Marketing and pricing strategies of online pharmacies

MARKETING AND PRICING STRATEGIES OF ON LINE PHARMACIES 1. INTRODUCTION Internet and e-commerce have profoundly changed society, the economy and the world of health care, not without ethical and legal consequences. Information about health care is widespread, and a considerable number of patients regularly use this tool before consulting a physician and during treatment. The web increases the opportunity for patients to acquire information that may be used either to form an opinion on their health status or to reduce anxiety (Barigozzi and Levaggi, 2008). A US survey found that 64% of the online population had searched for health information at least once in the previous 12 months, and a European survey found 71% of Internet users had accessed it for health purposes (Andreassen et al. 2007, Hesse et al. 2005). “Drugs” (requiring prescription or over the counter) was the fifth health topic searched for in the Internet, in 37% of cases (Fox, 2006). Internet has also increased the opportunity for patients to purchase drugs on line (Gallagher 2000). In the nineties, several pharmacies began to operate over the internet selling drugs without virtually any control, even for active principles requiring a prescription. According to the US Food and Drug Administration (FDA) there are at least 400 websites that both dispense and offer prescribing services and half of these sites are located outside the US. As reported on the FDA website, it has been estimated that the number of websites selling prescription drugs may now be closer to 1,000. The number of websites, however, fluctuates from day to day and seems to be growing (FDA 2008a). A recent survey of online pharmacies found that their domain of registration was in 51% of the sample in the USA, and in 22% in Europe (Orizio 2009). It is very difficult estimate the number of people buying online, the volume of drugs traded and the economic extent of this business. A survey conducted in the US by means of telephone interviews found that 4% of Americans had purchased prescription drugs on the Internet (Fox 2004). Back in 2000 the National Association of Boards of Pharmacy estimated that total sales of prescription-only medicines on the Internet in the USA was $100 million, with an estimated increase in all pharmaceutical sales to $1 billion by 2003 (Lorman 2000). Moreover, the web has no geographical barriers and its global dimension makes it difficult to control virtual pharmacies since there is no international legislation regulating this complex issue (Fung et al 2004; Henney 2001; Castronova 2006; Makinen et al 2005; Crawford 2003; Bernath 2003; Brand 2007). The growth of an unregulated global drugs market may increase the risk of the spread of counterfeit medicines, coming from countries where drug production is not subject to the same quality control as in the US, Canada and Europe (Veronin 2004). The World Health 1 Organization has declared that “medicines purchased over the Internet from sites that conceal their physical address are counterfeit in over 50% of cases” (WHO 2006). In addition, many online pharmacies offer the consumer drugs without medical prescription and this can expose people to the risks associated with the intake of inappropriate drugs and can harm the patient-doctor relationship, transforming patients into consumers (Lineberry 2004; Bostwick 2007). In fact, the prescription requirement aims to safeguard patients from risks associated with unnecessary drugs, and it assigns responsibility for cost/benefit assessment to the doctor, the professional who has the knowledge to make this decision. This rapid development of online pharmacies is related to several factors. Internet allows to reach suppliers worldwide at any hour without having to visit shops and it may allow to get commodities at a lower price. For on-line pharmacies there is another important factor related to the controversial theme of direct-to-consumer advertising (DTCA) of prescription drugs, where proponents highlight the potential in terms of patient empowerment, while opponents claim that DTCA distorts the patient-doctor relationship, generating demand without evidence of health benefits (Gilbody 2005; Almasi 2006). In recent years patients have increased their involvement in the decision-making process concerning their health, a process that in the past was entirely in the hands of the