Articles

Resources

A dictionary of words and phrases whcih are crucial to the world of clinical trials.

CDISC Glossary: Clinical Trials Terminology

Somewhere in a closed conference room, the IT wizards are plotting the data strategy of your company or institution. They are wrestling with the vast amount of raw data that is being collected and the challenges of making that data available as information to help in making decisions. After all, what is the point of collecting all of that data if you aren't going to use it?	

Clinical development organizations are expert at taking the raw data from a clinical trial and turning it into highly refined data listings and data tables, which are ideally used to support a drug's approval by the regulatory authorities. That is, after all, one of the main functions of a pharmaceutical company. However, the greater value of clinical data can only be unleashed by combining it together with other data for analysis and exploration. The hundreds of clinical trials, involving tens of thousands of patients, contain a treasure trove of information about disease and wellness, and the effects of pharmacological therapy. Now, if only this data could be easily aggregated and analyzed. 	

The data strategy being plotted by your IT staff hopefully goes far beyond mere clinical data. There is a vast amount of data collected in many disparate systems that, when linked together, can be invaluable in understanding and planning the entire process of drug development, not to mention making the process more efficient, increasing quality, and decreasing cost. 	

This data comes from clinical data systems, financial systems, clinical trial management systems, randomization systems, project planning systems, laboratory information systems, and even human resource systems. When taken as a whole, the data from these systems can be used to answer questions about the quality, efficiency, and cost of the clinical development process. 	

The information can be used to support strategic and tactical changes to business operations, to monitor the effects of current or changing processes, and to identify new ways of performing clinical development and clinical trials.	

Our data strategists will eventually leave their closed room and present their strategy to us. A barrel-shaped drawing labeled "Data Warehouse" might be in the middle of a complicated diagram. This, they will tell us, will be the way that we can combine all of that data from all of the different systems. Through the data warehouse we will be able to extract the aforementioned gold mine of information-the analytic and summary data that helps us make and assess the tactics and strategies of our company and its processes. 	

It is a grand vision, but will it work? There is no good answer to this question beyond "it depends." Certainly, though, its success is likelier if the end users of the data warehouse understand the purpose and limitations of such a project, provide knowledgeable input over the two to three years of its design and construction, and continue the input during the ongoing process of revision. To participate in this process, it is helpful to understand a little bit about databases in general and database design in particular.	

Most of us are unaware of the many databases we interact with on a daily basis. Everything from buying a product in a retail store to reading a Web page or our email involves hundreds of database transactions and database reporting. The marvel of modern computing is that the databases are so isolated from the end users' experiences that we're typically not even aware of their existence.				

When those of us who are not in a data management role speak of a database, we often envision a big spreadsheet or table. The address list for a party, for example, can be created where each row is a person and each column contains one of their attributes, such as name, street address, town, zip code, and phone number. This type of database, known as a "flat file" database, is very valuable and useful for specific purposes. Databases of this sort can contain thousands of rows and many columns. In skilled hands, very useful analytic pivot tables and reports can be produced from flat file databases, which can often provide important business information.	

There are, unfortunately, many limitations of flat file databases. First, they very quickly get too large to manipulate practically. Searching, sorting, or reporting a large flat file database can take a very long time. When the data being collected are more complex, the flat file database becomes untenable. For example, if you were the manager of a large shopping center, you might want to keep a database of stores, including the category of store, the name, phone number, and address. It is easy to capture this in a simple row and column database-each row would represent a store, and columns would contain name, address, phone, and type of store. Unfortunately, you would be re-using a particular store category for several different stores (e.g. bookstore), increasing the potential for errors (like a "bookstove").	

Now, imagine that you also wanted to track employees and information about sales events. Do you add a new column for each employee, labeled Employee1, Employee2, and so on? Obviously, this would be very unwieldy. Some stores have many, many employees; others have only a few. There would be many blanks in the table. In addition, such a database would offer no way to track employee information such as addresses and birthdates.	

Another approach might be to make a repeat row for each employee. This would create many rows for some companies. If you now wanted to collect information on sales events, you would have to have a separate row for each of these, for each employee. The number of rows would rapidly expand to an unreasonable number. In addition, this approach would create a lot of repeated information. Again, the same issue of repeated typing of locations or companies comes up, and the possibility of mistyping, errors, and other data corruption.	

