Prices of antihypertensive medicines in sub-saharan Africa and alignment to WHO s model list of essential medicines

Tropical Medicine and International Health doi:10.1111/j.1365-3156.2009.02453.x volume 15 no 3 pp 350 361 march 2010 Prices of antihypertensive medicines in sub-saharan Africa and alignment to WHO s model list of essential medicines Marc Twagirumukiza 1,2, Lieven Annemans 3, Jan G. Kips 2, Emile Bienvenu 4 and Luc M. Van Bortel 2 1 Department of Internal Medicine, Faculty of Medicine, National University of Rwanda, Butare, Rwanda 2 Heymans Institute of Pharmacology, Ghent University, Ghent, Belgium 3 Center for Health Economics, Department of Public Health, Ghent University, Ghent, Belgium 4 Department of Pharmacy, Faculty of Medicine, National University of Rwanda, Butare, Rwanda Summary objective To investigate compliance of National Essential Medicines Lists (NEMLs) with the WHO Essential Medicines List (WHO EML) in 2007 and to compare prices of antihypertensive drugs in and between 13 sub-saharan African countries. methodology Data on NEMLs and drug prices were collected from 65 public and 65 private pharmacies (five of each per country). Prices were compared with the International Drug Price Indicator Guide (IDPIG). The cost of drug treatment within a country was calculated using defined daily doses (DDD) and between countries using DDD prices adjusted for purchasing power parity-based gross domestic product per capita. results All surveyed countries had a NEML. However, none of these lists were in complete alignment with the 2007 WHO EML, and 38% had not been updated in the last 5 years. Surveyed medicines were cheaper when on the NEMLs; they were also cheaper in public than in private pharmacies. Prices varied greatly per medicine. A large majority of the public prices were higher than those indicated by the IDPIG. Overall, hydrochlorothiazide is the cheapest drug. conclusion There are substantial differences in NEML composition between the 13 countries. The proportion of NEMLs not regularly updated was double the global United Nations estimates. Prices of WHO EML-advised drugs differ greatly between drugs and for each drug within and between countries. In general, the use of drugs on the NEML improves financial accessibility, and these drugs should be prescribed preferentially. keywords sub-saharan Africa, drug costs, antihypertensive agents, hypertension, drugs, essential medicines, World Health Organization Introduction Accessibility of medicines is a matter of great concern for health services all over the world, especially for less developed countries. It is an explicit target of the United Nations (UN) Millennium Development Goals (United Nations 2008). WHO estimates that almost 2 billion people one-third of the global population do not have regular access to essential medicines, and in sub-saharan Africa (SSA) this is estimated at nearly half the population (WHO 2004a). Essential medicines are defined by WHO as those drugs that satisfy the health care needs of the majority of the population; they should, therefore, be available at all times in adequate amounts and in appropriate dosage forms, at a price the community can afford. A model list first assembled in 1977 and revised every 2 years since gives member countries an example to adapt a National Essential Medicine List (NEML) according to the country s needs. These NEMLs are meant to help health authorities, especially in developing countries, to optimize pharmaceutical resources (Kindermans & Matthys 2001; Reidenberg 2007; United Nations 2008). However, according to the UN report, a large gap still exists in the availability of medicines in both the public and private sectors, as well as a wide variation in prices, which render essential medicines unaffordable to poor people (United Nations 2008). Limited access to medicines mainly affects adherence to long treatment diseases (Mendis et al. 2007). This is particularly the case for cardiovascular diseases (Elliott 1996; Fischer & Avorn 2004), specifically hypertension (Shulman et al. 1986). Hypertension in SSA places an increasing economic burden on the limited resources of these developing countries (Montgomery 1998; Cooper et al. 1998; Seedat 2007), not only because 350 ª 2010 Blackwell Publishing Ltd

of treatment and management costs (Lindholm & Werko 1996; Pardell et al. 2000; Spurgeon 2004) but also because of loss of life associated with hypertension that occurs among adults of working age (Walker & Wadee 2000; Gaziano 2005). In the treatment of hypertension, medicine costs have been identified as the major determinant of cost of care (Odell & Gregory 1995; Dias da Costa et al. 2002), namely around 80% of the total cost of hypertension care within the first year of treatment (Shulman et al. 1986; Kaplan 1990). Although research has been undertaken in the Asia Pacific region (Balasubramaniam 1996) on prices of medicines used in communicable disease (Mugisha et al. 2002; Guimier et al. 2005; Gelders et al. 2006; Mendis et al. 2007; HAI 2008), independent studies evaluating the patient prices of antihypertensive drugs or comparing them with international price standards in SSA countries are scarce and sometimes limited to few drugs (HAI 2008) or just to one SSA country (Mendis et al. 2007). A very recent publication (Cameron et al. 2009b) draws attention to key barriers to access medicines: it discusses policy options to lower prices and improves treatment affordability by analysing reports from 36 countries including 10 from SSA, but only three antihypertensive drugs were on the analysed medicine list. The present survey had 3 aims: (i) to examine the availability of a NEML; (ii) to compare prices of antihypertensive medicines advocated by the WHO Essential Medicines List (WHO EML) within and between SSA countries and (iii) to compare these prices with the International Drug Price Indicator Guide (IDPIG) (MSH and WHO 2007) set by the Health Action International (HAI) World Health Organization (WHO). The IDPIG is published by Management Sciences for