physician, leading to general concern about the effect of such strategy on patient health (Donohue 2007). In a world where even the concept of health has changed, shifting from the absence of disease to well-being and wellness, pharmaceutical product promotional strategies may be aimed at creating a demand for drugs, generating consumers/patients. This is a very serious hazard which may be exacerbated by the sale of drugs via the Internet. The sale online of drugs is open to everybody who has an Internet access and online pharmacies sell all types of drugs, including prescription drugs. To our knowledge, the only cost analysis of online drugs was performed in the U.S., in a comparison between traditional pharmacies and OPs which found no economic advantage in buying online, both in year 1999 and in 2006 (Bloom and Iannacone, 2006 and 1999). In our study we investigate the economic features of online drug sales, analyzing marketing and pricing strategies by online pharmacies for some “marker drugs” that were chosen either because of their high intrinsic risk if used inappropriately or because of their widespread use. To our knowledge, this is the first study which focuses on the relationship between OP features and the pricing patterns. Marketing and price strategies are studied using descriptive and econometric tools. 2. METHODS 2 The study was conducted in May 2007 by the University of Brescia, as a collaboration between the Institute of Hygiene, Epidemiology and Public Health and the Department of Economics. It belongs to the Internhe@lth Study, a wider project aimed to investigate the impact of internet on the world of health 1. Using the information of the Internhe@lth database of an analysis of online pharmacies (Orizio 2009) belonging to the Internhe@lth Study, we investigated the selling characteristics of four active principles which we selected as “marker drugs” because of their high intrinsic risk if used without medical control. The risk may derive from an excessive dosage or because of the interactions with other medications the patient has been prescribed, but it can be even due to individual susceptibility. Two of them act on the nervous system: amitriptyline (an old generation antidepressant) and fluoxetine (a new generation antidepressant). Sildenafil is the best-known and most widespread active principle in erectile dysfunction therapy and tramadol is an atypical opioid painkiller. For each active principle we searched for their availability in all the online pharmacies (OP) selling the four drugs a sample (qui specificherei che non è un campione ma sono tutte quelle che sono state trovate) of 100 online pharmacies (OP) that we found selected with the “Google” search engine using as keywords the following combinations: “online pharmacies”, “online drugstore”, “drugs online” and “medicines online”, as described in Orizio (2009). We analyzed their general characteristics using an ad hoc Codebook, according to the Content Analysis method (Riffe 2005) and for each pharmacy we recorded the characteristics that were relevant for our analysis. One of the most controversial areas for e-commerce is what legislation should be applied to these transactions. For this reason, we investigated whether the OPs declared a physical location (state, town, street and number) or whether they presented themselves only as a virtual interface. We recorded data on the type of drug sold (brand or generic) and on the declared shipping location. The first variable should capture the market segment chosen by the online pharmacy while the second may be interpreted as a proxy of the production site, and therefore of the quality of the product.2 To analyze the type of patient targeted by online pharmacies we investigated their prescription requirements. The drugs we are investigating should be prescription only; the decision to by-pass this requirement can be interpreted as a market strategy to reach a group of patients who for various reasons do not wish to use the appropriate protocol. 1 The first result of the Internhe@lth Study have been presented at several Scientific Conferences, as the 6th Interdisciplinary Conference on Communication, Medicine and Ethics “COMET 2008” (Cape Town, South Africa 2-4 July 2008) and the 16th European Conference on Public Health of the European Public Health Association, EUPHA (Lisbon, 6-8 November 2008). 