The solution for this dilemma is known as the relational database. In essence, a relational database is a series of database tables designed to maximize the efficiency of the database for a particular task. 				

Dr. E.F. Codd first developed the relational data model in 1969 in an IBM research report. He continued working on the development of relational databases until the mid-'80s, when he published a list of 12 rules that concisely define an ideal relational database. These rules have provided a guideline for the design of all relational database systems ever since. There are certainly many other types of transactional database models in use today, but the vast majority of them are relational.	

To understand the relational database, consider our new store database, which might have separate flat file databases called tables for Store, Employee, Sale, and Category. In this relational database, each table has a unique name or "primary key" that defines each entry (for example, Store1, Store2, Store3). A table can be "related" to another table by linking to the other table's primary key as a column-these links are called "foreign keys." For example, the Employee table would have primary keys of Employee1, Employee2, Employee3, with names and addresses for each listing. The Employee table would also have a column for StoreID as the foreign key, which would correspond to the store that employs them. Since one store has many employees, but each employee only works at one store, the two tables are related by the "one to many" relationship. This description is highly simplified, but describes the essence of a relational database. 	

Data entry screens may match the particular tables in the database, making data entry very efficient. A modern database management system (DBMS) can also use roll-back functions that ensure that a multifield data entry transaction is entirely completed. Without this, partial data entry from a particular transaction can lead to corrupted databases and lost data. 	

Data retrieval from a transaction-oriented relational database can be complex, involving the careful "joining" of two or more tables. Fortunately, there is a common language (alas, with many dialects) that is used for adding, retrieving and manipulating data in databases. The language is known as SQL, for Structured Query Language. Learn SQL and you can get data from pretty much any professional database.	

The power of the relational database is many-fold. First, a properly designed database should only require the entry of particular data one time. This prevents errors from redundant data entry. In addition, the relational database is set up to maintain referential integrity. If a store is deleted from the database, all the linked tables that reference that store can also be removed. Finally, the relational database is designed to perform efficiently as part of the day-to-day operations of a business, enabling data entry and data reporting.	

As mentioned, Dr. Codd described a series of rules for "normalization" that are used to optimize databases. The rules formalize the process of eliminating redundancy in data, and ensure the presence of primary keys and dependency of all data on these keys. A database that is normalized to the "third normal form (3NF)" is considered to have a good design. 3NF databases typically have little or no repeated data. 	

For example, in our database many stores share the same address.  Therefore, all the stores at the same address would have the same "AddressID" entry. A separate Address table would contain a single instance of each address. In this way, the address wouldn't be repeated over and over again in the Store table. Practical considerations sometimes force deviations from this, but most good transactional databases have a 3NF form.	

Once we have all the data entered into a relational database, how do we get the data out? SQL queries with proper table joins will generate views of the entered data, or even nicely formatted and summarized reviews of the dataset. Most operational databases have a variety of prewritten and often flexible reports that allow users to view data. 	

However, consider a theoretical database of this sort with data on every shop in every mall in the country, or even the world. Imagine this database also contains tables with the sales and expenses for each store, for each day, for years. 	

The size of such a database would be staggering. A query that summarizes the quarterly sales and expenses for each of the different store categories, and allows viewing of the data for each of the different regions of the country and world, could require an immense amount of computing power to search through the tables, find and summarize the data, and report it in summary form. The relational database is superb for complex transactional data processes, but is quite limited in complex data reporting and analysis.	

Sophisticated reporting and analysis is where the data warehouse comes in. The data warehouses (and the smaller, more targeted data marts) are database designs, created to answer questions about business processes and results using massive amounts of data.				

Many people think of data warehouses as a "repository" for all of your business data, available for browsing or query development, but this is misleading and typically incorrect. Rather, it is best to think of a data warehouse as a database built to answer specific and frequently asked questions about business processes and results. The data warehouse typically involves the collection of data from many disparate transactional databases involved in the conduct of business.	

From the beginning, the data warehouse is designed with a very significant understanding of the questions that it is going to be asked. The table structure is typically constructed in a distinct "star" form, or data cube. A complex query in a relational database might require many table joins and hence very poor performance. 	

The star schema allows this query to be accomplished efficiently and quickly. This is because the tables in the data warehouse are purposefully "denormalized," and usually contain much repeated data, violating Dr. Codd's rules.  	