Health (MSH) in collaboration with WHO and contains a spectrum of prices from pharmaceutical suppliers and or buyers, international development organizations and government agencies. Methods The survey was carried out from June to July 2007 in 13 SSA countries: Rwanda, Burundi, Tanzania, Uganda, Democratic Republic of Congo (DRC), Kenya, Cameroon, Congo, Gabon, Ivory Coast, Senegal, Niger and Mozambique. We ascertained whether an NEML existed and the prices of antihypertensive medicines. All hypertension medicines advocated by the 15th WHO EML (WHO 2007) were surveyed: amlodipine, atenolol, enalapril, hydralazine, hydrochlorothiazide and methyldopa. The survey also included advocated drugs from previous lists (12th 14th WHO EML) (WHO 2003, 2004b, 2006), which were still used in more than half of the sampled countries (captopril, nifedipine). Data were collected by 13 local investigators (one correspondent investigator in each country). All investigators were either medical doctors working in the public sector (10 countries), or pharmacists working in the public sector (in countries: Rwanda, Uganda and DRC). All investigators received the instructions manual and report form prior to proceed to the data collection. They were asked to get the information on the existence and content of NEML from the Ministry of Health or National Pharmaceutical Office or governments central medical stores. The prices for the patient of different medicines were collected from five public hospital pharmacies and five private pharmacies, randomly selected in the capital town of the country. To randomize, all existing pharmacies in every country capital city were listed by outlet (public or private), and a number was allocated to each. An external person was then asked to choose blindly five numbers in the public basket and five in the private basket. The prices were elicited according to WHO and HAI guidelines (WHO and HAI 2009), i.e., only generic medicines with known origin and original products could be included. All collected prices were converted into US Dollars with the national official rate of July 2007. Defined daily dose (DDD) prices were used to compare prices between medicines. DDD prices were calculated, based on DDDs given by the WHO Collaborating Centre for Drug Statistics Methodology database (WHO 2008). A price adjusted for standard of living, inflation or deflation was estimated by adjusting the DDD price for purchasing power parity (PPP)-based gross domestic product (GDP) per capita (World Bank 2007). Although no reliable official source was found in countries about the salary of the lowest-paid unskilled government worker (LPGW) advocated by WHO HAI (WHO & HAI 2008) to estimate affordability, PPP-based GDP per capita adjusted price used in other surveys (Bloor et al. 1996; Cameron et al. 2009b) is also reasonable because like LPGW it takes into account the relative cost of living and the inflation rates of the countries (WHO & HAI 2008). This income-adjusted price in the country was calculated as: Income adjusted price ¼ DDD price in the country Lowest PPP-based GDP per Capita in sampled countries PPP-based GDP per Capita country ª 2010 Blackwell Publishing Ltd 351

Adjustment was performed to the country with the lowest PPP-based GDP per capita. This income-adjusted price allows comparison between the sampled countries and facilitates relative price interpretation. Data analysis and other calculations For every medicine in every country, minimum, median and maximum prices were identified. Descriptive analysis was performed on all data, and the influence of being on the NEML on the medicine price was analysed using the Mann Whitney-U-test with a-significance set at 0.05. The median price of the five samples in the private and public sector, respectively, was calculated to compare prices in the private and the public sector and to compare those prices with corresponding median prices in the IDPIG. Results WHO EML and NEML content The current (2007) WHO EML (WHO 2007) advises four first-line antihypertensive medicines from four classes: diuretics, b-blockers, angiotensin converting enzyme (ACE) inhibitors and calcium channel blockers. The four medicines are hydrochlorothiazide, atenolol, enalapril and amlodipine. Other advocated antihypertensive drugs are methyldopa and hydralazine. Advocated antihypertensive drugs did not differ between the 2007 and the 2005 (WHO 2006) list. But in 2005, WHO EML shifted from nifedipine sustained release to amlodipine and in 2003 from captopril to enalapril (WHO 2004a, 2006). Also in 2003, reserpine was dropped from the list. All 13 countries had a NEML. All medicines found in public pharmacies that are on the NEML are supplied from the Government s Central Medical Store. Table 1 shows the presence of medicines on the different NEMLs. No country had all four first-line medicines advocated by the 2007 WHO EML on its NEML. When WHO EML lists from 2002 (WHO 2003) on were taken into account, six countries had all four advocated first-line drugs on its NEML and six other countries had three on their NEML. These observations suggest a delay in the adaptation of the NEMLs. Indeed, five NEMLs had not been updated for 5 years. Of the other antihypertensive drugs, methyldopa was on the NEML in all 13 countries, while hydralazine was on nine of 13 NEMLs. Reserpine was advocated until 2002 and is still on the list of four countries. Although the short-acting nifedipine has been removed from the list in 2002, 11 of 13 countries still have it on their NEML. Table 1 also shows for each country the number of antihypertensive drugs aligned to the latest 2007 WHO EML and aligned to the last four WHO EMLs from 2002 through 2007. Prices in public and private sector per tablet and international reference prices Our method resulted in 10 surveyed drug formulations. The overall median price in the 13 countries was always lower in the public than in the private sector (Table 2). The price variation expressed by the price high-low ratio is also lower in the public than in the private