2 According to the most recent literature on the matter, the majority of counterfeit medicines are produced in developing countries such as China, India, Russia and the Philippines (EASSM 2008). 3 We analyzed and recorded their strategies to promote the sale of drugs, through several selling arguments (expediency of buying drugs on-line, bulk discounts, fidelity bonuses). Finally, an analysis of the declared side effects was conducted, analyzing whether the OPs declared at least one side effect; if so, the number of reported side effects was assessed. Most online pharmacies make several quotes for the same active principle, which varies according to the quantity and other characteristics of the offer. In our analysis if an OP was selling different quantities of the same drug we recorded each single offer; in order to compare prices, we selected a specific concentration for each active principle and evaluated the unit cost of the tablet for each offer. In this way, we obtained a complete dataset which records the following characteristics for each offer 3: • Unit price - p • Quantity offered – q Characteristic of the drug o BR: generic (0); brand (1) o SH: shipped from US/Canada/Europe (0); elsewhere/not declared (RW, rest of World) (1) Characteristic of the pharmacy o PH: Physical (1) – Virtual (2) o LR: Declared physical location in US/Canada/Europe (0); RW- not declared (1) Target of patients o PR: Sold without prescription (0); with prescription (1) Selling arguments 4 o list of collateral effects o free trial of other drugs o type of selling arguments Analysis of OP marketing strategies was performed using a descriptive and analytical approach aimed at assessing differences in the selling arguments according to the type of pharmacy, the drug sold and the consumer targeted. The data were analyzed using STATA Software 5. Pricing strategies were analyzed using a model that relies on the literature on hedonic price estimation (Oczowsky, 2001) using SHAZAM Software 6. Although the theoretical background is different 7, the econometric approach to the estimation is very similar since we wish to study which variables are important in price formation using a dataset with an important number of dummy variables. The technical aspects of the estimation method used are presented in appendix one. 3 A description of the variables is shown in table A1 in the Appendix. Given the varieties of the selling arguments, it has not been possible to create specific binary variables for the latter category which has been used only in the descriptive analysis. 5 STATA Statistical Software Release 8.0, 2003; Stata Corporation, College Station, Texas. 6 SHAZAM User Reference Manual, Version 10, North West Econometrics 2004. 7 The aim of this literature is to study the willingness of the consumer to pay for specific characteristics of the product, while in our case we study the pricing and marketing strategies of on-line pharmacies. 4 4 3. RESULTS 3.1 The OP sample: general features and marketing strategies In the sample of 100 OPs, 19 of them ask for a medical prescription to complete the order (prescription online pharmacies, PR); 2 of these specified they accepted only national prescriptions. , while 81 of themOPs do not require a prescription (non-prescription online pharmacies, NPR); among these, around two third were asking to fill in an online medical questionnaire in order to get an online prescription. The main differences between PRs and NPRs are shown in table 1. 56 OPs have only a virtual interface; of the 44 OPs that gave full details of their physical location (country, town, street and number), 26 (59.1%) declare they are located in the United States and Canada, 14 (31.8%) in Europe and 4 elsewhere (9.1%). The physical location is declared in 73.7% of PRs, compared with 37.0% of NPRs with a statistically significant difference (Fisher exact test: P= .005). Only 41 OPs declare their shipment location: 18 OPs (43.9%) in Asia, 11 in the United States (26.8%), 5 in Europe (12.2%), 3 in Canada 3 (7.3%) and 4 in others (9.8%). Selling arguments used by OPs to promote their products include privacy issues, service and drug quality statements, price offers and the suggestion that you can get a drug avoiding a visit to the doctor. For privacy issues, the OPs deal with the use of personal data (92% of OPs) and discreet packaging (59% of OPs). The quality of the service is assured via short delivery times (90%), online tracking of the