What is Global Trial Bank? Global Trial Bank (GTB) is being developed by the American Medical Informatics Association (AMIA, www.amia.org) to become the world’s peer-reviewed repository of protocols and results from clinical trials of all types. Once established, doctors, nurses, and other health professionals, industry, the public, and scientists will be able to perform powerful search and data exploration on our advanced database without charge, and GTB will become a major global resource for knowledge management in biomedicine. Why is Global Trial Bank important? A clinical trial is a research study in which human volunteers are treated and/or observed to answer a particular health question. Clinical trials of both drugs and devices are one of the most valuable sources of knowledge about which therapies are provedto be safe and effective or worthless and/or dangerous. GTB seeks to make public and searchable the results from all clinicla trials conducted worldwide, regardless of whether the results would favor their ultimate use in general patient care or not, so as to provide an unbiased knowledge base for improving human health. Furthermore, using advanced medical informatics technologies, GTB will make the contents computer-understandable so that information and decision support systems can directly access the knowledge over the Internet. In this way, GTB will improve the management, dissemination, and use of the valuable knowledge from clinical trials. What are the likely benefits of Global Trial Bank? GTB promises both short and long-term benefits from more complete and usable reporting of clinical trial information. A. Patients and the General Public Patients and the general public will be the major beneficiaries. Trial participants will be reassured that knowledge derived from their contributions will be impartially available to anyone having a stake in improving human health. GTB will have a patient-friendly search engine and resources to help patients make sense of trial results. B. Clinicians Clinicians will be able to search precisely for relevant trials. Also, decision support systems that are developed as adjuncts to GTB will be able to electronically match patients to relevant trials to help clinicians apply trial results to take better care of their patients. C. Medical Researchers The computability of GTB data will make it much easier to retrieve and compare related trials, evaluate their quality, assess their scientific value, and combine the evidence. Moreover, GTB data can be automatically “mined” for new findings and insights that cannot now be detected across trials. D. Clinical Trial Sponsors Because of the accessibility of GTB information for search and analysis, trial sponsors can avoid unnecessary duplication of trials, thus reducing costs as well as exposure of human participants to potential added risk. E. Medical Informatics The size and nature of the GTB database will help medical informaticians and computer scientists develop more rigorous and effective methods for ‘mining’ and displaying trial data, and for using trial data in clinical decision support. GTB offers all parties an independent third-party forum that allows trial sponsors to achieve the public’s expectation of open, unbiased results reporting. GTB’s peer review mechanism offers high quality reporting. How is Global Trial Bank likely to be different from other trial registries and trial reporting Web sites? 1. GTB is expected to contain sufficient protocol and results information to make sense of a trial In general, trial registers collect only enough information about a trial to uniquely identify its existence and to help patients enroll in studies. Results are included usually only as links to publications, but it is known that publications are biased towards reporting only favorable results and trials. GTB seeks to be unique in capturing all the protocol and results information needed for the express purpose of supporting high quality scientific assessments of any study’s strengths and weaknesses. 3. GTB contents will be peer reviewed All results information in GTB will be peer reviewed, either by mainline medical journals or by the Public Library of Science (PLoS, www.plos.org) through GTB’s companion journal PLoS Clinical Trials. To our knowledge, no other trial reporting website has an integrated peer-review mechanism. 4. GTB contents are computable In addition to textual information, GTB will contain structured information that is coded using a standard clinical vocabulary (SNOMED-CT). With such computable information, GTB will be able to support very specific queries such as “retrieve all trials on women over age 60 with heart disease in which a beta-blocker was studied, and that report heart attack rates at 3 and 5 years for which at least 80% of the participants were followed up.” Also, the results can be analyzed directly with GTB or third-party analytic software without having to manually transcribe any data from text articles to another computer. No other trial reporting website of which we are aware would support such detailed and powerful queries and analyses. CONTENT Which trials will Global Trial Bank include? GTB aims to include all clinical trials from all sponsors worldwide. This includes trials conducted by universities, government, and pharmaceutical and other companies. We will focus on hypothesis-testing randomized trials, but will also accept early phase exploratory trials. GTB will also accept trials that were completed and reported before GTB’s launch. What trial information will Global Trial Bank include? All GTB entries will include a minimum set of information about the study’s protocol and