Available online at www.sciencedirect.com
Health Policy 85 (2008) 305–313
Prices and distribution margins of in-patent drugs in pharmacy: A comparison in seven European countries Livio Garattini ∗ , Nicola Motterlini, Dante Cornago CESAV, Centre for Health Economics, “Mario Negri” Institute for Pharmacological Research, Villa Camozzi, Via Camozzi 3, 24020 Ranica (BG), Italy
Abstract Objectives: To compare prices of in-patent active ingredients (AIs) in Europe at three levels (ex-factory prices, net distribution margins and third party payers’ prices). Methods: We compared the prices in seven EU countries (Belgium, France, Germany, Italy, the Netherlands, Spain and the UK) of the 20 in-patent AIs most sold on the Italian retail market in 2004, based on “sell in” sales data. We calculated the average ex-factory price per unit of each compound in each of the seven countries, weighted by the volumes of all reimbursable package sizes and strengths. We estimated net distribution margins according to the 2004 domestic regulations by deducting any type of mandatory discount. Finally, we added VAT to calculate “third party payer’s prices”. All prices were expressed in index numbers (Italy = 100). Results: Italy had the lowest average ex-factory prices, the Netherlands and particularly the UK had by far the lowest distribution margins, while Germany had by far the highest third party payers’ prices. The Netherlands and particularly UK showed a steep decrease from ex-factory to third party payers’ prices, while Belgium, Italy and Spain gave the opposite pattern. Conclusions: Our study suggests that public authorities can deal with drug prices both by strictly controlling ex-factory prices and by establishing appropriate distribution margins. The latter might be facilitated by liberalizing the distribution sector. © 2007 Elsevier Ireland Ltd. All rights reserved. Keywords: Distribution margins; Europe; Pharmaceuticals; Prices
1. Introduction During recent years drug expenditure has risen faster than other healthcare costs in many EU countries [1]. This increase is often induced by new high-priced drugs ∗ Corresponding author. Tel.: +39 035 4535360; fax: +39 035 4535372. E-mail address:
[email protected] (L. Garattini).
which mainly replace off-patent and thus cheaper products [2]. Rising drug costs are a challenge to healthcare policy makers for two main reasons [3]. First, there might be an opportunity cost if resources devolved to finance drug expenditure could be instead allotted to improve public health. Second, high prices put governments under budgetary pressure as they try to maintain universal access to medicines for their population at a sustainable cost—in Europe about 75% of phar-
0168-8510/$ – see front matter © 2007 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.healthpol.2007.08.005
306
L. Garattini et al. / Health Policy 85 (2008) 305–313
maceutical expenditure is covered by public funding [4]. Many EU countries have opted for tight price control of reimbursable medicines to contain expenditure. Drug prices can be controlled by health authorities at ex-factory level or at wholesale and pharmacy levels by regulating distribution margins [5], while taxation (e.g. VAT) is usually set by the tax authorities and is beyond the control of health care systems. To evaluate pharmaceutical policies on in-patent drugs in Europe, we carried out a two-step analysis in seven EU countries (Belgium, France, Italy, Germany, the Netherlands, Spain and the United Kingdom) which implied: (i) a first phase of comparative analysis of how pricing and reimbursement of new drugs is regulated (published separately [6]), and (ii) a second phase of price comparison on a basket of active ingredients in the same EU countries (shown here). Difficulties in international drug price comparisons are well known [7–9] and several methodological issues need to be addressed for this type of study [10]. Recent literature in this field is limited [3,9,11]. To the best of our knowledge our price comparison is the first conducted in EU countries at three levels (ex-factory prices, net distribution margins and third party payers’ prices), and based on sales data. To clarify and justify our calculations, we provided an overview of the regulation of distribution margins and taxation in each country too.