sector, except for atenolol (Table 3). Considering median prices in each country, in four cases, the median price in the private sector was double the public price or more: atenolol 50 mg in Burundi, atenolol 100 mg in Benin, hydralazine 25 mg in Uganda and hydrochlorothiazide 25 mg in Mozambique. However, we also found a few exceptions where the median price was slightly higher in the public than in the private sector: methyldopa 250 mg in Kenya, enalapril 5 mg in Ivory Coast and nifedipine sustained release 20 mg in Senegal. For all 10 investigated drug formulations, the overall median public and private prices in the 13 countries were higher than the corresponding median price reported by the IDPIG, with the biggest difference for enalapril, where the median public and private price was 22.5 and 34.8 times more expensive than the IDPIG price, respectively. This higher price in private as well as in public sector compared to the IDPIG price also holds in every surveyed country for three formulations (enalapril 5 mg, hydrochlorothiazide 25 mg and nifedipine sustained release 20 mg), in all but one country for hydrochlorothiazide 50 mg and in 11 countries for atenolol 50 mg and hydralazine 25 mg. Treatment prices To compare expenses of treatment between medicines, the price of the DDD per medicine was calculated using the median price of each medicine in the public pharmacies or in the private pharmacy if this price was lower. In the case of atenolol and hydrochlorothiazide, the lowest median price of the two formulations was used. Table 4 shows that hydrochlorothiazide was the cheapest treatment in nine of 13 countries, while this was atenolol in three countries (in Uganda equal to hydrochlorothiazide) and captopril in two countries. Enalapril was the most expensive treatment in all but one country. To compare the prices of drug treatments between countries, adjusted prices were calculated by adjusting all DDD prices for the PPP-based GDP per capita of DRC, the country with the lowest 2007 PPPbased GDP per capita. Based on the median of all incomeadjusted prices in the 13 countries (Table 4), enalapril was 352 ª 2010 Blackwell Publishing Ltd

Table 1 Antihypertensive drugs on the National Essential Medicine Lists (NEMLs) Medicine classes Medicines (tablets) On WHO EML list* Benin Burundi Cameroon Congo Democratic Republic of Congo Ivory coast Kenya Mozambique Niger Rwanda Senegal Tanzania Uganda Countries with medicine on NEML n % CCB Amlodipine, 5 mg 2007 Y 1 8 Diltiazem No Y Y Y Y Y Y 6 46 Nifedipine 10 mg) No Y Y Y Y Y Y Y Y Y Y Y 11 85 Nifedipine 20 mg (SRA) 2003 Y Y Y Y Y Y Y 7 54 Nicardipine No Y Y 2 15 Verapamil No Y Y Y Y 4 31 BB Atenolol 50 or 100 mg 2007 Y Y Y Y Y Y Y Y Y Y 10 77 Acebutolol No Y 1 8 Propranolol No Y Y Y Y Y Y Y Y Y Y Y Y 12 92 Carvedilol No Y Y Y 3 23 ACEI Enalapril 5 mg 2007 Y Y 2 15 Captopril 2002 Y Y Y Y Y Y Y Y Y Y Y Y 12 92 Ramipril No Y 1 8 DIU Hydrochlorothiazide 2007 Y Y Y Y Y Y Y Y Y Y Y 11 85 25or 50 mgà Bendroflumethiazide No Y 1 8 Furosemide 60 mg (SR) No Y Y 2 15 Acetazolamide 250 mg No Y 1 8 Amiodarone No Y Y 2 15 Spironolactone No Y Y Y Y Y Y Y Y Y 9 69 Altizide + Spironolactone No Y Y 2 15 (15 + 25) mg CA Methyldopa 2007 Y Y Y Y Y Y Y Y Y Y Y Y Y 13 100 Clonidine No Y Y Y Y Y 5 38 ARB Losartan 80 mg No Y 1 8 Valsartan 80 mg No Y 1 8 Others Hydralazine 2007 Y Y Y Y Y Y Y Y Y 9 69 Reserpine (Serpasil) 2002 Y Y Y Y 4 31 Summary Adherence to 2007 3 2 5 4 4 2 5 4 1 3 4 5 4 WHO EML Adherence to WHO EML 2002 2007 5 4 6 6 6 4 6 5 4 3 6 6 6 Some listed drugs still used in sub-saharan Africa are no longer first line antihypertensive in developed countries. CCB, calcium channel blockers; BB, beta-blockers; ACEI, angiotensin converting enzyme inhibitors; DIU, diuretics; CA, centrally acting drugs; ARB, angiotensin II receptors blockers; SR, sustained release; Y, drug available on NEML; No, the medicine is not on the WHO EML lists of 2002, 2003, 2005 or 2007; n, number of countries having the drug on their NEML; EML, Essential Medicine Lists. *Year of the latest WHO EML list with the medicine on the list. WHO EML mentions 2.5 mg tablets. àwho EML only mentions 25 mg tablets. Number out of six antihypertensive drugs advised by the current (2007) WHO EML on the country s NEML. Number out of nine antihypertensive drugs from different classes advised by at least one of the WHO EML between 2002 and 2007 present on the country s NEML. by far the most expensive treatment, being 5.3 times more expensive than amlodipine, the second most expensive treatment. Hydrochlorothiazide is the most affordable, being 2.7 times cheaper than captopril, the second most affordable (Table 4). Influence of NEML on medicine prices Adjusted medicine prices show that being on NEML reduces the cost of a medicine (Table 5). This was statistically significant for medicines on the current (2007) ª 2010 Blackwell Publishing Ltd 353

Table 2 Price per tablet in public and private sector per country Median Medicine prices per tablet in US Dollarcent Countries Outlets MET 250 mg AML 5mg ATE 50 mg ATE 100 mg CAP 25 mg ENA 5mg HYD 25 mg HCT 25 mg HCT 50 mg NIF 20 mg Benin Public 4.2 28.0 0.7 4.4 0.6 17.7 2.4 0.5 1.0 2.9 Private 5.2 39.5 1.0 10.2 0.8 28.5 3.6 0.5 1.3 2.9 Burundi Public 4.0 25.5 7.0 8.0 3.1 19.0 2.6 7.0 12.0 9.7 Private 4.7 33.9 14.0 13.0 5.1 23.4 5.0 9.0 11.7 11.4 Cameroon Public 4.0 10.0 20.0 20.0 2.0 37.0 2.5 7.0 15.0 9.3 Private 4.2 19.8 25.0 22.0 2.1 52.1 3.9 7.6 27.4 18.4 Congo Public 3.4 16.5 11.0 0.9 6.0 56.0 2.0 5.1 4.0 7.4 Private 3.5 23.0 11.2 1.6 11.9 83.4 2.0 5.5 4.0 8.0 Democratic Public 4.1 11.0 2.4 6.0 10.0 45.0 2.4 8.0 9.6 9.1 Republic of Congo Private 5.9 21.1 3.8 6.6 10.4 69.7 3.6 9.7 15.5 18.0 Ivory coast Public 7.2 108.5 24.0 25.0 30.0 100.0 3.4 2.5 5.0 12.9 Private 9.8 197.5 43.2 27.5 31.2 90.0 4.8 3.7 9.0 16.5 Kenya Public 1.2 20.0 2.4 1.8 1.0 95.0 2.9 3.0 5.0 11.0 Private 1.1 37.2 2.9 2.5 1.5 163.9 3.1 4.7 7.9 14.2 Mozambique Public 1.6 18.4 4.2 5.5 1.5 11.5 1.9 5.0 11.1 7.1 Private 2.5 29.8 5.8 6.8 1.5 21.6 1.9 10.1 13.2 7.7 Niger Public 4.3 12.5 11.5 9.0 2.0 