state of the orders (76%) and testimonials by people who have already bought online (33%). Statements about drug quality are found in 89% of OPs and reassurance that buying on the web is legal in 31% of OPs. As regards selling arguments, 89% of OPs encourage bulk purchases, 64% of OPs advertise lower prices in comparison with “brick and mortar” pharmacies, 62% propose fidelity bonuses, 40% offer free delivery and 9% advise buying specific drugs together with the selected ones, because of special discounts offered or because the two drugs are usually bought in combination. Of the NPRs, 73% of them suggest that is not necessary to undergo a doctor’s examination, avoiding embarrassment and waste of time. The differences in selling arguments between NPRs and PRs concern discreet packaging (69.1% vs 15.8%, Fisher exact test, P < .001), bulk discount (93.8% vs 68.4%, p=0.006), bonuses for repeated purchases (69.1% vs 31.6%, P = .004), free delivery (34.6% vs 63.2%, P = .03) and online order checking (82.7% vs 47.4%, P = .003). 5 Amitriptyline is sold in 72% of OPs, fluoxetine in 82%, sildenafil in 92% and tramadol in 75%. There are no statistically significant differences in the availability of these drugs between PRs and NPRs, with the exception of amitriptyline which is more available in PRs (NPRs 67.9%, PRs 100.0%, p=0.003). About 25% of OPs record no side effects for these drugs; comparing PRs and NPRs, a statistical difference was found only for sildenafil, for which NPRs more frequently declared at least one side effect (p=0.03). We analyzed the distribution of the number of side effects declared for each drug in PRs and NPRs. The number of declared side effects was significantly higher in NPRs for fluoxetine and amitriptyline, whereas for sildenafil it was significantly higher in PRs. (Table one about here) 3.2 Pricing strategies Online pharmacies may differentiate their pricing strategies in two several directions: through the type of drug sold (whether a brand or a generic), through prescription requirement, through the quantity offered and by using a mix of these strategies. In this section we present an overview of the main results of our model and an in-depth analysis of the pricing strategy of each active principle; to assist interpretation we have presented in table one the price differential of each single characteristic of the pharmacy and in figure 1 to 4 the estimated price-quantity schedule. The main results can be summarized as follows: • Firms use mixed marketing strategies (i.e. they differentiate their price both for brand and for customers having a prescription) in selling their drugs with the exception of sildenafil. In the latter case there is a clear-cut strategy for brand and generic drug. The former is usually sold by pharmacies requiring a prescription, the latter by those that do not ask for • this requirement. The prescription has a value on the market. A consumer who wants to buy a drug without prescription has to pay a higher price, unless the active principle is generic sildenafil. In the latter case, the consumer will pay a higher price if he has a prescription. Given the health risk connected with inappropriate use of this active principle, the mark-up to sell the drug • without prescription is important and depends on the type of drug sold. The unit cost of the active principle decreases with quantity; this may be due to decreasing marginal cost, but it may also suggest a strong demand inducement effect (Evans, 1973; Dranove, 1988). For some active principles the firm differentiates the discount according to the type of consumer (with or without prescription) or the type sold (generic or brand). This 6 result is in line with what was observed in analysis of the selling arguments where specific forms of demand inducement include offering “try packages” for other drugs, advising the combined use of several active principles (saying something like “people that bought this • drug often order this one as well”). Active principles that do not originate from the US/Europe/Canada, because the drug is shipped from elsewhere, cost less. The shipping origin may be related to the quality of the drug sold. In some countries the quality standards for producing active principles are fairly slack; it is then possible to produce a drug at a lower price, but of a poor quality (EASSM 2008). (Table two about here) Amitriptyline For amitriptyline, we have recorded 