Emirates, Lebanon, Egypt, Saudi Arabia and Bahrain from the Eastern Mediterranean region. Countries in the Asia Pacific [Australia and New Zealand] were excluded from this study, despite being part of the Western Pacific region, as they had already been covered in previous studies [7]. The included countries were classified into Low income [LIC], low middle income [LMIC], high middle income [HMIC] and high income [HIC] countries based on their GNI per capita, using the cut off points provided by the world bank website [8]. The United Nation’s 3-letter standard abbreviations [ISO ALPHA-3 code] were used to represent country names in Tables 1, 2 and 3. [9]

The price of the cancer drugs in the ten included countries was retrieved from official pricing authorities or the respective Ministry of Health or equivalent websites [Table 1]. The authors ensured that the prices were retrieved from the most recent price lists published by the respective countries. The prices retrieved for Bahrain, Lebanon, Oman and Taiwan were published in 2016 and the prices for the remaining countries were published in 2015. However, the exact publication date of the pricing data was unclear.

The GNI per capita [USD] of nine of the ten included countries was retrieved from the World Bank website. For Taiwan, however, the data had to be retrieved from Taiwan’s national statistics bureau [10] since the country is not a member of United Nations and was not listed on the World Bank website. The respective GNI per capita and country classifications are shown in Additional file 1: Table S2.

A cross–country comparison of 31 selected samples of cancer drugs was made by Vogler et al. [7] whereby the selection of drugs was also dependent on the data availability and availability of comparable products in the market in at least ten countries. In this study, the same 31 drug presentations were chosen initially as a guide. The inclusion criterion for selecting a drug presentation was the availability its pricing data in at least four out of the ten countries.

The price data were reviewed by the six co-authors. As the unit of measurement, we selected retail price per unit [i.e. per tablet/capsule or per vial]. Retail prices were chosen in this study because they represent the patients’ actual out of pocket expenditure. Another advantage of using retail prices is that they include all the add-ons such as taxes, distribution and pharmacist fees; these add-ons sometimes lead to the final price costing more than double the actual cost of the drug [11, 12]. The other reason for the use of retail prices was that ex-factory prices are hard to measure accurately, especially in countries that have no publicly funded drug coverage. [12]

Prices were only included if they referred to the same presentation in terms of pack size, strength and pharmaceutical form. Pricing data were originally extracted and presented in national currencies but later converted to USD using purchasing power parity (PPP) to allow fair comparison between different countries. Microsoft Excel was used for PPP-related calculations. Purchasing power parity rates were retrieved from the World Bank website [8]. The purchasing power parity conversion rates used are presented in Additional file 1: Table S1. The mean unit prices were calculated for each drug presentation as well as for each country. The high/low ratio for each presentation was calculated by dividing the highest unit price of the presentation by the lowest unit price of the same presentation to analyze the inter-country variability of the price for every presentation. A high/low ratio of 10 implies that the highest unit price is ten times more expensive than the lowest unit price. All calculations were performed in Microsoft Excel. All statistical analyses were descriptive.

In order to present the price comparison more effectively, the results have been arranged into the World Bank categories and comparisons are made accordingly. Of the ten included countries, one country [Pakistan] is classified as a low-income country, one [Egypt] as middle-income, three [Thailand, Malaysia and Lebanon] as higher-middle income, and five [the United Arab Emirates, Bahrain, Saudi Arabia, Oman and Taiwan] as high-income countries.

Table 3 represents prices per package as well as the calculated unit prices in USD using PPP. Upon estimating the PPP, cabazitaxel was the most expensive drug with a unit price of $17,304·9/vial [in Oman]. Nilotinib 150 mg had the lowest unit price in USD [$19·22/tab in Egypt].

The PPP adjusted USD unit prices for each medication are represented in Fig. 1. The different unit prices are connected to illustrate variations between countries. To ensure that the variations among prices is easy to discern for both low price and high price medications, the two have been presented in two different figures. Figure 1a represents cancer drugs with unit price between $0–1000, while Fig. 1b represents cancer drugs with unit prices between $1000–20,000.