150.0 4.1 5.2 9.6 15.5 Private 6.9 21.8 16.1 10.3 4.0 298.4 6.1 9.9 15.4 16.8 Rwanda Public 0.8 18.5 2.0 5.6 1.5 50.0 3.7 0.5 0.8 13.9 Private 0.9 26.1 2.4 8.1 2.5 95.0 5.1 0.5 0.8 22.9 Senegal Public 1.6 40.0 15.0 7.7 1.0 10.0 2.9 5.0 5.1 10.9 Private 2.5 59.5 15.7 9.2 1.3 12.5 4.6 5.5 5.1 10.7 Tanzania Public 1.0 10.0 12.0 3.5 3.0 100.0 2.3 5.0 5.8 8.6 Private 1.4 19.8 16.5 5.9 4.3 179.0 3.2 5.0 6.2 10.3 Uganda Public 0.5 10.0 0.7 7.0 1.5 30.0 3.9 0.5 4.0 14.7 Private 0.8 18.8 0.9 7.7 1.6 41.2 8.0 0.5 4.4 24.9 Overall median price Public 3.4 18.4 7.0 6.0 2.0 45.0 2.6 5.0 5.1 9.7 Private 3.5 26.1 11.2 8.1 2.5 69.7 3.9 5.5 7.9 14.2 IDPIG 3.2 13.3 1.5 1.2 1.1 2.0 2.1 0.4 0.9 2.5 Prices per country are median of tablet price in five pharmacies in the public and five in the private sector. Prices in bold are lower than or equal to IDPIG prices. MET, methyldopa; AML, amlodipine; ATE, atenolol; CAP, captopril; ENA, enalapril; HYD, hydralazine; HCT, hydrochlorothiazide; NIF, nifedipine sustained release; Overall median price, median of all 65 (13 countries five samples) prices in private and public pharmacies, respectively; IDPIG, median price advocated by the International Drug Price Indicator Guide. WHO EML, and tended to be statistically significant for captopril and nifedipine sustained release, which were on the 2003 and 2002 WHO EML, respectively. The cost of a treatment with a medicine, which is not on NEML, was on average 1.69 ± 0.44 (ranging from 1.27 to 2.61) times higher than the same medicine on the NEML. Discussion This study compared drugs on NEMLs in SSA countries with drugs advocated by the WHO EML, investigated the impact of being listed on NEMLs on prices in SSA, and drug prices in 13 SSA countries as well as data on the most frequently used hypertensive drugs (10 formulations of eight drugs). All 13 countries had NEMLs, but none of these lists were in complete alignment with the 2007 WHO EML, and 38% had not been updated in the last 5 years. Although reasons for this non-compliance were not investigated, studies in developing countries suggest political will, insufficiency of human resources or funding and conflict of interests (Hughes & Fiander 1989; Courtois & Dumoulin 1995; Bloor et al. 1996; Balasubramaniam 1996; Turshen 2001; Anis et al. 2003; Aaserud et al. 2006). The number of out-of-date NEMLs is double the UN report estimates of 19% of developing countries who need to establish an EML or update an existing one (United Nations 2008). Outdated NEMLs may handicap medicine access and standardization of hypertension management in 354 ª 2010 Blackwell Publishing Ltd

Table 3 Price variation per tablet in public and private sector per country Variation in Medicine prices per tablet expressed as High Low Ratio (HLR) Countries Outlets MET 250 mg AML 5mg ATE 50 mg ATE 100 mg CAP 25 mg ENA 5mg HYD 25 mg HCT 25 mg HCT 50 mg NIF 20 mg Benin Public 1.6 3.1 2.6 8.7 5.0 11.8 3.7 2.0 3.0 1.4 Private 2.2 2.2 4.8 9.5 5.5 12.5 3.6 2.0 2.6 1.9 Burundi Public 1.1 2.6 4.9 9.8 1.7 3.9 4.0 2.0 1.8 19.4 Private 1.5 5.9 9.8 17.6 3.6 3.2 4.2 4.4 2.6 22.6 Cameroon Public 6.3 9.0 3.5 4.9 1.5 8.4 3.6 1.8 4.5 17.3 Private 12.3 6.0 6.5 11.1 2.2 8.8 4.7 3.3 9.5 32.9 Congo Public 1.5 2.6 3.5 4.7 3.4 3.7 3.7 3.0 8.0 7.5 Private 2.0 3.0 7.4 4.5 4.5 3.6 5.3 3.0 8.0 5.4 Democratic Public 1.1 1.9 13.6 2.4 3.3 12.5 3.7 1.6 1.3 7.2 Republic of Congo Private 1.6 1.9 23.8 4.0 5.2 9.4 3.5 1.8 1.4 7.2 Ivory coast Public 1.4 5.3 12.0 6.2 3.3 6.3 3.8 4.5 1.6 5.6 Private 2.1 5.2 9.0 15.5 3.7 7.9 4.2 4.3 1.4 5.3 Kenya Public 2.8 3.8 4.2 3.2 1.9 2.5 3.7 1.3 1.6 9.0 Private 3.4 4.8 3.1 3.4 2.8 2.8 3.2 2.0 2.1 11.1 Mozambique Public 1.3 2.4 2.1 8.7 6.0 10.7 3.6 1.6 1.5 13.9 Private 2.0 3.4 3.1 21.7 6.0 11.4 4.0 2.8 2.4 8.2 Niger Public 2.4 2.7 8.3 2.4 3.0 4.2 3.6 3.6 9.1 5.8 Private 2.0 3.3 10.0 6.1 3.2 2.7 5.2 7.0 9.1 5.3 Rwanda Public 8.1 5.6 2.0 3.1 1.7 2.2 3.7 3.0 3.5 5.5 Private 10.4 9.6 4.1 7.8 1.5 2.3 3.7 3.2 4.2 4.9 Senegal Public 3.3 8.9 1.7 7.5 2.5 5.7 3.6 3.0 3.2 17.4 Private 3.0 15.2 2.5 8.4 4.5 7.0 2.7 3.2 3.3 22.1 Tanzania Public 2.2 2.4 1.6 2.9 4.2 2.2 5.0 3.4 1.0 3.1 Private 2.6 2.8 2.3 3.7 3.9 2.9 3.8 3.4 1.3 3.4 Uganda Public 1.5 3.1 4.5 5.6 4.5 8.0 3.6 3.6 9.0 2.4 Private 1.5 2.0 5.4 7.8 4.7 8.9 4.0 4.0 9.0 2.5 All countries Public 9.0 52.9 40.0 17.3 12.0 70.3 9.0 9.0 41.0 122.0 Private 14.0 86.1 31.7 35.3 19.0 74.0 10.8 14.0 41.0 234.2 IDPIG HLR 3.1 13.9 2.3 5.5 5.6 12.3 1.8 2.2 43.9 3.8 HLR in bold is lower than or equal to IDPIG HLR. MET, methyldopa; AML, amlodipine; ATE, atenolol; CAP, captopril; ENA, enalapril; HYD, hydralazine; HCT, hydrochlorothiazide; NIF, nifedipine sustained release; HLR, High low ratio is the ratio of the highest price over the lowest price; IDPIG HLR, HLR according to the International Drug Price Indicator Guide. sub-saharan countries. Listing of medicines on NEMLs favours their use but does not mean that they are more often prescribed by physicians. However, as we showed that NEML drugs are cheaper, their listing on NEML is a powerful tool on price reduction and control and may increase their use. The patient price of medicines largely differs between countries for each drug and between drugs in each country. In the majority of countries, hydrochlorothiazide was the cheapest drug. In low- and middle-income countries, limited access to medicines is mainly because of two factors (Smith & Tickell 2003): poor availability and relatively high cost (Quick et al. 2002; WHO and HAI 2009; Mendis et al. 2007). Our study confirms others (WHO 1999; Quick et al. 2005) showing that appearance of a medicine on the NEML lowers its price. Of the many factors that can influence drug price, three are especially important: (i) mode of procurement and existence of local manufacturers (supplier, travel distances, custom fees and other taxes); (ii) patent issues and (iii) existence of the medical insurances or government subsidiaries if the drug is on the NEML (Perez-Casas et al. 2001; Levison & Laing 2003; Linjer et al. 2005). If a medicine appears on the NEML, the add-on costs are greatly reduced as taxes and custom fees are lowered, suppliers are carefully chosen, and medicine prices are subsidized (Levison & Laing 2003). None of the 13 investigated countries has a mandatory social security or a national inclusive medical insurance covering the whole population (Annemans et al. 1997). ª 2010 Blackwell Publishing Ltd 355