232 offers; the average price is 0,84$ per tablet (min: 0.10 $ max.3.30 $). The average number of tablets offered is equal to 178 (range 30-1200). The price variation is very high in this case. The price is lower for generic and for offers requiring a prescription . For this active principle, online pharmacies discriminate price by type of drug (brand or generic) and by patient (with or without prescription) . Drugs shipped from RW cost less, but the price differential drops if the consumer has a prescription. The characteristic of the online pharmacy, in particular whether they are simply a virtual interface, affects the price, but the price differential mainly depends on the type of drug sold. (FIGURE ONE ABOUT HERE) As regards demand inducement, the discount offered by online pharmacies is presented in figure 1. The price discount is greater for generic than brand and the scheme is more generous for drugs sold with prescription. For amitriptyline, OP strategy seems to be oriented towards inducing the average patient to get the prescription. This policy is reasonable since the drug has a fairly low cost, but a number of side effects when not used appropriately. Fluoxetine For fluoexetine, we have recorded 296 offers; the average price is 2.12 $ per tablet (min: 0.12 $ max.8.21 $). The average number of tablets offered is equal to 97 (range 14-800). The price variation is very high in this case. The price is lower for generic and for offers requiring a prescription . 7 For this active principle the selling strategy is oriented towards differentiation between the type of consumer addressed. If they have a prescription they pay less, but the price differential is higher for generic than brand. The shipping origin is important while other characteristics of the pharmacy seem to be marginal. For fluoxetine, firms offer a discount only to patients that do not have a prescription as shown in figure 2. This means that for this active principle, the higher price charged for not having a prescription is a sort of fixed cost and the price differential drops if the consumer buys a considerable number of tablets. (FIGURE TWO ABOUT HERE) In fact, firms charge a relatively higher price (about 80 cents more per tablet) but they offer up to 25% discount for large quantities. As for amitriptyline, the discount is higher for generic drugs. Tramadol For Tramadol, we have recorded 290 offers; the average price is 1.52$ per tablet (min: 0.18 $ max. 5.13 $). The average number of tablets offered is equal to 90 (range 10-300). The price variation is very high in this case. The price is lower for generic and for offers requiring a prescription . The pricing strategy for selling tramadol is less sophysticated. Firms charge a lower price for generic drugs and to patients with a prescription. The only significant difference in the pricing strategy for brand and drug is represented by a lower price (23 cents) offered by firms shipping from RW to the consumers of brand drugs. The discount offered for bulk purchase is significant, but in the case of tramadol it is relatively higher for customers with a prescription for offers up to 100 tablets; from this point onwards, the pattern is similar to the one observed for other drugs. It is interesting to note that for tramadol the offers are more concentrated. They range from 10 to 300 tablets, a fourth of what is observed for fluoxetine, where the maximum number of tablets offered is 1200. (FIGURE THREE ABOUT HERE) Sildenafil For sildenafil the selling strategy is rather different from those we have described so far. OPs discriminate their strategy according to the type of drug sold. We have recorded 220 offers for the generic product and 321 for the brand one. For generic the average price is 4.21$ per tablet (min: 1.16 $ max.13.5 $). The average number of tablets offered is equal to 58 (range 40-360). For brand the average price required is 15.21 (range 11.5 to 42.7) and the average number offered is 29.5 (range 30-360) For brand having a prescription has not a significant impact on the unit price and the bulk discount is fairly limited (2.5%) as maximum. For generic, the strategy is quite different: OPs promote sales 8 without prescription by making patients pay a higher amount (about 21 cents per tablet) if they have a prescription and by granting a significant discount if a large quantity is purchased. (FIGURE FOUR ABOUT HERE) 4. DISCUSSION AND