BCT.org: Feasibility of a Clinical Trial Matching Tool

Institution: University of California, San Francisco
Investigator(s): John Park, M.D. - Morton Lieberman, Ph.D. -
Award Cycle: 2003 (Cycle IX) Grant #: 9DB-0070 Award: $99,200
Award Type: TRC Pilot Award
Research Priorities
Community Impact of Breast Cancer>Sociocultural, Behavioral, and Psychological Issues: the human side

Initial Award Abstract (2003)
Clinical trials enable new breast cancer therapies to be tested under controlled conditions before they are made widely available to patients. However, only 3% of cancer patients participate in clinical trials creating a major obstacle to their timely completion. Further compounding this problem is the increased demand for study subjects as more drugs continue to flow from the laboratory to the clinical setting. As a means of facilitating clinical trial enrollment, we have developed BreastCancerTrials.org (BCT.org), an Internet-based tool for matching breast cancer patients to clinical trials. The heart of BCT.org is the “online registry,” a vehicle for comparing patient self-reported medical histories with clinical trial eligibility requirements. BCT.org has been carefully designed to ensure data security and patient privacy. When a match occurs, patients must initiate contact with clinical investigators; no medical records are released without patient consent.

Our long-term hypothesis is that BCT.org can shorten the duration of clinical trials and promote participation among underserved women including seniors, racial and ethnic minorities, rural residents, and lower income families. For this pilot project we intend to study BCT.org in several key areas: This pilot will allow us to fine-tune BCT.org by sequentially testing and revising its utility for clinical trial matching. For accuracy we will compare the self-reported medical histories of 30 Stage IV patients to their medical charts. A calculated accuracy score will serve as the basis for future improvement. For acceptance and usability we will initially evaluate these measures among “intended users” in a test setting prior to public release. In total, 32 patients and 10 investigators will share information in focus groups about data security, privacy protection, site navigation, content, and general usability. Additional testing will be done after BCT.org is released over the Internet using online focus groups, system statistics and phone interviews. For the outreach aspect of the study we will develop community partnerships to help us adapt BCT.org for underserved women.

Innovative elements of the project in non-technical terms:
Although there are other websites to help patients find clinical trials, BCT.org has several innovative features. Accuracy: With its focus on one disease, BCT.org uses screening criteria that are unique to breast cancer and which can be easily updated to keep pace with new treatments. Accessibility: BCT.org is available to patients who do not have access to computers; they can use surrogate contacts or register on print forms by mail. Usability: Developed by a team of patients and investigators, BCT.org addresses both sets of users with regard to data security, privacy, and ease-of-use. Outreach: BCT.org is both a nonprofit service and a research program. It serves a practical need and at the same time provides a tool for studying patterns of clinical trial enrollment and outreach.

Final Report (2006)
We developed BreastCancerTrials.org (BCT) to help breast cancer patients find clinical trials. Registrants complete a detailed online Personal Health Record (PHR), which is matched to trial eligibility criteria. Match results contain a trial summary and site information; users can also invite sites to view their online PHR. BCT was developed by the NCI, UCSF Comprehensive Cancer Center, and Center of Excellence for Breast Cancer Care. A pilot was launched in June 2005 with trials in the San Fran.Bay Area and Sacramento.

BCRP funding was used evaluate the feasibility of the BCT online clinical trial matching tool. Measures were acceptance, usability, quality of patient self-reported histories, and potential efficacy. Acceptance: web statistics showed most BCT users between 35 and 64 years of age, Caucasian, and with a college degree; 39% were 55 years or older, 17% had only a high school diploma. Over 14-months, 733 patients registered with BCT and 614 provided consent/started a PHR. Research sites were academic centers, community hospital practices, Kaiser, CCOPs, and private practices. BCT has listed over 52 treatment trials for DCIS and invasive cancer.

Usability: Among patients who started a PHR, 70% completed it and 95% matched to at least one trial; 17% invited a study site to view their online PHR. Online help was clicked almost 1700 times. Among patients responding to our satisfaction survey (20% response rate), 88% reported it was easy to complete a PHR; 84%, that it was easy to navigate the site; 63%, that the PHR was useful other than for matching; and 77%, that they would recommend BCT to a friend. When asked about BCT safeguards for privacy, only 3% had concerns.

Quality of patient self-reported data: We launched a trial to compare a patient's BCT record with her UCSF clinic chart. We have recruited 18 of our projected 30 UCSF patients and will report results at study closure.

Potential Efficacy: Among respondents to our survey, 23 contacted a research site; 52% (12) were told they were eligible for a trial and 42% (5) of those eligible reported enrollment. A planned efficacy study was postponed to focus on feasibility. It will be redesigned to coincide with a nationwide rollout of BCT.

In addition, we conducted a paper survey of UCSF and San Francisco General Hospital patients to assess knowledge about their cancer history and attitudes to clinical trials. Results showed that education and race/ethnicity affected both parameters.

