1. Paying for Pharmaceutical Regulation Warren Kaplan Richard Laing Boston University School of Public Health http://heapol.oupjournals.org/cgi/reprint/18/3/237.pdf

2. ABSTRACT We analysed DRA new drug registration fees for 34 countries Many DRA registration fees for new drug applications for developing/non- OECD countries are less than the current GNP/capita of that country. For many countries new drug application fees are between 1–5 times GNP per capita or between $17 000 and $80 000 for each $1000 spent per capita on healthcare. There is little relationship between DRA registration fees and drug approval times in developing countries. Drug regulatory authority user fees are an important policy instrument to improve access to medicines but for the most part, they are underutilized.

3. The financial sustainability of a DRA is critical to implementing its regulatory functions but financial sustainability of a government DRA DRAs can collect registration fees and various other charges for the services they provide. There are 50 registration fees for the United Kingdom DRA and over 90 for Australia. These DRAs rely entirely on registration fees for their support Policy tradeoffs are important. Differential registration fees, presumably designed to encourage locally produced versus imported products, may violate international trade regulations. Moreover, certain DRA registration fees may provide perverse incentives for the pharmaceutical industry. Developing countries should require that DRA registration fees be based on accurate accounting of the cost of services provided. At present levels, these fees could probably be increased without disincentive to the pharmaceutical industry. BACKGROUND AND SETTING

4. Study Aims  ON WHAT BASIS DO DRAs SET THEIR REGULATORY FEE STRUCTURE? ARE THEY REALLY DESIGNED AS COST RECOVERY TOOLS?  Analyze the relationship between aggregate economic indictors and DRA drug registration fees in developing and some developed countries  Provide recommendations with regard to appropriate fee levels, differential fees for generics, and the policy consequences of registration fees for improving DRA effectiveness and improving access to generic drugs.

5. Methods (II) Based on a questionaire sent to 71 DRAs and from the literature, we analyzed several indices for new drug registration fees.  Ratio of the DRA registration fee to the GNP per capita (GNP/capita: World Bank, calculated by the Atlas method)  Ratio of DRA registration fee to total amount per capita spent on healthcare (WHO 2000: data in international dollars) to new drug registration fee  Ratio of DRA registration fee to total pharmaceutical market size for the various countries (in retail value; over-the- counter plus prescription sales). (Market size information came from the US Department of Commerce Commercial Service (United States Department of Commerce 2002) and the IMS-HEALTH websites (IMS-HEALTH 2002)).

6. GNP/capita, total healthcare/capita and DRA fees for selected countries GNP/capita Total healthcare/capita Initial registration fee (US$) (2002) Fee index (US$) (A) US$ (B) NEW DRUG GENERIC (B/A) Algeria1560122641038504.1 Argentina74608231200040001.6 Australia 20240160112650025006.3 Bulgaria 152019311249370.74 Costa Rica 3810489500500 0.13 Cuba 8001097007000.91 Denmark 320201940840681420.26 Japan 35620175966676448641.91 Malaysia 33802021001000.03 Singapore 247407505002200.02 South Africa 30203962873975–12320.95 Uganda 300443003001 UK24430119310293326804.2 USA34100372430964700 9.1 Zimbabwe46064100010002.2 Notes: South Africa: range is for generics on EML ($975) and not on EML ($1232) United States: No registration fees for generic products

7. RESULTS (II)  Registration fees for new chemical entities are often higher than those for generic products.  Median approval times of Canada, Australia, Europe and the United States for non-fast track procedures tend to converge on 15 to 20 months. Japan is the exception.  Data is limited for developing countries but average approval times are often nearly twice as fast major market countries  The limited data for developing countries suggests little or no relationship between the magnitude of the new drug registration fees and DRA approval time

8. RESULTS (I)  DRAs partly or completely financed via full cost recovery/user fees: United States and the United Kingdom, the DRA fee for an initial new drug filing is between 4 and 10 times the GNP/capita, for Australia and Zimbabwe it is about 6 and 2, respectively.  For all other countries, most new drug DRA fees are between only 0.21 and 2.1 of GNP/capita.  DRAs partly or completely financed via full cost recovery/user fees: Between $17500 and $80000 new drug fee for every $1000 spend on healthcare per capita  For non OECD countries, $5283 fee for each $1000 per capita healthcare spent  For OECD countries, about $10000 fee for each $1000 per capita healthcare spent

9. Results (III) Range of DRA new drug fees by national health/economic indicator IndicatorCountryScaling factor Gross National Product OECD countries Median fee = 1.25 times GNP per capita Per capita Interquartile range: 0.4–2.4 times GNP Non OECD CountriesMedian fee = 0.45 times GNP per capita Interquartile range: 0.07–1.0 times GNP Total healthcare expenses Per capita OECD CountriesMedian fee: $10 983 for each $1000 Interquartile range: $4372–37 905 Non OECD CountriesMedian fee: $5283 for each $1000 per capita Interquartile range: $760–11 029 Total National Pharmaceutical All countriesFee: $1000 for each $384 million of national market Market (Sales) Excluding $1B MarketsFee: $1000 for each $85 million of national market Notes: OECD Countries: Australia, Czech Republic, Denmark, Japan, Netherlands, New Zealand, Poland, Portugal, Slovak Republic, Sweden, United Kingdom, United States. Billion dollar markets: Argentina, Australia, Denmark, India, Ireland, Netherlands, Poland, Portugal, Singapore, Sweden, United Kingdom, United States, Venezuela.

11. Summary  DRAs of developing countries charge lower new drug registration fees in relationship to their per capita national income than do developed countries  No empirical evidence that DRA new drug registration fees are actually set by policy according to either the GNP per capita, the total healthcare expenditure per capita, the size of the pharmaceutical market – or in fact any other national level economic indicator

12. Discussion  Higher drug registration fees could easily be charged by developing country DRAs. This would not unduly increase costs for either the pharmaceutical industry, the government or the consumer.  If new and higher fees were to be charged, it is critical that these additional financial resources go back to the DRA to improve quality of services. In cases where the generated revenue ends up in the central government account, it is far more difficult to make a case for increased fees.  On balance, increased fees (from whatever source) must provide the DRA with sufficient resources to retain competent staff and pay outside consultants and experts to meet DRA performance standards. Without more staff, there is likely to be an increased workload which causes high staff turnover. Poor staff retention can also result in a loss of critical ‘institutional memory’.  Differential (i.e. lower) registration fees for generic drugs would encourage most manufacturers to submit registration dossiers as the cost would be minimal. It is possible that such fees may also encourage low quality manufacturers as well, who might submit applications haphazardly.

13. Conclusions and Recommendations  For most developing countries DRA registration fees probably bear little relationship to the true drug regulatory costs required for the drug approval process.  Promoting generics through differential, reduced fees should be encouraged in developing countries as a matter of public policy.  To justify increased fees, stakeholders must be able to document improvements in service and cost efficiencies.  Key DRA policy discussions should focus on the magnitude of the registration fees and on the balance between the registration fee portion and the budgeted or appropriated portion to avoid the DRA becoming entirely dependent upon the industry it is meant to regulate.