CONCLUSIONS The analysis presented in the previous section shows that OPs have a precise and sometimes fairly sophisticated strategy for selling their products, very similar to those that can be observed in the market for other commodities. The uncertainty about the legislation to apply to internet transaction, coupled with the lack of transparency that e-commerce may allow, opens up a florid market for OPs which is reflected in their strategies. These strategies differ according to variety (brand or generic), quality, quantity and - what is more worrying from the point of view of the regulator - the type of consumer. In general, OPs are willing to sell a product without prescription, but the consumer will have to pay extra. The OPs interpret this as a sort of fixed cost: via an accurate bulk discount policy, the pharmacy quotes the same unit price for selling (depending on the active principle considered) a large quantity without a prescription or a much lower quantity with a prescription. For the most popular active principle (sildenafil), the use of the generic variety is promoted by offering a lower price to patients that do not have a prescription. The pharmacies proposing such offers often declare a limited number of side effects and promote bulk purchase. In this way they induce the consumer to think that the product sold is a drug that will improve his/her lifestyle without many side effects. Such behavior should worry regulators because it means that OPs induce demand through their pricing strategies. Furthermore, today’s adults, a population familiar with internet and e-commerce, will be the elderly of the near future. Since the elderly are those who use drugs most, we can expect the potential market reached by Ops to expand. Information and health services are already freely available on the web and an increasing number of patients use this tool. The web may offer opportunities to improve health care, but it may also represent a big health hazard. This market is basically unregulated and offers very low consumer protection. For health care this a serious problem because in this market, more than for other commodities, the consumer is not able to detect the quality of the product sold. The analysis presented in this paper shows that OPs differentiate their strategies for active principles and target patients, but as a whole they act roughly in the same way, i.e. “rough sites” copycat the pricing strategies of the genuine ones, so that price is not a quality indicator. OPs can affect national health care systems, potentially leading to unequal access to health care and to increased health hazards. For appropriate prescriptions, the development of on-line sales, 9 especially if coped with a parallel reduction of conventional pharmacies may reduce access to drugs because not everybody has access to internet. On the other hand the sale of drugs without prescription and control may endanger patients. The appearance of a “global free drugs market” could seriously endanger the internal equilibrium of individual countries’ public welfare and health systems, which is why the European Union’s Court of Justice has stated that “Article 30 EC may be relied on to justify a national prohibition on the sale by mail order of medicinal products which may be sold only in pharmacies in the Member State concerned in so far as the prohibition covers medicinal products subject to prescription. However, Article 30 EC cannot be relied on to justify an absolute prohibition on the sale by mail order of medicinal products which are not subject to prescription in the Member State concerned” (EU 2003). The Court’s sentence was issued in connection with the case of Doc Morris, a legal Dutch OP trading mostly in the Netherlands and Germany, which had been taken to the European Court of Justice by the regional court of Frankfurt after the accusation of illegal practice by the German Association of Pharmacists who opposed foreign competition in a trade regulated at national level (Makinen 2005). The main limitation of the present study is not having attempted to buy the drugs on offer, assessing their effective shipping to consumers. This is the next step we intend to perform. The results found in this analysis regarding the marketing and pricing strategies of OPs confirm the concern expressed by the FDA, which has defined OPs a “public health risk”, with huge potential effects both at individual