An example might be useful. In our theoretical, international database of shopping malls we will construct a data mart, which is a small, purpose-built data warehouse, for answering our question about sales and expenses at different store categories summarized by region of the world. The center of our "star" is a "fact table" which represents the business process that is being measured. The fact table contains a primary key (fact1, fact2, fact3), the core data facts, and secondary keys for a series of surrounding "dimension" tables, thus the star design. The core data facts of daily sales or units sold are typically numeric values that can be manipulated arithmetically (summed, multiplied, and divided). 	

In our example, the fact table might contain a FactID, a quarterly sales amount, a quarterly expense amount, and secondary IDs that serve as links to a store table, and a category table, and others. While the center of the star, the fact table, is normalized, each of the dimension tables are not. 	

A dimension table for Store, for example, might include store name, postal code, city, state, region, country, and continent. This table would have many repeat values in the columns; the state may be repeated in hundreds or thousands of rows. The denormalized form of the dimension tables is purposely created to allow rapid and efficient SQL queries. A nontechnical user can quickly see each attribute of a dimension table can be used to "sum up" the data. In our store table, we can look at quarterly sales and expenses by store, postal code, country, region, state or city. No complex joining of tables needs to happen-the design of the database is created to support this. 	

Similarly, a "category" dimension table may contain category (dress store, bookstore), product (hardcover, paperback, blouse), and other factors. Again, sales can be summed by product, category, and other factors. 	

Several important things are worth noting about data warehouses. First, the data are drawn from the transactional databases of a company. They typically don't have to be "real time" so they are often uploaded, in bulk, on a regular daily or weekly basis. The work of cleaning, aggregating, and manipulating the data for uploading to a data warehouse can be significant.	

To use a simple example, the combining of two databases that use different terminologies for male and female (M and F versus 1 and 2) can require modification of all the records in one of the databases. When this must be done for many different fields, it can become a very large task. In some cases the adjustments can be automated, but they always requires a lot of thought and planning. 	

Since the data are moved from an existing database, they don't typically have to be updated individually, can have repeating identical fields, and don't have to have "referential integrity" built in. A data warehouse is NOT a data repository-it is a special purpose analysis database designed for efficient reporting. 	

Nevertheless, the types of queries and reports that can be derived from a data warehouse depend a great deal on the planning and design of the database before it is built or populated. The structure of the tables determines the types of reports that can be generated, and so must be designed to contain the appropriate data. For efficiency, this data is very often aggregated-monthly updates in a data warehouse may be combined from individual data transactions before it gets sent. Alternately, a data warehouse may have fact tables for weekly, monthly, and quarterly facts, each calculated from a single daily fact table. In either case, the aggregated dimension table makes it possible to get rapid access to data without excessive processing or database queries.	

If implemented and managed well, a data warehouse can provide very important information for smarter decisions from concise understanding of business processes. In clinical data, a data mart for a particular question about a particular therapeutic area may use the wealth of data available to better understand a drug or drug class. However, those who plan to make use of this data must understand that the data may not be as clean and ordered as is typical in a clinical database.	

If you are asked to participate in a data warehouse project, make sure you understand what you are getting into. The process is going to take several years to create and implement. The project will certainly expand beyond your initial expectations, and will involve a substantial amount of data cleaning and quality control of data sources. 	

Finally, your end users need to be properly trained and motivated to use the data warehouse. It is quite common for data "consumers" to be uncomfortable with digging into a data warehouse. 	

Without knowledgeable and tenacious data consumers, the data warehouse won't be nearly as valuable. However, with carefully defined goals and a committed end-user community, it can provide invaluable information and knowledge from a previously untapped resource-your data. 	

Technology Viewpoint

Somewhere in a closed conference room, the IT wizards are plotting the data strategy of your company or institution. They are wrestling with the vast amount of raw data that is being collected and the challenges of making that data available as information to help in making decisions.

Data Warehouses: Your Company's Future in the "Stars"

A full listing of college and university courses, training classes, and other educational opportunities.

Education and Training Directory

This past year has been one of tremendous activity at the Food and Drug Administration, and at other government agencies involved with healthcare, biomedical research, and national security. Scientists completed sequencing the human genome, generating optimism for a period of unprecedented advances in biomedical discovery. Congress completed a five-year effort to double funding for the National Institutes of Health (NIH) to more than $27 billion, prompting that agency to propose major changes in how it funds and encourages innovative biomedical research [see ACT View from Washington, November 2003]. The need to protect the nation from biological warfare also promises to stimulate development of new vaccines and medical treatments [see sidebar]. 	