2. Methods 2.1. Data analysis We compared the prices of 20 in-patent active ingredients (AIs) delivered through the retail channel in seven EU countries; in-patent drugs dispensed through hospitals only, off-patent drugs, and OTCs were excluded. We selected the basket of countries according to two main criteria: France, Germany, Italy, Spain and the UK because of their importance in the EU, and Belgium and the Netherlands since they are two old EU members often referred to by other EU countries as benchmarks for pricing. Drug selection involved identifying the 20 most sold in-patent AIs on the Italian market in 2004. Since domestic patent protection of drugs is still affected by a
national extension allowing longer coverage than under EU regulations [12], we checked in advance that all 20 AIs were still under patent protection in the other six EU countries too at the time of the study. All these drugs were launched during the decade 1994–2004, except for enoxaparin which was available earlier (in 1987 in France for the first time). Our sample represented 13.3% of public pharmaceutical expenditure of all the seven EU countries as an average in 2004, ranging from 9.6% in the UK to 16.2% in the Netherlands. We conducted the analysis from the public authorities’ perspective, so we considered only the packages reimbursed by each national health system. A total of 792 packages were included. Sources of figures for sales in volumes and values to estimate prices were the IMS national health surveys. We processed the “sell in” data of IMS, which are available in all EU countries and are calculated on the orders dispatched by pharmacies to wholesalers and manufacturers (for direct sales). Volumes were expressed in standard units to make the different forms comparable (tablets, pills, capsules, etc.). All prices in the analysis were expressed in Euro, the currency adopted in six of the seven EU countries considered—English prices in Pounds were converted to Euros (D 1 = £0.678671) at the 2004 official average exchange rate [13]. We applied purchasing power parities (PPP) based on 2004 gross domestic products (GDP) to our data [14]. GDP PPPs are rates of currency conversion that equalize the purchasing power of different currencies/countries for an economy-wide basket of goods [11]. First of all, we calculated the average ex-factory price per unit of each compound in each of the seven countries, weighted by the volumes of all reimbursable package sizes and strengths. The following formula summarizes the computation: Pa =
N Xb b=1
Yb
Yb
× N
b=1 Yb
N
= b=1 N
Xb
b=1 Yb
P = ex-factory price per unit (Euros), a = 1, . . ., 20 active ingredients, X = sales in ex-factory prices (Euros), Y = sales in units standardized by dosage, b = 1, . . ., N type of package for each active ingredient. Since distribution margins are calculated as percentage “mark-ups” on packages in most of the seven countries, we identified a “standard package” for each AI for our analysis at different price levels. We assumed
Table 1 Unitsa , strengths and unit prices of the most sold packages of the 20 drugs included in our sample, by country (2004) Drugs
Belgium
1 20 50 16 200 40 20 80 300 0.125 10 50 10 5 5 20 10 80 160 75
Germany
Italy
The Netherlands
Price Unit mg
Price Unit mg
Price Unit mg
3.79 1.16 4.74 0.63 0.92 3.53 0.94 0.74 0.88 0.31 0.47 0.61 1.10 0.44 1.88 0.67 1.25 0.79 0.79 0.88
4.34 0.64 3.81 0.57 0.76 5.45 1.00 0.82 0.59 0.25 0.42 0.62 1.10 0.34 3.37 0.57 1.11 0.68 0.60 0.33