Our results show that patients will use an online trial mataching tool. We are currently developing a nationwide BCT with an improved interface, expanded educational content, and outreach to diverse populations.

Symposium Abstract (2005)
Breast cancer patients benefit when effective treatments are rapidly incorporated into routine clinical practice. But barriers to clinical trial recruitment often delay the pace at which new interventions are rigorously screened for safety and efficacy. Overall, approximately 3% of eligible patients enroll as study subjects with representation among racial/ethnic minorities significantly lower. Strategies directed at increasing patient awareness about clinical trials and providing access to study investigators could potentially accelerate clinical trial recruitment and broaden the diversity of participating patients.

BreastCancerTrials.org (BCT), a personalized, Internet tool that matches breast cancer patients to the eligibility of clinical trials is one such strategy. Conceived by patients, BCT is sponsored by The National Cancer Institute (NCI), the UCSF Comprehensive Cancer Center, and the Center of Excellence for Breast Cancer Care. Patients who register with BCT enter their cancer history onto web-based forms that are compiled into a Personal Health Record (PHR). When a match occurs, BCT provides the patient with a lay summary of the trial and contact information for the clinical investigator, each displayed in a password-protected Message Center. With consent, patients can also authorize investigators to view their PHRs. Patients do not have to visit BCT to learn about new matches; the system runs automated matching whenever a new trial enters the system or a patient updates a PHR and alerts patients via email to login to their Message Center. Patients have complete control over their protected health information; BCT does not release patient information without prior patient consent. BCT is a pilot for caMATCH, an NCI-initiative to develop clinical trial matching applications that conform to its Cancer Biomedical Informatics Grid.

BCT was launched as a 6-month research pilot focusing on trials in the greater San Francisco Bay Area and Sacramento on May 31. In addition to showing the site, we will report on the collection of feasibility data: patient acceptance within diverse populations, usability of the patient interface, the ability of patients to complete a PHR, and the ability of the system to execute valid matches. Data will be collected via web statistics, online surveys, and HELP desk interviews. In addition, we will evaluate the quality of patient self-reported data in 60 patients; 30 from the UCSF Breast Care Center and 30 from San Francisco General Hospital (SFGH), a public hospital with a large underserved population. This will be addressed by comparing each patient’s BCT-generated PHR to his or her clinic chart. In a related study, an anonymous paper survey is being conducted at UCSF and SFGH to evaluate what breast cancer patients know about their medical history as a function of ethnicity and education variables. This information will be used to develop strategies for improving BCT usability among various populations.

Symposium Abstract (2007)
Background: Difficulties accruing patients to clinical trials is a major impediment to the timely approval of new treatments for breast cancer. As an alternative to current recruitment strategies, we have developed BreastCancerTrials.org (BCT), a web-based service that allows patients to find trials that are customized to their medical histories. BCT is a nonprofit, patient-centered, tool that was developed in collaboration with the National Cancer Institute (as part of caMATCH), UCSF Comprehensive Cancer Center, Center of Excellence for Breast Cancer Care and patients advocates. It was launched in June 2005 as a pilot project listing trials in the San Francisco Bay Area and Sacramento.

Materials and Methods: Patients registering with BCT complete a detailed Personal Health Record (PHR) that is matched to trial criteria. Matches are displayed in a secure Message Center with trial summaries and contact information. Patients can contact research staff directly or invite them to view their PHR online. BCT provides online HELP, phone/email support, and alerts patients to newly registered trials. Nine research consortia in the Greater San Francisco Bay Area participated in the pilot. Measures were acceptance, usability, satisfaction, and quality of patient self-reported data. Most data were collected from web statistics and an online survey of those who completed a PHR. We are currently investigating quality of BCT patient data in an ongoing trial at UCSF by comparing a BCT user's self-reported record and clinic chart.

Results: At 14 months post-launch, 733 patients registered with BCT, 614 (84%) provided consent/started a PHR, and, among these, 428 (70%) completed a PHR. Online HELP was clicked 1687 times. Patients learned of BCT via Internet links (65%) or word of mouth (35%); 25% had not completed college and 37% were over 55 years of age (8% over 65). Of the patients who completed a PHR, 95% received at least one match and (17%) invited researchers to view their PHR; additional patients may have contacted sites directly. At one year, 20% of patients responded to our survey, 80% found it easy to complete their PHR and only 3% had concerns about privacy. Among survey respondents, 37% contacted a site and 74% were satisfied with site responsiveness. Among the respondents who contacted a research site, 52% were eligible for a trial and among these, 42% enrolled. We will report on the quality of patient self-reported data when the study concludes.

Discussion: This pilot demonstrates that an Internet trial matching service is feasible and has the potential to impact clinical trial accrual. Based upon our experience with the BCT Pilot, we are revising our user interface and data collection methods in anticipation of expanding BCT to a nationwide service. In concert with this effort, we will develop outreach to underserved populations.