At the same time, public demand for more affordable medicines has shaped a wide range of policy initiatives. Congress tackled the daunting task of crafting a prescription drug benefit for the nation's 40 million+ Medicare beneficiaries. FDA has responded with regulatory and management initiatives to lower drug costs, while also challenging efforts to import less expensive drugs from foreign sources.

Although out of the limelight, government officials continued efforts to improve oversight of clinical research. Work began on a process to register and obtain more information on institutional review boards. Academic and healthcare organizations began to sign up for voluntary accreditation of their human research protection programs. Congressional leaders worked on legislation to oversee conflicts of interest, international research, and other related issues, but failed to reach consensus. One milestone announced in October is that 1,000 privately sponsored clinical studies have been posted on the ClinicalTrials.gov web site, expanding the listings to more than 8,000 clinical trials.	

In his first year in office, FDA commissioner Mark McClellan unveiled a steady stream of initiatives designed to streamline FDA internal regulatory procedures and to clarify requirements for testing new therapies. A prime goal has been to reverse the recent slow-down in applications for important new drugs. McClellan emphasizes the importance of applying risk management strategies to FDA policies and programs, a concept that he explained further in a Strategic Action Plan, issued in August. Risk management involves targeting FDA's limited resources to oversight of higher risk products and processes. Initiatives that reduce the time, cost, and uncertainty of developing new therapies can improve public health, according to McClellan, by bringing new products to market more efficiently and thus facilitating consumer access to effective new treatments. 				

Accelerate application reviews. Last January the commissioner unveiled a plan for "Improving Innovation in Medical Technology." It seeks to clarify regulatory standards and test data requirements, and to process more applications in only one review cycle. 

Focus on chronic diseases. The agency is developing more guidances on requirements for developing new treatments for diabetes, cancer, and obesity-top priority disease areas for government health agencies. 

Improve clinical trial design and methods. FDA is encouraging development of new methods for characterizing novel products and for assessing drug toxicity and immune reactions [speech Oct. 9, 2003 at the Phases conference in Washington, D.C.]. One idea is to reduce the need for some animal toxicity testing by refining methods for using tissue cultures and cellular assays that can better predict drug safety.

Streamline manufacturing processes. FDA officials continue to implement a two-year campaign to overhaul FDA regulations governing good manufacturing practices (GMPs). An important component involves revising policies governing electronic records and signatures, which are important for clinical trial operations.

Ensure medication safety. FDA held a three-day workshop in April to examine how risk management approaches can promote the safe use of medical products. Agency officials, manufacturers, and health professionals commented on FDA proposals for more extensive pre-approval testing of new products to better identify and study safety concerns. Participants also discussed how manufacturers could develop pharmacovigilance plans to detect and analyze safety issues that appear after a product reaches the market. FDA wants sponsors to do more to identify risk "signals" and to develop post-marketing risk management programs for drugs that pose increased safety concerns for patients. The agency is finalizing rules on these issues, and implementation will be a major challenge in coming months. 

Consolidate oversight of drugs and biologics. FDA transferred regulation of biotech therapies from the Center for Biologics Evaluation and Research (CBER) to the Center for Drug Evaluation and Research (CDER) to gain more uniformity in oversight programs. The consolidation became official Oct. 1, and has involved shifting to CDER some 200 staffers who evaluate applications for monoclonal antibodies, thrombolytics, cytokines, and growth factors. CBER staffers who review clinical data in applications formed a new office in CDER's Office of New Drugs, while CBER researchers and manufacturing experts created the Office of Biotechnology Products. 

A key concern of FDA staffers is whether significant budget cuts for 2004 will support these initiatives, many involving extensive use of personnel and resources. CDER plans to hire fewer additional staffers under its drug user fee program, and anticipates cuts in information technology personnel and research programs. This could create difficulties for plans to integrate fairly different CBER and CDER information and electronic submission systems. 								