4.34 1.11 5.08 0.75 0.88 6.34 0.97 0.73 0.65 0.31 0.31 0.65 1.27 0.46 4.83 1.44 1.92 0.79 0.65 1.15
28 28 30 28 30 6 28 28 28 50 28 28 28 28 28 28 28 28 28 30
1 10 50 8 200 40 20 80 150 0.125 10 50 10 5 10 20 10 80 80 37.5
100 100 90 98 100 10 60 100 98 150 100 98 100 100 56 60 56 98 98 100
1 20 50 16 200 40 20 40 150 0.125 10 50 10 5 10 40 20 80 80 75
28 30 28 28 20 6 14 28 28 50 28 28 28 28 28 14 14 28 28 14
Price Unit mg
1 3.53 30 20 0.53 30 150 12.85 30 16 0.63 30 200 0.75 30 40 3.89 10 40 1.23 30 80 0.67 28 150 0.53 28 0.125 0.27 50 10 0.35 28 50 0.59 30 10 1.03 30 5 0.33 28 10 3.10 28 20 0.54 30 20 1.03 28 80 0.63 28 80 0.54 28 75 0.76 30
1 10 50 4 200 40 40 40 150 0.125 10 50 10 5 5 40 20 40 80 75
Spain
United Kingdom
Price Unit mg
Price Unit mg
4.27 0.79 5.18 0.37 0.92 3.69 1.39 0.60 0.60 0.32 0.37 0.61 1.20 0.45 1.82 1.26 1.33 0.62 0.58 0.72
2.89 0.66 3.61 0.63 0.84 3.02 0.93 0.85 0.78 0.30 0.38 0.58 1.05 0.39 3.29 0.59 1.15 0.66 0.83 0.86
28 28 30 28 30 10 14 28 28 50 28 28 28 28 56 28 28 28 28 30
1 10 50 16 200 40 20 80 300 0.125 10 50 10 5 10 20 20 80 160 75
28 28 28 28 60 10 28 28 28 50 28 28 28 28 28 28 28 28 28 28
Standard package Price Unit mg
1 3.83 28 10 0.83 28 150 11.05 30 4 0.45 28 100 0.46 30 20 4.37 10 20 0.85 28 40 0.59 28 150 0.76 28 0.125 0.31 50 10 0.29 28 100 1.05 28 10 1.20 28 5 0.45 28 10 4.49 28 20 0.59 28 10 1.05 28 40 0.58 28 80 0.76 28 75 1.10 30
1 10 50 16 200 40 20 80 150 0.125 10 50 10 5 10 20 10 80 80 75
L. Garattini et al. / Health Policy 85 (2008) 305–313
Unit mg Anastrazole 28 Atorvastatin 84 Bicalutamide 28 Candersartan 98 Celecoxib 30 Enoxaparin 10 Esomeprazole 56 Fluvastatin 98 Irbersartan 98 Latanoprost 150 Lercanidipine 28 Losartan 98 Montelukast 28 Nevibolol 28 Olanzapine 28 Pantoprazole 56 Rabeprazole 28 Telmirsartan 98 Valsartan 98 Venlafaxine 28
France
mg = milligrams. a All units refer to tablets, except for enoxaparin (syringes) and latanoprost (drops).
307
308
Table 2 Regulation of wholesale and pharmacy margins, mandatory discounts and VAT in the seven EU countries (2004) Net distribution margins (A) + (B) − (C)
VAT (%) Pharmacy margin (B)
Mandatory discount (C)
PP ≤ D 24 → 15.08% MSP, 24.01 ≤ PP ≤ 38.97 → D 2.18, PP > D 38.97 → D 2.18 + 0.68% (PP − D 24)
4.5% of patients’ co-payments on reimbursed drugs
6
France
MSP ≤ D 22.9 → 10.3%, D 22.91 ≤ MSP < D 150→ +6%, MSP ≥ D 150 → +2%
NA
2.1
Germany
MSP ≤ D 3 → 15%, D 3.01 ≤ MSP ≤ D 374 → D 0.45, D 3.75 ≤ MSP ≤ D 5.00 → 12%, D 5.01 ≤ MSP ≤ D 6.66 → D 0.60, D 6.67 ≤ MSP ≤ D 9.00 → 9%, D 9.00 ≤ MSP ≤ D 11.56 → D 0.81, D 11.57 ≤ MSP ≤ D 23.00 → 7%, D 23.01 ≤ MSP ≤ D 26.82 → D 1.61, D 26.83 ≤ MSP ≤ D 1200 → 6%, MSP > D 1200 → D 72 6.65% PP
PP ≤ D 24 → 51.7% MSP, 24.01 ≤ PP ≤ 38.97 → D 7.44, 38.98 ≤ PP ≤ 66.67 → D 7.44 + 2.32% (PP − D 24), PP > D 66.67 → D 7.44 + 2.81% (PP − D 24) MSP ≤ D 22.9, D 22.91 ≤ MSP < D 150 → 26.1%, MSP ≥ D 150 → +10% + D 0.53 per pack → +2% D 8.10 per pack + 3% WSP
D 2 per pack
16
26.7% PP
PP ≤ D 25.82 → 3.75%, D 25.83 ≤ PP ≤ D 51.65 → 6%, D 51.66 ≤ PP ≤ D 103.29 → 9%, D 103.30 ≤ PP ≤ D 154.94 → 12.5%, PP ≥ D 154.95 → 19% 6.82% PP (up to a maximum of D 6.80) 4.39% PPa
10
Italy
The Netherlands Spain United Kingdom
NA MSP ≤ D 78.34 → 9.6% WSP, MSP > D 78.34 → D 8.32 12.5% PP
D 6.10 per pack MSP ≤ D 78.34 → 27.9% PP, MSP > D 78.34 → D 33.54 £0.946 per pack
11.47% PP
MSP = manufacturer’s selling price, NA = not applied, PP = public price, WSP = wholesaler’s selling price. a This percentage was calculated as the ratio of the total discount paid by Spanish pharmacies in 2004 and the total NHS expenditure in 2004 (VAT excluded).