Table 4 DDD Prices and purchasing power parity (PPP)-based gross domestic product (GDP) per capita adjusted DDD prices in US Dollarcent Countries 2007 PPP-GNI Per capita Medicines Methyldopa (Aldomet) Amlodipine Atenolol Captopril Enalapril Hydralazine tablets Hydrochlorothiazide DDD (mg) 1000 5 75 50 10 100 25 30 Nifedipine (long acting) High Low HLR Benin 1310 DDD Price 16.8 28.0 0.5 1.2 35.4 9.6 0.5 4.4 ENA HCT 70.8 Adjusted price 3.7 6.2 0.1 0.3 7.8 2.1 0.1 1.0 Burundi 330 DDD Price 16.0 25.5 5.3 6.2 38.0 10.4 5.9 14.6 ENA HCT 7.2 Adjusted price 14.1 22.4 4.6 5.4 33.4 9.1 5.1 12.8 Cameroon 2120 DDD Price 16.0 10.0 15.0 4.0 74.0 10.0 7.0 14.0 ENA CAP 18.5 Adjusted price 2.2 1.4 2.1 0.5 10.1 1.4 1.0 1.9 Congo 2750 DDD Price 13.6 16.5 1.4 12.0 112.0 8.0 2.0 11.1 ENA HCT 83.0 Adjusted price 1.4 1.7 0.1 1.3 11.8 0.8 0.2 1.2 Democratic Republic of Congo 290 DDD Price 16.4 11.0 1.8 20.0 90.0 9.6 4.8 13.7 ENA ATE 50.0 Adjusted price 16.4 11.0 1.8 20.0 90.0 9.6 4.8 13.7 Ivory coast 1590 DDD Price 28.8 108.5 18.0 60.0 180.0 13.6 2.5 19.4 ENA HCT 72.0 Adjusted price 5.3 19.8 3.3 10.9 32.8 2.5 0.5 3.5 Kenya 1540 DDD Price 4.4 20.0 1.8 2.0 190.0 11.6 2.5 16.5 ENA ATE 105.6 Adjusted price 0.8 3.8 0.3 0.4 35.8 2.2 0.5 3.1 Mozambique 690 DDD Price 6.4 18.4 3.2 3.0 23.0 7.6 5.0 10.7 ENA CAP 7.7 Adjusted price 2.7 7.7 1.3 1.3 9.7 3.2 2.1 4.5 Niger 630 DDD Price 22.4 17.2 10.4 6.0 448.4 20.4 6.3 24.2 ENA HCT 74.7 Adjusted price 7.9 5.8 4.0 1.8 138.1 7.5 2.2 10.7 Rwanda 860 DDD Price 3.2 18.5 1.5 3.0 100.0 14.8 0.4 20.9 ENA HCT 250.0 Adjusted price 1.1 6.2 0.5 1.0 33.7 5.0 0.1 7.0 Senegal 1640 DDD Price 6.4 40.0 11.3 2.0 20.0 11.6 2.6 16.1 NIF HCT 20.0 Adjusted price 1.1 7.1 2.0 0.4 3.5 2.1 0.5 2.8 Tanzania 1200 DDD Price 4.0 10.0 5.3 6.0 200.0 9.2 2.9 12.9 ENA HCT 69.0 Adjusted price 1.0 2.4 1.3 1.5 48.3 2.2 0.7 3.1 Uganda 920 DDD Price 2.0 10.0 0.5 3.0 60.0 15.6 0.5 22.1 ENA ATE 120.0 Adjusted price 0.6 3.2 0.2 0.9 18.9 4.9 0.2 7.0 Median DDD Price 13.6 18.4 3.2 4.0 90.0 10.4 2.6 14.6 Adjusted price 2.2 6.2 1.3 1.3 32.8 2.5 0.5 3.5 High DDD Price 28.8 108.5 18.0 60.0 448.4 20.4 7.0 24.2 Adjusted price 16.4 22.4 4.6 20.0 138.1 9.6 5.1 13.7 Low DDD Price 2.0 10.0 0.5 1.2 20.0 7.6 0.4 4.4 Adjusted price 0.6 1.4 0.1 0.3 3.5 0.8 0.1 1.0 HLR DDD Price 14.4 10.9 34.3 50.0 22.4 2.7 17.5 5.6 Adjusted price 26.0 16.4 39.7 75.3 39.0 11.4 46.4 14.2 MET, methyldopa; AML, amlodipine; ATE, atenolol; CAP, captopril; ENA, enalapril; HYD, hydralazine; HCT, hydrochlorothiazide; NIF, nifedipine sustained release; DDD price, defined daily dose prices were calculated from median prices in the public pharmacies, or from median price in the private pharmacies if this price was lower; Adjusted price, PPP-based GDP per capita adjusted DDD price. 356 ª 2010 Blackwell Publishing Ltd

Table 5 Effect of presence on National Essential Medicines List (NEML) on drug price Medicines On NEML n* Mean prices Standard deviation No yes ratio P-valueà Methyldopa Yes 130 4.48 1.00 No 0 Amlodipine Yes 10 3.77 6.62 2.10 0.003 No 120 7.91 9.23 Atenolol Yes 100 1.58 1.03 1.55 <0.001 No 30 2.46 0.88 Captopril Yes 120 3.72 2.69 0.27 NS No 10 1.01 4.37 Enalapril Yes 20 29.23 13.21 1.29 <0.001 No 110 37.78 16.80 Hydralazine Yes 70 3.17 7.99 1.91 <0.001 No 60 6.04 10.04 Hydrochlorothiazide Yes 110 1.03 2.41 3.57 <0.001 No 20 3.68 2.61 Nifedipine (sustained release) Yes 60 7.42 3.25 1.13 NS No 70 8.39 6.58 NS, Not statistically significant. *Number of pharmacies where prices were collected. Mean adjusted defined daily dose prices per country. àstatistics from Mann Whitney-U-test. However, there are community-based, district-based or national medical insurances with a co-payment method. Within this system, the patient has to pay a small fraction of the medicine price as own individual contribution, ranging from 0% (as in Uganda for referred patient prescriptions) to 25%, the remainder being subsidized by medical insurance. The co-payment system has a positive impact on price reduction. Complete or partial absence of a social security emphasizes the need for a NEML, which provides more affordable drugs. Although the WHO EML is updated every 2 years, a government can decide not to follow the latest WHO EML because of a more affordable drug in the same class. This may be the case for keeping captopril on the NEML, which was overall 26.0 times (income-adjusted DDD price difference) less expensive than enalapril, the ACE inhibitor advocated by the latest WHO EML. Nevertheless, because of the frequent change of drug prices, as in case of patent expiry, governments should be encouraged to consider updating the NEML each time a new version of the WHO EML appears to guarantee maximal affordability of drugs. We found that five NEMLs had not been updated for 5 years, which caused older antihypertensives, which had been superseded by short-acting nifedipine to remain on the NEML. The majority of patient prices of medicines in this survey were higher than the corresponding buyer prices of the IDPIG. This finding may be in line with other authors (Laing 1991; Mallet et al. 2001; Levison & Laing 2003) who found that prices of medicines in most SSA countries are well above their production costs (Richards 1986; Richard 2004; Ewen & Dey 2007) and that the profits of those in the distribution chain (pharmacists, dispensing doctors, wholesalers and even some governments) are frequently high. The UN report (United Nations 2008) estimates that in the public sector, generic medicines cost on average 250% more than the international reference price and in the private sector, those same medicines cost on average about 650% more than the international reference price. After exclusion of the outlier enalapril, in the present survey, patient prices