Meanwhile, CDER leadership is shifting, at least temporarily. In October, CDER director Janet Woodcock started a six-month detail to McClellan's office to help implement "cross-cutting initiatives" outlined in FDA's strategic plan. Woodcock has headed CDER since 1994, but spent a few months in the commissioner's office to help develop FDA counter-terrorism plans following the 9/11 attack. She has played a lead role in overseeing the CDER/CBER consolidation, in developing risk management proposals, and in launching FDA's GMP overhaul. 	

CDER deputy director Steve Galson is heading the Center temporarily, assisted by Mark Goldberger, who moved up from the Office of New Drugs to be acting deputy director. The change gives Woodcock a break from the steady grind of managing a growing operation squeezed by serious budget constraints and provides her new opportunities to address broader policy issues, including biodefense initiatives and applying pharmacogenomics to drug development.	

While seeking to spur new product development, FDA also is fighting the campaign to allow reimporting of low-cost but unapproved drugs and biologics from Canada and other nations. McClellan insists that reimporting can bring unsafe counterfeit products into the United States, particularly through purchases via Internet pharmacies, many operating overseas. The debate has heated up during the past year, with the House approving a bill in July that permits drug reimporting-without any requirement that FDA first certify that such activity will not harm public health. 				

In September the Department of Justice filed suit against a leading pharmacy Web site, Rx Depot, for arranging drug imports from Canada. However, several states and local governments devised plans to buy Canadian medicines in order to reduce outlays on prescription drugs for Medicaid and local pharmacy benefit programs. The public generally believes that drugs and biologics from Canada are just as safe and effective as those available in the United States.	

FDA and manufacturers fear that illegal reimporting opens the door for counterfeit and unsafe products to enter the U.S. market. FDA sponsored a well-attended meeting 15 Oct., 2003 to discuss proposals for preventing counterfeit products from slipping into the U.S. drug distribution system. The agency proposes to attack counterfeiting through a mix of stricter licensing standards for wholesalers and distributors; new technologies for identifying bogus products; and more public education to help health professionals and consumers identify counterfeits more readily. 	

Another approach to deterring illegal importing, McClellan explains, is to reduce the differences between United States and foreign pharmaceutical prices. At a September colloquium on generic medicine in Cancun, Mexico, the commissioner noted that higher drug prices in the United States create pressure for importing. Such pricing disparities also reflect an unwillingness of other wealthy nations to pay a "fair share" of the costs of bringing new treatments to the world, he noted, explaining that Americans account for about half of all pharmaceutical spending worldwide, while Germany pays less than 5 percent. With drug development now so complex and costly, such price regulation overseas threatens to "grind the global development of new drugs to a halt." 	

The commissioner proposed an international solution that relates the prices of new treatments to a nation's income, with rich nations paying more to support R&D so that developing nations can pay less. Realistically, there is little chance that political leaders in Europe, Japan, and Canada will accept higher pharmaceutical prices at a time of escalating healthcare costs. But relating the reimporting problem to wide disparities in pharmaceutical prices beyond U.S. borders raises important questions about who will finance future biomedical R&D.	

View from Washington

This past year has been one of tremendous activity at the Food and Drug Administration, and at other government agencies involved with healthcare, biomedical research, and national security.

Drug Development and Access: Top Issues in the Past Year

These names, numbers, and titles are a phone directory for the regulatory world.

US and European Government Contacts

A listing of the alphabet soup of acronyms, abbreviations, and initials in the clinical trial world.

CDISC Glossary: Acronyms, Abbreviations, and Initials

Europe has a new system for pharmacovigilance for human medicines up and running. It's a data-processing network and database management system for dealing with Individual Case Safety Reports (ICSRs) in 18 countries-Iceland, Norway and Lichtenstein, in addition to the 15 European Union member states. It will be operational in 28 countries by the middle of next year, when Poland, the Czech Republic, Hungary, Slovakia, Estonia, Latvia, Lithuania, Slovenia, Cyprus, and Malta become EU members. 	

The new system, called EudraVigilance by The European Agency for the Evaluation of Medicinal Products (EMEA), should be "a significant milestone with major implications on the conduct of pharmacovigilance." It is open to industry as well as national regulatory authorities in all participating countries. 	

EudraVigilance is a key component in the EMEA's plan to coordinate the supervision of medicines within the EU. The dissemination of adverse drug reaction information through this pharmacovigilance database, the agency says, should help provide advice on the safe and effective use of these products. Already the existence of the network has aided rapid transmission of data between national authorities for alerts relating to faulty manufacture, serious adverse reactions, and other pharmacovigilance data. 	