6 4 0
L. Garattini et al. / Health Policy 85 (2008) 305–313
Wholesale margin (A) Belgium
L. Garattini et al. / Health Policy 85 (2008) 305–313
the size and strength of the standard package equal to that of the most sold package available at the same time in the majority of the seven EU markets. Table 1 shows the units, dosages and prices per standard unit of the most sold packages of each AI by country. We estimated wholesale and pharmacy margins according to the 2004 regulation for drugs reimbursed in each of the seven countries (see Section 2.2). We tried to estimate net margins by deducting any type of mandatory discount or claw-back from pharmacies to public authorities on the distribution margins set by law. Finally, we added VAT (where applied) to net distribution margins and ex-factory prices in order to calculate the virtual “third party payer’s prices” of each standard package. Ex-factory prices, net distribution margins and third party payers’ prices were expressed in index numbers for each AI. Here, we present the average index number per country for easier interpretation of results. Italy was the reference country (100), as the nation where AIs were selected. The t-test was applied to assess statistical differences in average index numbers between Italy and each of the other six countries. 2.2. Distribution margins Table 2 summarises the domestic regulations in force as of January 2004 for wholesale and pharmacy margins, mandatory discounts from pharmacies to public authorities, and VAT for reimbursed drugs. Therefore, it includes all information we used then to estimate net distribution margins and third party payers’ prices in our calculations. In Italy and Spain wholesale and pharmacy margins are a fixed proportion of respectively public prices (VAT excluded) and wholesale prices. In Belgium, France and Germany wholesale margins are calculated by applying regressive percentages to increasing exfactory prices. In Belgium and France pharmacies are remunerated by applying these percentages plus a small flat dispensing fee per pack in France only (included in the public price). In Germany pharmacy remuneration is a sum of a flat “fee for service” per package dispensed (included in the public price) and a proportion of the wholesaler’s selling price. Differently from the other countries analysed, wholesale margins are not set by law in the Netherlands. Since manufacturers pay the distributors, in practice
309
their margin is already included in the ex-factory price; according to the Royal Dutch Pharmaceutical Association of Distributors in 2004 it was a 12% discount on wholesale selling prices as an average, then equally shared by pharmacies and wholesalers. Pharmacies also receive a flat dispensing fee per package which is not included in public prices. In the UK the wholesale margin for in-patent products has been fixed at 12.5% of public prices. In practice, wholesalers pass most of their margin on to pharmacies as a discount, the amount varying from pharmacy to pharmacy depending on their bargaining power—the British Association of Pharmaceutical Wholesalers (BAPW) estimated an average discount of 9.0% in 2004. Pharmacies also receive an allowance per package from the National Health Service (NHS) which is not included in the public price. In all countries but France the authorities have introduced “claw-back” systems to reduce pharmacy margins, although the methods vary widely. In Italy the discount was a progressive percentage of public prices in 2004, while in the Netherlands it was a flat proportion up to a ceiling per package. In Germany a mandatory discount was set as a fixed amount per package sold, in Belgium it was a tax fixed as a proportion of patients’ co-payments on reimbursed drugs. In the UK and Spain the claw-back was a progressive percentage applied to the pharmacy’s monthly turnover, i.e. the higher the turnover the higher the discount. To assess the incidence of net distribution margins on different ex-factory prices, Fig. 1 shows the patterns computed on our sample of packages. Although trends are regressive to prices in all seven countries, their steepness differs widely and varies a lot according to price bands. Apart from the UK where reimbursed drugs are exempted from VAT, in the other six countries medicines dispensed through the retail channel are subject to different rates, ranging from 2.1% in France up to 16% in Germany.