of sampled antihypertensive medicines were even more expensive than estimated by the UN report, being on average 313% and 745% higher than international reference prices in the public and private sector, respectively. Like IDPIG, the present study showed that absolute patient prices and variation in prices in private pharmacies are in general higher than in public pharmacies. This can be explained by the fact that only public pharmacies are involved in the NEML program to provide more affordable drugs. However, in the present study, the difference between public and private pharmacy prices was smaller than what has been reported by HAI in the Medicine price monitor monitoring three pilot countries : Uganda, Kenya and Tanzania (HAI 2008). In the private sector, pharmacies and outlets are often poorly regulated, making the prices higher than in the public sector. However, in Kenya, the price of atenolol and hydralazine was lower in the private than in the public ª 2010 Blackwell Publishing Ltd 357

sector. This may be linked to the competition between several local manufacturers and import export companies, which suggests also a potentially important role for local manufacturing in increasing the accessibility of medicines. Furthermore, establishing an authority to continuously monitor prices, availability, affordability and linked components, might minimize prices fluctuations and help increasing health care access in SSA. International guidelines advise diuretics and calcium channel blockers as preferred drugs in blacks (Mancia et al. 2007). The diuretic hydrochlorothiazide was the best affordable in a majority of countries and in the 13 countries as a whole, being 2.6 times less expensive than the second most affordable drug captopril and 13.2 times less expensive than amlodipine. This means that for the cost of treating one patient with amlodipine, at least 13 patients can be treated with hydrochlorothiazide. Hence, hydrochlorothiazide should be, in line with previous reports (Helgeland et al. 1986; Philipp et al. 1997; Mayor 2003), the first drug to be considered. Therefore, sustained efforts should be made to provide hydrochlorothiazide at the best affordable price (Jayasinghe et al. 2003; Spurgeon 2004), especially in those countries where the price of hydrochlorothiazide is high, such as Cameroon and DRC. A chronic use of thiazide diuretics can lead to metabolic disturbances. However, the use of low doses (e.g. not exceeding 25 mg daily of hydrochlorothiazide), even for a longer period, is normally well tolerated (Waller & Ramsay 1989). Therefore, from a public health perspective, control of urates, potassium and blood glucose may not be obligatory up to 25 mg of hydrochlorothiazide per day). Because of their low cost, thiazide diuretics are important baseline drugs in the treatment of hypertension. The shift from captopril to enalapril and nifedipine long acting to amlodipine made in WHO EML may not be appropriate for some low-income countries. Yet the advocated drugs are long acting and may increase treatment compliance since they are given once daily, but they may not be affordable in several developing countries. In most countries, treatment with enalapril was the most expensive one, ranging from 7.2 to 250 times more expensive than a treatment with hydrochlorothiazide in Burundi and Rwanda, respectively. We adjusted medicine prices for PPP-based GDP per capita to estimate drugs affordability. However, the WHO HAI recommended method, used in several other reports, is the salary of the LPGW (WHO & HAI 2008; Cameron et al. 2009b). Discussions and criticism are very often raised about using LPGW to estimate affordability (Niens & Brouwer 2009). The reality is that in many SSA countries a substantial proportion of the population earns less than this amount. Further, the need for other nondiscretionary expenditures (e.g., food or housing), seasonal fluctuations in income, the number of dependants who live on this wage and the full costs of treatment are not accounted for. In addition, it would be very interesting to know country by country the proportion of the population able to afford a complete treatment of a chronic disease. However, this requires economic modelling with many assumptions, such as which proportion of the salary can be spent, the composition of the household and the number of drugs to be used to control the disease. These data are not readily available and therefore we decided not to proceed with such approach. Given the complex nature of affordability, multiple measures are valuable in gaining a robust understanding of this concept as it applies to medicine costs (Cameron et al. 2009a,b). Because of the complexity and size of the problem of hypertension in SSA, the main limitations of our survey lie in what was not assessed rather than what was investigated. As the outlets surveyed were chosen in each country from the capital city, data from distanced rural areas could change according to transport add-ons. The survey was limited to drugs on NEMLs, which were on the WHO EML between 2002 and 2007. Prices of other nonantihypertensive drugs on NEMLs were not analysed. The present study ignored the price data from informal channels, such as street vendors, which should interact with the prices in SSA countries. Apart from the price, the quality of medicines is also of utmost importance in treatment. However, the quality of medicine was out of scope of the present manuscript. Another limitation may be related to the use of the IDPIG as a price reference. The IDPIG has been advocated as the most useful reference standard given its wide availability, frequent updates and fair stability over time (WHO & HAI 2008; Cameron et al. 2009b). It has also claimed to represent actual procurement prices for medicines offered to low-income and middle-income countries by non-profit suppliers and international tender prices (Cameron et al. 2009b). Nevertheless, it may be criticized as it is based on a limited number of data for some medicines. Finally in this manuscript, the prices discussed are prices for monotherapy. But this does not necessarily reflect the cost of the management of hypertension because a patient with established hypertension regularly requires more than one antihypertensive drug and often also other cardioprotective drugs like acetylsalicylic acid or statins. Additionally, the present study is descriptive and not explanatory. It was not possible to analyse the reason of price disparities because the price components were not investigated due to logistical limitations and or governmental sensitivity in delivering such information. However, the results of this survey show the picture of real patient prices of 358 ª 2010 Blackwell Publishing Ltd