EudraVigliance's system accepts secure transmission of ICSRs via email and the Web through an electronic gateway, with appropriate security and encryption systems. The fully automated and secure system includes all aspects of privacy, authentication, integrity and non-repudiation of all transactions in pharmacovigilance. The database management system (DBMS) will acknowledge it receives safe data from users, and can also let users run routing and loading mechanisms, query the database, take a guided tour through an ICSR creation, and oversee the management and security mechanisms.	

The EudraVigilance Gateway, which follows ICH recommendations for the Electronic Standards for the Transfer of Regulatory Information (ESTRI-Gateway), offers one common portal for receiving regulatory submissions. Pharmaceutical firms that register on the system have a different common reporting point; each pharma transaction is routed through a "hub and spoke" configuration to the national authorities addressed and to the EMEA. National authorities, in turn, can use it as a secure reporting mechanism to the pharmaceutical industry and to the EMEA. 	

The heart of the EMEA's new system, EudraVigilance's DBMS, handles safety report information in full compliance with relevant ICH specifications. The DBMS continually updates a cumulative pharmacovigilance database compiled from the importing and consolidation of data from multiple sources, including information on medicines and adverse drug reactions. The extensive query and tracking/tracing capabilities embrace both the scientific and business perspectives.	

A Web-based information system known as EVWEB is fully integrated with the EudraVigilance Gateway and Database Management System. EVWEB gives freedom for ICSRs, which users can create, send, receive, and acknowledge getting. Certain users with access rights can even search the ICSR database. The EVWEB facility is specifically designed for smaller firms that do not have a fully ICH compliant pharmacoviglance system or ESTRI gateway in place. It offers similarly secure electronic reporting to the EMEA and all participating national authorities. 	

The system's benefits include standardization of secure data transmission and elimination of data transcription errors, faster communication for expedited safety reporting, and a reduction in the burden and cost of paper processing for authorities and companies. The design means messages can reach every participant quicker and cheaper than by establishing a connection between each and every endpoint.	

National authorities have access to all ICSRs stored in the European pharmacovigilance database. Pharmaceutical companies currently have restricted access to only the ICSRs it has submitted to the EMEA. Each ICSR stored in the database is flagged with the sending organization's ID.	

Each report submitted is automatically checked through a Document Type Definition (DTD) validation to ensure compliance with ICH recommendations on component parts of the message and agreed standard terminology. 	

The system is complemented by two other tools. One is a EudraVigilance Medicinal Product Dictionary (EVMPD), which supports standardized coding, reliable data retrieval, and scientific evaluation of medicinal products reported in ICSRs. The dictionary has been designed to support the collection, reporting, coding, and evaluation of medicinal product data in a standardized and structured way, based on authorized regulatory information. It includes proper style standards for active ingredients, excipients, pharmaceutical forms, routes of administrations, concentration ranges and units, country codes, and the identity of marketing authorization holders. It also provides a standardized approach to support updates and variations to product information through the defined responsible product owner. And to cope with Europe's numerous languages-22 official languages across the 28 participating countries-it is multilingual.	

The other complementary tool is MedDRA, which has been integrated into the EudraVigilance DBMS and EVWEB to support the coding, classification, retrieval, presentation, and communication of medical information presented in ICSRs.  MedDRA-the Medicinal Dictionary for Regulatory Activities-was developed as a clinically validated international medical terminology for regulatory authorities. MedDRA is used in the pharmaceutical industry during all phases of the regulatory process; it's the standard for clinical studies, reports of spontaneous adverse reactions, events, regulatory submissions and regulated product information. Use of MedDRA codes in participating countries became mandatory in 2003 for all adverse drug reaction reporting. Smaller firms based in the participating countries can obtain reduced-price access to MedDRA at just $1,000 a year, since the EMEA recognizes that the $3,500 lowest level of normal subscription can be prohibitively expensive. Very small firms, who submit fewer than 100 ICSRs per year, can even obtain free MedDRA access.	