3. Results Table 3 shows the ex-factory prices per unit by country. Italy had the lowest prices for nine of the 20 AIs, followed by Spain (5 AIs) and France (3 AIs). Ger-
310
L. Garattini et al. / Health Policy 85 (2008) 305–313
Fig. 1. Incidence of net distribution margins on ex-factory prices in seven EU countries, calculated on the sample of packages analysed (the lowest and highest 5% “tails” were excluded from calculations for graphical reasons).
many, the UK, and the Netherlands were the nations with the largest numbers of the highest prices (10, 7, and 5). These results were confirmed by calculating the price index by country on the whole sample of the average ex-factory prices (Fig. 2). Italy had the
lowest average value, slightly less than Spain and far lower than the Netherlands (p < 0.001), UK (p = 0.003) and Germany (p < 0.001). When GDP PPP rates were applied, the situation changed significantly only for Spain (p < 0.001).
Table 3 Ex-factory prices (D ) per standard unit of the 20 in-patent drugs in our sample, by country (2004)
L. Garattini et al. / Health Policy 85 (2008) 305–313
311
Fig. 2. Average ex-factory price index on the sample of 20 in-patent drugs (Italy = 100). B = Belgium; F = France; D = Germany; I = Italy; NL = The Netherlands; E = Spain; UK = United Kingdom; GDP = gross domestic product; PPP = purchasing power parity. *t-test statistically significant vs. Italy (p < 0.05).
Fig. 3 shows the indexes of the estimated net distribution margins. The Netherlands and particularly the UK had by far the lowest distribution margins, and Spain the highest. All these countries showed significant differences from Italy (p < 0.05); differences were even bigger once values were weighted by GDP PPPs. Fig. 4 shows the indexes of third party payers’ prices, estimated by adding net distribution margins and VAT to ex-factory prices. Germany had by far the highest, and the UK the lowest. Both countries differed significantly from Italy (p < 0.05) and this was confirmed once values were converted to GDP PPPs. Fig. 5 summarises the lowest/highest ranking of the three price levels on the basis of each country’s position. The Netherlands and particularly the UK showed a steep decrease from ex-factory to public price, while
Belgium, Italy and especially Spain gave the opposite pattern, mainly because of distribution margins.
4. Discussion This study compared three price levels in seven EU countries on a sample of 20 selected in-patent drugs. International price comparisons necessarily imply methodological choices which must be kept in mind when interpreting the results [6,9]. First of all, we limited the comparison to 20 in-patent drugs reimbursed by public authorities in seven European countries. Although the drugs selected were the most sold in one country only (Italy), they were a major sample in the other six too, their proportion on domestic public expenditure ranging from 9.6% to 16.2%. More-
Fig. 3. Average net distribution margin index on the sample of 20 in-patent drugs (Italy = 100). B = Belgium; F = France; D = Germany; I = Italy; NL = The Netherlands; E = Spain; UK = United Kingdom; GDP = gross domestic product; PPP = purchasing power parity. *t-test statistically significant vs. Italy (p < 0.05).
312
L. Garattini et al. / Health Policy 85 (2008) 305–313
Fig. 4. Average third party payers’ price index on the sample of 20 in-patent drugs (Italy = 100). B = Belgium; F = France; D = Germany; I = Italy; NL = The Netherlands; E = Spain; UK = United Kingdom; GDP = gross domestic product; PPP = purchasing power parity. *t-test statistically significant vs. Italy (p < 0.05).
over, the launch of each drug differed by not more than 3 years from one country to another, except for enoxaparin and nebivolol (6 years); thus their time lag on the market was really close. Given the wide variability among countries in the range of strengths and package sizes for each drug, strict matching criteria would have excluded most packages from analysis, making the sample very small and unrepresentative of public pharmaceutical expenditure. We therefore estimated an average price per AI of all reimbursable packages weighted by volumes sold, in order to capture their real impact on public expenditure. Our calculations might have led to very
different values from real prices. However, once we compared the ex-factory prices of the most sold packages by country with the standard packages calculated by the weighing formula, differences were very small (<±20%) in three quarters of the 140 cases and very striking (>±80%) only in six cases—the latter always due to large differences in dosages between the most sold package in the country and our standard package. We decided to estimate three levels of prices so as to analyse the different price components separately. Comparison was complicated by wide differences in building up public prices in the seven countries. We tried to take these differences into account to make
Fig. 5. Ranking of the seven EU countries by ex-factory prices, net distribution margins and third party payers’ prices.