results, to enable critical appraisal of the trial’s methods, analyses, and results. Additional required items will be defined in consultation with key stakeholders. Investigators may report more than the minimum requirements if they desire. A. Protocol Information The minimum set of protocol information will be at least the 20 item minimum set agreed to in the April 2005 WHO-mediated consensus. The following data items will be captured in coded form using SNOMED-CT: • Intervention(s) • Key conditions for eligibility, including age and ethnicity • Primary and secondary outcomes • Harm outcomes In addition, the e-document of the final protocol with amendments as approved by the ethics board of the main study site will also be requested. http://www.plos.org B. Results Information The minimum set of results information will be at least the information in the ICH E3 summary. The following data items will be captured in structured, computable form: • baseline characteristics for all intervention groups and subgroups • raw, intention to treat summary level results for all outcomes in all intervention and a priori subgroups at all prespecified timepoints • compliance data (i.e., how many in each intervention group were on assigned intervention at each timepoint) Results for efficacy or per-protocol analysis can also be reported in structured form. How will Global Trial Bank assure the validity of the contents? GTB will assure the validity of its contents via peer review conducted through its collaboration with the Public Library of Science. First, GTB will ensure that all protocol and results submissions meet the minimum reporting requirements. Protocol information will be reviewed using a simple checklist of reporting requirements. Two GTB advisory board members and one person from the company in order to meet reporting requirements and to attest that the self-stored protocol is the one used in the trial will review protocols that firms wish to keep in their own possession. Protocol information will not receive any additional scientific review. Results information will be peer reviewed through either of 2 mechanisms. Results may be peer-reviewed and published by traditional scientific journals, in which case GTB will not perform any additional peer review. Results may also be submitted directly to GTB for publication in GTB’s companion journal PLoS Clinical Trials. PLoS will rigorously peer review these submissions, including an evaluation of the methods and statistics, and will annotate the submission with a structured commentary. PLoS Clinical Trials anticipates accepting all trials that meet a set of minimum requirements, regardless of outcome or “clinical interest,” to maximize the number of trials that are reported to the public. PLoS will also provide a discussion forum for open, post-publication peer review of all GTB entries of completed trials. How are Global Trial Bank entries related to entries in trial registries? Before being deposited into GTB, all trials that began enrollment after July 1, 2005 will be required to be registered at a trial registry approved by the WHO’s International Clinical Trial Registration Platform. Ongoing trials must be registered at some point in an approved registry before the trial is closed. The July 1, 2005 date is aligned with the clinical trial registration policy of the International Committee of Medical Journal Editors. GTB entries will be cross-indexed to registry entries through the trial’s registration number(s). How are Global Trial Bank entries related to study reports in journal articles? Going forward in time, the GTB entries will be linked to trial reports published in peer-reviewed journals, including PLoS Clinical Trials. The publishing of protocol and results information in both a journal and a trial bank is called trial-bank publishing. Release of protocol and results information in GTB will not violate any copyright that traditional journals may claim. However, all trials published in PLoS Clinical Trials will be published under the Creative Commons open-access license and a copy of each article will be deposited in PubMed Central, the free full-text repository of biomedical articles. DEPOSIT AND ACCESS POLICIES Will there be charges to deposit into Global Trial Bank? There will be a modest fee for depositing information into GTB, waivable for hardship cases and developing countries. GTB will offer institutions (e.g., NIH, universities, pharmaceutical company) subscriptions with an annual rate to cover all GTB deposits, or a bundled rate (i.e., a fixed rate for a set number of trials), or a proportion of the fees (e.g., 50%) for each deposit from that institution. Publication charges for PLoS Clinical Trials are tentatively set at $2000 See: http://www.plos.org When is trial information expected to be reported to Global Trial Bank? The expectation is that trial investigators will report the minimum set of protocol information into GTB within a month after the trial is registered with an approved registry. Investigators will also be encouraged to report protocol amendments to GTB as they occur or at the time results are reported. The minimum set of results information is expected to be reported to GTB within 3 years after trial completion, with allowances for studies that are under peer review at a journal. What definition of open access will GTB use? GTB uses the open access definition of the Budapest Open Access Initiative, which includes free availability on the public internet, permitting any users to read, download, copy, distribute, print, search, or link to GTB entries, crawl them for indexing, pass them as data to software, or use them for any other lawful