Six presentations [23·1%] had a mean unit price below $100.00 and nine drugs [34·6%] had a mean unit price between $100·00 and $500·00. Eight drugs [30·7%] had a mean unit price of over $1000.00, of which, one [cabazitaxel] had a mean unit price of over $5000·00 [$11,832·9/vial]. Overall, Taiwan had the lowest mean unit prices[$492·61] and Oman the highest [$2355·6]. So Overall, Taiwan had the lowest mean unit prices for all presentations.

The average unit prices by country income category were as follows: LMICs $814.07, HMICs $1150·63, HICs: $1148·19. Using PPP-adjusted mean unit prices, the three most expensive presentations were found to be cabazitaxel [$11,832·93], trastuzumab 440 mg [$4779·35], and panitumumab [$4146·99]. The three cheapest oncology drugs were lapatinib ditosylate [$40·08], pazopanib disodium heptahydrate [$52·20], and imatinib [$56·92].

The high/ low ratio data included in Table 3 allowed us to look at the price deviation between countries. The high/low unit price ratio was less than 3 for fourteen drugs [53·80% of the 26 total included products], between 3 and 6 for eight drugs [30·77%] and more than 6 for four drugs [15·38%]. The smallest high/low ratio was 1·07, which belonged to sunitinib malate 25 mg, and the largest high/low ratio belonged to Interferon alpha-2B [13·68] [Table 3].

Box plots have been constructed based on the unit $ price of the included drugs. The boxplot for low and high price cancer medicines have been presented separately in Fig. 2a and b to ensure that the differences are easy to discern. The box plot displays the inter-quartile range [IQR] as calculated by Microsoft Excel; the bottom and top of the box are the 25th and 75th percentiles [the 1st and the 3rd quartiles, respectively], and the band in the middle of the box is the median. The extended lines describe the bottom and top whiskers.

Drug prices varied significantly across the included countries. Figure 3 represents the frequency of unit prices, as of February 2016, ranked in quartiles [Note: quartile 1 and 4 are not inclusive of the minimum and maximum values; these values have been categorized and represented separately using unique colours of their own]. Thailand and Taiwan had prices at the lower end, prices in Lebanon mainly fell in the first quartile, while prices in the United Arab Emirates and Bahrain fell in the second and third quartiles. Prices in Oman and Saudi Arabia were at the upper end [Fig. 3]. Prices in Oman were ranked the most expensive for eight presentations (Additional file 1: Table S3).

Prices in high income countries are at the upper end, and are ranked as the maximum for 13 of the included presentations, while prices in low middle income countries were ranked the most expensive for six presentations only [Fig. 4].

Cross-country comparison of pharmaceutical prices is challenging because of the differing level of sales, frequent changes in exchange rates and the differences in the pharmaceutical presentations such as strength, pack size, dose and dosage form. Of the 57 countries in the South-East Asia, Eastern Mediterranean and Western Pacific regions (excluding Australia and New Zealand), we only managed to find reliable pricing information for ten countries only due to: [i] under-developed/incomplete/not-user-friendly websites of official pricing/heath authorities; [ii] use of languages other than English in official websites; [iii] lack of public access to official drug prices; and [iv] absence of an official institution to summarize/compare prices across Asia. However, we managed to include representative countries from different ranges of GNI per capita. The other key limitations of this study are as follows: Firstly, the prices may not reflect the true cost of medications because the retrieved data are the official prices as published by the pricing authorities without consideration of [usually confidential] discounts and rebates. Secondly, this study used retail prices, which include add-ons such as taxes and distribution fees. Understanding of the amount and sources of add-ons would identify potential targets for price reduction. Unfortunately, data on add-ons was limited and hence we were unable to estimate them. Thirdly, the use of PPP calculations for price comparison required the assumptions that the value of goods and services are homogeneous across countries and that international shipment of goods takes place instantaneously, and with no cost. Unit prices were used to compare results in this study, when interpreting the results of this study, it should be kept in mind that one unit may refer to the daily dose of a tablet, or monthly vial for injection or a weight based two weekly injections. Future studies should use the data provided in this study and perform a price comparison using monthly dose or total treatment cost, as a unit for measurement and comparison. Finally, pricing revisions are done at irregular intervals and the price lists may not be updated immediately. However, most recently available prices were used for calculations.