antihypertensive medicines and NEMLs in SSA countries and highlight the role NEMLs can play in regulating medicine prices in SSA. In contrast to previous studies (Balasubramaniam 1996; Mendis et al. 2007; HAI 2008; Cameron et al. 2009a), data were collected independently rather than through government surveys. Our conclusions, though in line with others published earlier, are more robust than previous reports. Conclusions In conclusion, all surveyed SSA countries have a NEML that partially aligns with WHO EDL. There are substantial between country differences in NEML. Prices of drugs advised by WHO EDL differ largely between drugs and for each drug within and between countries. Use of NEML drugs is likely to improve financial accessibility of care in SSA countries. The proportion of NEMLs that need updating was double the UN report estimate. Governments should be encouraged to consider updating the NEML each time a new version of the WHO EML appears to guarantee maximal accessibility of drugs. In general, hydrochlorothiazide is the most affordable drug and should be the drug to be considered first. Sustained efforts should be made to provide hydrochlorothiazide at the best affordable price. This survey adds to what is known about access-tomedicines patterns in developing countries. Such information is key when setting up a public health policy or taking a decision on treatment delivery for hypertension. Our findings are relevant especially in areas where like in SSA hypertension and other cardiovascular diseases are nowadays rapidly increasing because of demographic changes and urbanization. The information given will help clinicians and other health professionals to make decisions in treatment delivery especially in settings where patient purchasing power is limited. Acknowledgements We thank all local physicians and pharmacists who voluntarily contributed to the collecting of medicine price data, as well as all contacts within Government s Central Medical Store in the sampled countries for their contribution. References Aaserud M, Dahlgren AT, Kosters JP et al. (2006) Pharmaceutical policies: effects of reference pricing, other pricing, and purchasing policies. Cochrane Database of Systematic Reviews 19, CD005979. Anis AH, Guh DP & Woolcott J (2003) Lowering generic drug prices: less regulation equals more competition. Medical Care 41, 135 141. Annemans L, Crott R, De Clercq H et al. (1997) Pricing and reimbursement of pharmaceuticals in Belgium. Pharmacoeconomics 11, 203 209. Balasubramaniam K (1996) Health and Pharmaceuticals in Developing Countries: Towards Social Justice and Equity. Consumers International-Regional Office for Asia and the Pacific, Penang, Malaysia. Bloor K, Maynard A & Freemantle N (1996) Lessons from international experience in controlling pharmaceutical expenditure. III: regulating industry. BMJ 313, 33 35. Cameron A, Ewen M, Auton M & Laing R (2009a) Better measures of affordability required Authors reply. Lancet 373, 1081 1082. Cameron A, Ewen M, Ross-Degnan D, Ball D & Laing R (2009b) Medicine prices, availability, and affordability in 36 developing and middle-income countries: a secondary analysis. Lancet 373, 240 249. Cooper RS, Rotimi CN, Kaufman JS, Muna WF & Mensah GA (1998) Hypertension treatment and control in sub-saharan Africa: the epidemiological basis for policy. BMJ 316, 614 617. Courtois X & Dumoulin J (1995) Sale of drugs and health care utilization in a health care district in Zaire. Health Policy and Planning 10, 181 186. Dias da Costa JS, Fuchs SC, Olinto MT et al. (2002) Cost-effectiveness of hypertension treatment: a population-based study. São Paulo Medical Journal 120, 100 104. Elliott WJ (1996) The costs of treating hypertension: what are the long-term realities of cost containment and pharmacoeconomics? Postgraduate Medicine 99, 241 242. Ewen M & Dey D (2007) Medicines: Too Costly and Too Scarce. Briefing paper for World Health Assembly delegates 2007. http:// www.haiweb.org/medicineprices/2005/pricingbriefingpaper FINAL.doc. Accessed 17 June 2009. Fischer MA & Avorn J (2004) Economic implications of evidencebased prescribing for hypertension: can better care cost less? JAMA 291, 1850 1856. Gaziano TA (2005) Cardiovascular disease in the developing world and its cost-effective management. Circulation 112, 3547 3553. Gelders S, Ewen M, Noguchi N & Laing R (2006) Price, Availability and Affordability: an International Comparison of Chronic Disease Medicines. http://mednet3.who.int/medprices/ CHRONIC.pdf. Accessed 20 February 2008. Guimier JM, Candau D, Garenne M & Teulieres L (2005) Why drug prices are high in sub-saharan Africa. Analysis of price structure: the case of Senegal. Sante 15, 41 52. HAI (2008) Monitoring Medicine Prices, Availability and Affordability : Pilot Countries Reports. http://www. haiweb.org/medicineprices/. Accessed 5 September 2008. Helgeland A, Strommen R, Hagelund CH & Tretli S (1986) Enalapril, atenolol, and hydrochlorothiazide in mild to moderate hypertension. A comparative multicentre study in general practice in Norway. Lancet 1, 872 875. ª 2010 Blackwell Publishing Ltd 359