purpose, without financial, legal, or technical barriers other than those inseparable from gaining access to the internet itself. Will all GTB information be open access? All results information deposited into GTB will be open access. We expect that most study protocols reported to GTB will also be open access. However, some clinical trial sponsors (e.g., pharmaceutical or device companies) may have legitimate proprietary interests concerning the protocols of their ongoing trials. To accommodate such concerns, GTB will allow protocol information to be deposited into a lockbox, in one of 2 forms A. GTB-stored Into GTB in structured format, along with the protocol e-document to be stored on the GTB server, all under access control B. Self-stored protocol e-document to be stored at sponsor’s server. Protocol information that is in the trial registry registering this trial will be put in GTB in structured format Lockbox protocols will be made open access when the results are reported, or some period of time (e.g., 3 years) after the completion of participant enrollment. Thus, all results information and the study protocols of all reported results will be open access. How will lockbox security be maintained? GTB-stored protocols will be stored on a secure server and accessible only via the SAFE protocol. GTB will implement all reasonable security measures to prevent unauthorized access to GTB-stored lockbox protocols, but shall not be liable for any incursions into the GTB server. For self-stored lockbox protocols, although the details are not yet finalized, we anticipate that study sponsors will keep the protocol on their own information systems in encrypted form and will provide GTB with the encryption key. The encryption key will assure GTB that the “locked” protocol remains unchanged. What kind of database access will Global Trial Bank provide? The general public will be able to search and browse all open access GTB entries on the Inter¬net. There will also be tools and resources designed to help patients understand what information they are reviewing and the limitations of that information. In addition, we anticipate that users will be able to bulk download information for use with¬out restriction, and that GTB will provide standards-compliant interfaces to other information systems. We anticipate that the GTB will become a valuable database for use in informatics research as well as clinical research. ADMINISTRATIVE How is Global Trial Bank funded? Initial funding will come from a mix of public and private sources from outside the pharmaceutical industry. Deposit fees to be determined, as well as additional funding from public and private sources, will be sought to address ongoing operational as well as research and development needs. AMIA, GTB’s home organization, is a 501(3)(c) non-profit professional scientific organization of approximately 3500 clinical informatics specialists from 45 nations around the world. What is the current status of GTB? A seed grant has already been received from one foundation, and additional foundation sources are being explored. A two-stage process of development is anticipated. A first round of funding will allow a detailed business plan to be developed. This stage will require 8 months. Over another 18 months, the plan will be implemented. GTB will be developed and managed under AMIA’s purview until the independent organization emerges. PLoS and other potential partners will be engaged formally in a manner to be negotiated before the end of summer 2005. GTB recognizes that the pharmaceutical industry is a key stakeholder in trials registration and reporting. GTB will seek to identify and work with interested pharmaceutical and/or device companies to establish an example of acceptable and accountable protocol reporting, even in the absence of any mandating legislation. We will then incrementally expand the companies that report to GTB. GTB will continue to keep abreast of and provide input to relevant legislation in the USA and other countries, but GTB will not rely on specific legislation for achievement of its project milestones. Global Trial Bank Steering Committee GTB is concerned with all clinical trials worldwide. As a result, GTB’s lead and advisory personnel will reflect international perspectives as much as possible. Don Detmer, MD, MA President & CEO, American Medical Informatics Association, (AMIA) Bethesda, MD and Professor of Medical Education, University of Virginia, Charlottesville, VA William Detmer, MD, MA President & CEO, Unbound Medicine, Inc, Charlottesville, VA Christopher G. Chute, MD, DPh Professor of Medical Informatics and Associate Professor of Clinical Epidemiology, Mayo Clinic College of Medicine, Rochester, MN Ken Getz, MBA Research Fellow, Tufts Center for the Study of Drug Development, Boston, MA Karen Greenwood Executive Vice President, AMIA, Bethesda, MD Charles Jaffe, MD, PhD Senior Global Strategist, Digital Healthcare Group, Intel Corporation, West Chester, PA Isaac Kohane, MD, PhD Assoc. Prof. Pediatrics and Health Sciences and Technology, Harvard Medical School, Boston, MA Ronald L. Krall, MD Sr. VP, Worldwide Development, Research and Development, GlaxoSmithKline Pharmaceuticals, King of Prussia PA Ida Sim, MD, PhD Assoc. Prof. Medicine, & Assoc Dir. Medical Informatics, Program in Biological & Med. Informatics, UC San Francisco, CA Richard Smith, MD, MBA Vice President, United Healthcare Europe, London, UK Emma Veitch, PhD Publications Manager, PLoS Clinical Trials, Public Library of Science, Cambridge, United Kingdom GTB will retain consultants on an ad hoc basis during the start-up phase, in anticipation of forming a formal International Advisory Board.