The major update of the EU rules on pharmaceuticals (see "No End to European Evolution" in this issue) was still being vigorously fought over as this issue of ACT went to press. The conflicting points of view argued by research-based and generic segments of the pharmaceutical industry in Europe have led to haggling over the details of intellectual property protection in the current draft text. But the tone and scope of the debate-and especially how it has drifted toward the political-can also be gauged from another principal protagonist in the discussions, the patients' lobbies. 				

One of the most visible organizations in this broad coalition of pro-patient (and often anti-industry) groups is the self-styled Medicines in Europe Forum, a Paris-based pressure group of scientists and political activists. As a counterpoint to the industry views on how EU rules should be changed, the Medicines in Europe Forum has been calling loudly for "a policy on medicines that puts patients first." The time for listening to the pharmaceutical industry is over, it says, suggesting that the EU is largely under the influence of international drug firms. 	

And it urges that patients' needs include "knowing which are the active ingredients of a drug, being informed on the added therapeutic value of a new drug, benefiting from a compassionate use programme when it is vital, reporting an adverse effect, and contributing to the transparency of regulatory agencies' activities by having a place on management boards."	

As it says, with irrefutable logic, "ultimately medicines are used and paid for by patients/citizens/national insurance contributors." Accordingly, the EU rules review should place the emphasis on the needs of patients and of European citizens in general. So, they should have the information that is "key to making the best possible use of medicines." With open disdain for what it sees as secretiveness among drug firms and regulators, it asserts: "The days when doctors and pharmacists spoke Latin and prescribed secret remedies whose properties were known only to them are long over." Today, patients have the right to know what decisions are taken about their health and to be told about the expected effects of their treatments, and this includes being informed of the precise active ingredients of medicines, the Forum claims. So, it insists that packaging and in advertisements should feature International Nonproprietary Names. And patients should know "whether the medicines they are using really are the most suitable for their particular condition." One of the Forum's most strongly expressed complaints is that "many new medicines are sold without having been compared to the most appropriate drugs for the same disease already available on the market."	

The remedy, it argues, is for independent information to be compiled and disseminated. "Health authorities must be able to supply comparative information on the added therapeutic value of new medicines-or on the lack of it," it says. And "this information can be obtained only if the appropriate clinical trials are carried out."  What is not wanted is "information put out by the pharmaceutical companies and designed for promotional purposes."	

A similarly democratic strand runs through the Forum's arguments on representation of civil society on the boards of medicines agencies. "Each day European and national medicines agencies take decisions that concern patients' health: the evaluation of new medicines and their marketing authorization; monitoring adverse effects and withdrawing drugs from the market for safety reasons; the production of information for health professionals and patients," says the Forum. And "given that the pharmaceutical industry represents huge financial stakes, it is essential for the decisions taken by the agencies to be transparent, and for this duty of transparency to be enshrined in law." In concrete terms, what the Forum wants is patient advocates in healthcare and social protection systems so they can continually "remind these institutions of their duty to protect the public interest."	

The wide range of the Forum's arguments also take in the need for more scope for compassion "for patients in a therapeutic impasse," which it defines as "patients with terminal diseases, and for whom there are no effective existing treatments." They should have access to promising medicines that have not yet been authorized, it says. The principle will become practice only if national authorities are "forced to set up compassionate use programmes," by creating application procedures to allow health professionals or patient groups to seek compassionate use, and if such procedures are fully funded.	

The other plea the Forum makes is for patients to be encouraged to get involved in direct reporting of adverse effects. Doctors, pharmacists and other health professionals are not all strongly motivated to report adverse effects, and do not necessarily place the same degree of importance on some adverse effects as the patients themselves, it argues. Patients should be allowed to report directly to the public pharmacovigilance system any adverse effects-for instance by filling in a form available from community surgeries and pharmacies.	

The sometimes-cloistered world of clinical trials may imagine that the discussions in Brussels on future legislation are focused on the academic and scientific matters that lie at the heart of their professional activity. But, as the Forum's lucubrations demonstrate, it just isn't like that when it comes to European politics.	

View from Brussels

Europe has a new system for pharmacovigilance for human medicines up and running. It's a data-processing network and database management system for dealing with Individual Case Safety Reports (ICSRs) in 18 countries-Iceland, Norway and Lichtenstein, in addition to the 15 European Union member states.

New Pharmacovigilance, and New Inputs to Legislation

Find out about the purposes, rates, meetings, and publications of over two dozen socieites and associations.

Applied Clinical Trials-12-01-2003