L. Garattini et al. / Health Policy 85 (2008) 305–313
virtual net distribution margins and third party payer’s prices as closely comparable as possible for all seven countries. This implied some assumptions for each component: for instance, we had to set a single “standard package” for each AI and to include pharmacy margins in third party payers’ prices also for countries like the Netherlands and UK where they are not. Thus, depending on the regressiveness of distribution margins, we might have overestimated them in Belgium and especially Germany, the two countries where the most sold packages had a much higher number of units for most AIs than the other countries (see Table 1). The UK, Germany and (partly) the Netherlands had very high average ex-factory prices compared to Italy and France, which have strict pricing control by public authorities. The first three countries are those which traditionally apply more flexible price schemes and where industry usually launches its products first, as confirmed by our sample too (11 AIs in the UK, 5 in Germany and 3 in the Netherlands). However, the results changed considerably when assessing net distribution margins and third party payer’s prices. Belgium, Italy and Spain rose in the ranking, while the Netherlands and UK moved down dramatically; both the latter are very advanced in the widespread process of liberalizing the distribution sector, e.g. opening pharmacies is not subject to statutory restrictions and chains are allowed. It is difficult to compare studies in this field since the results are always sensitive to the authors’ different methodological choices on crucial issues [8]. Moreover, the international literature on price comparisons across different countries is not flourishing if one limits the search to Europe and innovative drugs. To the best of our knowledge, the only recent study was by Martikainen et al. [3] who compared the prices of eight new reimbursable pharmaceuticals approved by the European Commission in 2000 on a basket of nine EU member states (Belgium, Denmark, Finland, France, Ireland, the Netherlands, Spain, Sweden and the United Kingdom). This survey applied different criteria for the selection of both countries and drugs, and the three price levels analysed (wholesale price, pharmacy retail price with and without VAT) are not fully comparable with ours. This study did not consider sales data and compared prices by pack (although not all of
313
them were available at the same time in all nine countries). Taking account of these differences, Martikainen et al’s results are in fact broadly similar to ours, since they found wholesale prices were lower in countries with strict price control (e.g. France and Spain), while the ranking changed considerably for retail prices due to pharmacy margins and taxation differences in each country. In conclusion, our study suggests that public authorities can deal with drug prices both by strictly controlling ex-factory prices and by establishing appropriate distribution margins. The latter might be facilitated by liberalizing the distribution sector.
References [1] OECD Health Data 2003, 3rd ed. [2] de Joncheere K, Haaijer-Ruskamp FM, Rietveld AH, Dukes MNG. Scope of the problem. International Journal of Risk and Safety of Medicine 2002;15:7–13. [3] Martikainen J, Kivi I, Linnosmaa I. European prices of newly lanched reimbursable pharmaceuticals—a pilot study. Health Policy 2005;74(3):235–46. [4] Ess SM, Schneeweiss S, Szucs TD. European healthcare policies for controlling drug expenditure. Pharmacoeconomics 2003;21(2):89–103. [5] Rietveld AH, Haaijer-Ruskamp FM. Policy options for cost containment of pharmaceuticals. International Journal of Risk and Safety of Medicine 2002;15:29–54. [6] Garattini L, Cornago D, De Compadri P. Pricing and reimbursement of in-patent drugs in seven European countries: a comparative analysis. Health Policy 2007;82:330–9. [7] Garattini L, Salvioni F, Scopelliti D, Garattini S. A comparative analysis of the pharmaceutical market in four European countries. Pharmacoeconomics 1994;6(5):417–23. [8] Danzon PM, Kim JD. International price comparisons for pharmaceuticals. Measurement and policy issues. Pharmacoeconomics 1998;14(Suppl. 1):115–28. [9] Danzon PM, Chao L-W. Cross national price differences for pharmaceuticals, how large and why? Journal of Health Economics 2000;19:159–95. [10] Anderson F. Methodological aspects of international drug price comparison. Pharmacoeconomics 1993;4(4):247–56. [11] Danzon PM, Furukawa MF. Prices and availability of pharmaceuticals: evidence from nine countries. Health Affairs 2003;22(6):W521–36. [12] Garattini L, Tediosi F. A comparative analysis of generics markets in five European countries. Health Policy 2000;51(3):149–62. [13] Ufficio italiano dei cambi: http://www.uic.it. [14] Eurostat: http://epp.eurostat.ec.europa.eu.