level and at the level of national health systems (FDA 2008b). Hence there appears to be an urgent need for international regulation on this issue, aimed at preventing risks and controlling the phenomenon, both by increasing control and awareness of the risks associated with buying drugs on the Internet, implementing public health policies and reinforcing the trust-based relationship between patient and General Practitioner. The latter policy may prove to be even more effective. The regulation of a global market is very difficult: it requires a coordinated effort by all the nations and even in this case, its implementation may be too expensive. On the other hand the increase of patients’ awareness as concerning the risk of getting drugs without a proper prescription may in the long run be the winning factor to limit the size of on-line trade to appropriate prescriptions. 10 REFERENCES Almasi EA, Stafford RS, Kravitz RL, Mansfield PR. What are the public health effects of direct-toconsumer drug advertising? PLoS Medicine 2006; March, 3(3): e145. PMID:16076787 Andreassen HK, Bujnowska-Fedak MM, Chronaki CE, et al. European citizens' use of E-health services: A study of seven countries. BMC Public Health 2007, 7(147):53. Barigozzi, F. e Levaggi, R. Emotions in Physician Agency (2008) Health Policy, forthcoming Bernath P. Regulation of online pharmacy:an Australian perspective. J Law Med 2003; 10: 339-63. Bessel TL, Silagy CA, Anderson JN, Hiller JE, Sansom LN. Quality of global e-pharmacies: can we safeguard consumers? Eur J Clin Pharmacol 2002; 58: 567-572. Bloom BS, Iannacone RC Changing availability and cost of internet physician consultation and prescription medications. Med Inform Internet Med. 2006; 31: 247-53. Bloom BS, Iannacone RC Internet availability of prescription pharmaceuticals to the public. Ann Int Med. 1999; 131: 830-833. Bostwick JM, Lineberry TW. Do cheap internet drugs threaten the safety of the doctor-patient relationship? Expert Opin Drug Saf. 2007; 6: 9-13. Brand R. Drugs: just a click away. Online pharmacies con make dangerous drugs easy to get, but also can promote better health care. Should we regulate them? State Legis. 2007; 33: 45, 47, 49. Castronova JR. Operation cyber chase and other agency efforts to control internet drug trafficking. The “virtual” enforcement initiative is virtually useless. J Leg Med 2006; 27: 207-24. Crawford SY. Internet pharmacy: issues of access, quality, costs, and regulation. J Med Syst 2003; 27: 57-65. Donohue JM, Cevasco M, Rosenthal MB. A decade of direct-to-consumer advertising of prescription drugs. N Engl J Med 2007;357:673-81. Dranove, D. (1988), "Demand inducement an the physician-patient relationship", Economic Inquiry, 24: 282-298. EASSM, European Alliance for Access to Safe Medicines, The Counterfeiting Superhighway. 2008, Meidcom Group Ltd, UK. Available from: www.eaasm.eu. EU Court of Justice judgement of 11/12/2003, case C-322/1 Deutscher Apothekerverband versus 0800 Doc Morris NV. Available from: http://eur-lex.europa.eu. Accessed 10/09/2007. Evans, R. G. (1974), "Supplier- induced demand: some empirical evidences" in Perlman, M. (ed.), The economics of Health and Medical Care, New York, John Wiley. FDA 2008a, U.S. Food and Drug Administration website, www.fda.gov, at the page: http://www.fda.gov/oc/buyonline/faqs.html#faqs3. Last accessed 29-09-2008. FDA 2008b, U.S. Food and Drug Administration website, www.fda.gov, at the page: http://www.fda.gov/oc/buyonline/faqs.html#faqs4. Last accessed 29-09-2008. Fox S. Online Health Search 2006. Pew Internet and American Life Project. www.pewinternet.org, October 2006 (last accessed 03-09-2008). Fox S. Prescription drugs online. Pew Internet and American Life Project. Oct 2004. Available from: www.pewinternet.org. Fung CH, Woo HE, Asch SM. Controversies and legal issues of prescribing and dispensidng medications using the internet. Mayo Clin Proc 2004; 79: 188-94. Gallagher J, Collaizi J. Issues in Internet pharmacies practise. Ann Pharmacother 2000; 34: 1438-5. Gilbody S, Wilson P, Watt I. Benefits and harms of direct to consumer advertising: a systematic review. Qual Saf Health Care 2005; 14: 246-250. PMID:16076787. Halvorsen, R., & Palmquist, R. (1980). The interpretation of dummy variables in semilogarithmic equations. The American Economic Review, 70, 474-475. Henney JE. Cyberpharmacies and the role of the US Food and Drug Administration. J Med Internet Res 2001; 3(1): e3. 