Hughes D & Fiander A (1989) The Bamako initiative. BMJ 299, 683. Jayasinghe S, Jayasinghe S & De Silva P (2003) State can make drugs available cheaply to poor people. BMJ 326, 553. Kaplan NM (1990) The cost-effectiveness of antihypertensive drugs. JAMA 263, 2888 2889. Kindermans JM & Matthys F (2001) Introductory note: the access to Essential Medicines Campaign. Tropical Medicine and International Health 6, 955 956. Laing R (1991) Essential drugs programmes in Africa. Africa Health 14, 32 33. Levison L & Laing R (2003) The hidden costs of essential medicines. Essential Drugs Monitor 33, 20 21. Lindholm LH & Werko L (1996) Comparing hypertension guidelines. Cost effectiveness analyses have been carried out in Sweden. BMJ 313, 1203 1204. Linjer E, Hedner T, Jonsson B et al. (2005) Cost analysis of different pharmacological treatment strategies in elderly hypertensives. Blood Pressure 14, 107 113. Mallet HP, Njikam A & Scouflaire SM (2001) Evaluation of prescription practices and of the rational use of medicines in Niger. Sante 11, 185 193. Mancia G, De Backer G, Dominiczak A et al. (2007) 2007 Guidelines for the Management of Arterial Hypertension: the Task Force for the Management of Arterial Hypertension of the European Society of Hypertension (ESH) and of the European Society of Cardiology (ESC). Journal of Hypertension 25, 1105 1187. Mayor S (2003) Thiazides could achieve major cost savings in uncomplicated hypertension. BMJ 327, 521. Mendis S, Fukino K, Cameron A et al. (2007) The availability and affordability of selected essential medicines for chronic diseases in six low- and middle-income countries. Bulletin of the World Health Organization 85, 279 288. Montgomery RW (1998) Hypertension treatment and control in Sub-saharan Africa. Amount spent on health care per capita is same as cost of a McDonald s. BMJ 317, 76. MSH & WHO (2007) International Drug Price Indicator Guide 2007. 2007 edn. MSH &WHO, Geneva. Mugisha F, Kouyate B, Gbangou A & Sauerborn R (2002) Examining out-of-pocket expenditure on health care in Nouna, Burkina Faso: implications for health policy. Tropical Medicine and International Health 7, 187 196. Niens LM & Brouwer WB (2009) Better measures of affordability required. Lancet 373, 1081 1082. Odell TW & Gregory MC (1995) Cost of hypertension treatment. Journal of General Internal Medicine 10, 686 688. Pardell H, Tresserras R, Armario P & Hernandez Del Rey R (2000) Pharmacoeconomic considerations in the management of hypertension. Drugs 59(Suppl. 2), 13 20. Perez-Casas C, Herranz E & Ford N (2001) Pricing of drugs and donations: options for sustainable equity pricing. Tropical Medicine and International Health 6, 960 964. Philipp T, Anlauf M, Distler A et al. (1997) Randomised, double blind, multicentre comparison of hydrochlorothiazide, atenolol, nitrendipine, and enalapril in antihypertensive treatment: results of the HANE study. HANE Trial Research Group. BMJ 315, 154 159. Quick JD, Hogerzeil HV, Velasquez G & Rago L (2002) Twentyfive years of essential medicines. Bulletin of the World Health Organization 80, 913 914. Quick JD, Boohene NA, Rankin J & Mbwasi RJ (2005) Medicines supply in Africa. BMJ 331, 709 710. Reidenberg MM (2007) World Health Organization program for the selection and use of essential medicines. Clinical Pharmacology and Therapeutics 81, 603 606. Richard V (2004) Funding for healthcare in sub-saharan Africa: cost recovery. Médecine Tropicale 64, 337 340. Richards T (1986) Drugs in developing countries: inching towards rational policies. British Medical Journal 292, 1347 1348. Seedat YK (2007) Impact of poverty on hypertension and cardiovascular disease in sub-saharan Africa. Cardiovascular Journal of Africa 18, 316 320. Shulman NB, Martinez B, Brogan D, Carr AA & Miles CG (1986) Financial cost as an obstacle to hypertension therapy. American Journal of Public Health 76, 1105 1108. Smith MK & Tickell S (2003) The essential drugs concept is needed now more than ever. Transactions of the Royal Society of Tropical Medicine and Hygiene 97, 2 5. Spurgeon D (2004) NIH promotes use of lower cost drugs for hypertension. BMJ 328, 539. Turshen M (2001) Reprivatizing pharmaceutical supplies in Africa. Journal of Public Health Policy 22, 198 225. United Nations (2008) Millennium Development Goal (MDG) Gap Task Force 8: Delivering on the Global Partnership for Achieving the MDGs, Report 2008. http://www.un.org/esa/ policy/mdggap. Accessed 5 April 2009. Walker AR & Wadee AA (2000) In the rationing of healthcare in indigent African populations, which services should come first? Journal of the Royal Society of Health 120, 152 157. Waller PC & Ramsay LE (1989) Predicting acute gout in diuretictreated hypertensive patients. Journal of Human Hypertension 3, 457 461. WHO (1999) Operational Principles for Good Pharmaceutical Procurement. WHO, Geneva. WHO (2003) The selection and use of essential medicines 2002. 12th Model list of essential medicines. World Health Organization Technical Report Series 914, 58 59. WHO (2004a) The World Medicines Situation. WHO, Geneva. WHO (2004b) The selection and use of essential medicines 2003. 13th Model list of essential medicines. World Health Organization Technical Report Series 920, 80 82. WHO (2006) The selection and use of essential medicines 2005. 14th model list of essential medicines. World Health Organization Technical Report Series 933, 76 78. WHO (2007) The selection and use of essential medicines 2007. 15th Model list of essential medicines. World Health Organization Technical Report Series 946, 104 105. WHO (2008) ATC DDD Index 2008. http://www.whocc.no/ atcddd/. Accessed 25 May 2008. 360 ª 2010 Blackwell Publishing Ltd

WHO & HAI (2008) Medicine Prices: a New Approach to Measurement. http://www.haiweb.org/medicineprices/manual/ manuals/medicineprices.pdf. Accessed 5 December 2008. WHO & HAI (2009) Measuring Medicine Prices, Availability, Affordability and Price Components. WHO, Geneva. World Bank (2007) World Development Indicators (WDI) Database, 2007. http://www.congre.co.jp/neurosci2009/english/ registration/images/gni_per_capita.pdf. Accessed 5 September 2009. Corresponding Author Luc M. Van Bortel, Heymans Institute of Pharmacology, Faculty of Medicine and Health Sciences, Ghent University, Belgium. E-mail: Luc.VanBortel@UGent.be ª 2010 Blackwell Publishing Ltd 361