44s Abstracts

s12 STANDARDIZED COMPUTER-BASED REPORTING OF CLINICAL

TRIALS: IN SUPPORT OF META-ANALYSIS

Ida Sim Stanford Medical School

Stanford, Carifornia

Performing a good meta-analysis is difficult, partly because clinical trial reports are often incomplete and inaccurate. With the advent of electronic publication, now is an opportune time to build an integrated, electronic, trial reporting system that will support improved, standardized trial reporting; improved retrieval of trial reports; and direct computer access to both descriptive and summary results data of trials. For such a system to be truly integrated, the electronic trial databanks, or Trial Banks, must all use a common conceptual model of clinical trials. This conceptual model should be able to represent information about the design, execution, and results of trials in any stage of execution, from planned to fully completed.

I have incorporated trial reporting recommendations from trial registries, structured trial reporting, and other clinical trial commentators into an initial conceptual model of randomized trials. The model consists of 156 concepts, which can be classified as being about population, treatment, outcomes, treatment allocation, statistics or administrative matters at the design, execution or results stage of randomized trials. I have implemented this model as a Filemaker Prom database. Preliminary use shows that the model has fully captured four randomized trials for the task of contingency table meta-analysis, as well as partially capturing four prospective cohort trials. Future work will include further refinement of the model, its implementation as a Web-based Trial Bank, and formal evaluation of the model’s expressivity and utility for meta-analysis.

Al3 EXPLORING THE INFORMATION SUPERHIGHWAY: USE OF THE INTERNET IN CLINICAL DRUG TRIALS

Sarah Carr, Vicky Foster, Matthew Wiener and Aimee Blagg

Pharmatech, Inc. Denver, Colorado

The biomedical product development industry has a need to reduce the “time to market” and the cost of product development. The rate limiting step for clinical trial completion can be the timely recruitment of subjects. This leads to the necessity for expanded and innovative subject recruitment procedures for the clinical trials. The Internet is a potential resource for subject enrollment. The Internet has over 35 million users in North America, 50 million worldwide, and is said to be the fastest growing information medium. The expansion of the Internet, and particularly the world wide web, into the public domain has led to its utilization for communication with a variety of stakeholders: subjects, investigators and sponsors. A “home page” , or a more complex world wide “website”, allows for the dissemination of general information, but also provides detailing depth contact via e-mail should a respondent request greater detail. Multiple hotlmks on websites give the interested “surfer” referral resources. Pharmatech has developed a website which includes information of biomedical research and specific trials (http://www.pharmatech.com). The initial month of operation reflected 15 visits a day, yielding 6 specific inquires daily. Pharmatech recently utilized

Abstracts

this website in an Astbma program. Twenty five patients were recruited over the Internet during a period of 45 days. Five were referred to clinical investigators in their respective geographic areas. Typical recruitment efforts for this type of protocol have yielded a screening: referral ratio of 16:1, at a cost of approximately $500 per patient. This internet recruitment ration was 51 at a cost of < $100 per patient. The efficiency and savings were significant. The Information Superhighway is being promoted by the media and utihied by an ever-increasing number of households. It provides the clinical trials industry with an additional medium of communications and interaction.

Al4 RECRUITMENT FOR A LARGE-SCALE CLINICAL

STUDY VIA THE INTERNET

K. Zadnik, M.A. Bullimore, J.T. Barr, M.O. Gordon and T.B. Edrington

Collaborative Longitudinal Evaluation in Keratoconus (CLEK) Study

Universily of California, Berkeley, California University of Cali@omia, Davis, California

The Ohio State Universily, Columbus, Ohio Washington University School of Medicine, St. Louis, Missouri

Southern Calz@omia College of Optometry, Fullerton, California

The Collaborative Longitudinal Evaluation of Keratoconus (CLEK) RuDose. Study is a five-year observational study with an ambitious recruitment goal of 1,000 keratoconus patients in ten months. To the best of our knowledge, the CLEK Study is the fust NIH-sponsored study to recruit subjects via a page on the World Wide Web (WWW): htpp://spectacle.berkeley.edu/CLEK/CLEK.html.

M&oda. The CLEK Study WWW Page provides a brief description of keratoconus and the CLEK Study with a fit of the locations of the 16 CLEK Participating Clinics across the US. It includes pointers to the NE1 and the NIH Web Site. Initially, we contacted all individuals with eye-related Web Sites. Most were willing to add a pointer to the CLEK Study page. We also contacted all the Web Searchers/Directories. Now, if a patient searches for “keratoconus”, he or she will find the CLEK Study Web Site.

Rest&a. Over a 40day period (9120195 to 10/30/95) the CLEK WWW home page was visited 399 times. More interested visitors probably went to the “CLEK Clinics” page, which was visited 168 times. Eighty-six visitors’ home bases could not be identified, 153 were commercial users (e.g., aol), and 57 were non-US based. Sixty-nine inquiries were from universities, and 21 were from government-based e-mail addresses. Eiht were from organizations. Between 6126195 and 11/27/95, the Study Chairman was contacted by e-mail by 33 interested patients or potential referring doctors.

Conch&on. The Internet is an innovative and effective means of communicating with eligible patients for large-scaleclinical studies. Supported by NE1 grants UlO-10419, UlO-10069, and UlO-10077.