11 Hesse BW, Nelson DE, Kreps GL, et al. Trust and Sources of Health Information. Arch Intern Med 2005;165(22):2618-24. Landon, S. and Smith, C. E. (1997) ‘The Use of Quality and Reputation Indicators by Consumers: The Case of Bordeaux Wine’, Journal of Consumer Policy 20, 289–323. Lineberry TW, Bostwick JM. Taking the physician out of “physician shopping”: a case series of clinical problems associated with internet purchases of medication. Mayo clin Proc 2004; 79: 10314. Lorman AJ. Internet pharmacies catch on with consumers; pose problems for the regulators. Health Care Advisory, 2000 February. Available from: http://mints.com. Makinen MM, Rautava PT, Forsstrom JJ. Do OPs fit European internal markets? Health Policy 2005; 72: 245-252. Oczkowski, E. (2001) ‘Hedonic Wine Price Functions and Measurement Error’, Economic Record 77, 374–82. Orizio G, Schulz PJ, Domenighini S, Caimi L, Rosati C, Rubinelli S, Gelatti U. Cyberdrugs: a cross sectional study of online pharmacies characteristics. The European Journal of Public Health, 2009 Jan 28, doi: 10.1093/eurpub/ckn146 Riffe D, Lacy S, Fico FG. Analyzing Media Messages. Using Quantitative Content Analysis in Research. Mahwah, New Jersey: Lawrence Erlbaum; 2005 Schamel G., Anderson K. (2003), “Wine Quality and varietal, regional and winery reputations: hedonic prices for Australia and New Zealand”, The Economic Record, 79 (246): 357–369 Steiner, B. (2004) Valuing labelling attributes with hedonic price analysis: Australian wines in the British wine market , Agribusiness: An international journal, 20(3): 287-307 Veronin MA, Youan BC. Magic bullet gone astray: medications and the internet. Science 2004; 305: 481. WHO, Fact sheet n° 275, revised 14 November 2006, available from: http://www.who.int/mediacentre/factsheets/fs275/en/index.html. 12 Formatted: English (U.S.) Formatted: Normal, Don't adjust space between Latin and Asian text, Don't adjust space between Asian text and numbers Formatted: Font: Not I talic, English (U.S.) Formatted: English (U.S.) Formatted: Font: I talic Formatted: Font: Not I talic Table 1: Percentage and statistical significance of the main characteristics of PRs and NPRs. P (Fisher PRs (n=19) NPRs (n=81) exact test) Declaration of physical location 73.7% 37.0% .05 % % Selling arguments: declaration n.s. 90.1 100.0 - Safe use of personal data <.001 15.8 - Discreet packaging 69.1 n.s. 91.4 84.2 - Short delivery times .003 47.4 - Online order checking 82.7 n.s. 37.0 15.3 - Testimonials n.s. 91.4 78.5 - Drug quality n.s. 32.1 26.3 - Reassurance that buying online is legal .006 68.4 - Lower price the more you buy 93.8 n.s. 65.4 57.9 - Lower price than in the traditional pharm. .004 31.6 - Affiliation program 69.1 .03 34.6 - Free delivery 63.2 n.s. 9.9 5.3 - Advice to buy another drug / 72.8 / - Suggestion that you get prescription drugs without prescription Number of OPs selling each drug % % - Amitriptyline 67.9 .003 100 - Fluoxetine 94.7 86.4 n.s. - Sildenafil 94.7 96.3 n.s. - Tramadol 89.5 81.5 n.s. Number of side effects declared* Median Median - Amitriptyline 49 .005 61 - Fluoxetine 25 .007 41 - Sildenafil 11 .008 21 - Tramdol 14 22 n.s. * Evaluated only for the OPs which declared side effects n.s.: not statistically significant 13 Table 2: Effects of binary variables on the price (UD$) 8 Amtriplytine Fluoextine Tramadol Sildenafil Base price Brand With Prescription generic With Prescription brand Shipping from RW Shipping from RW with prescription Shipping from RW brand Virtual pharmacy –Generic Virtual pharmacy –Brand Real pharmacy- RW 0.97 0.62 -0.70 -0.70 -0.49 -0.27 2.25 1.89 -0.89 -0.75 -0.37 0.28 0.11 0.18 1.23 1.63 1.24 -0.49 -0.49 Generic Brand 4.52 24.53 0.42 -0.25 -0.23 -0.14 Notes: Base price is the average price of a tablet containing generic active principle sold without prescription by a pharmacy that has its legal residence in US/CAN/EUR and ships the product from one of these locations. RW is “rest of the world”, meaning everywhere not in US/CAN/EUR or in place not declared.AP 8 Halvorsen and Palmquist (1980) and Landon and Smith (1997) explain how to interpret the coefficient of the dummy variable in terms of price. The presence of the factor represented by the dummy variable is equal to g = exp(c) – 1 where c is the estimated coefficient. When two dummies interact (for example having a prescription and selling a brand) the effect will be equal to g = exp(c+d) – 1. 14