Microsoft Server Product Portfolio Customer Solution Case Study Epilepsy Study Chooses Microsoft for EDC, Global Collaboration, and Data Management Overview Country or Region: United States Industry: Healthcare—Life sciences Customer Profile The University of California at San Francisco is dedicated to advanced biomedical research, graduate-level education in the life sciences and health professions, and excellence in patient care. It has 10,000 employees. Business Situation The university needed an unusually large and sophisticated technology platform to support hundreds of researchers in 20 centers around the world, while minimizing the burden on the informatics staff. Solution Project directors chose to base their public Web site, intranet portal, applications, and databases on Microsoft technologies from end to end. Benefits Cuts development time by 50 percent Provides exceptional ease of use Delivers near 100 percent uptime, minimizing support needs “This is such an advance for the field of clinical research informatics that we’re looking at making components of the architecture available to others in the research community.” Gerry Nesbitt, Director of Bioinformatics, the Epilepsy Phenome/Genome Project, University of California at San Francisco Unlocking the genetic secrets to epilepsy might provide the key not only to better treatments but also to prevention. To pursue that possibility, researchers at the University of California at San Francisco and at New York University are heading up the Epilepsy Phenome/Genome Project, which spans hundreds of researchers at 20 centers worldwide. Their challenges in collecting, processing, storing, accessing, and analyzing massive amounts of data, and in collaborating with colleagues around the globe, are enormous. To gain the patient-tracking, electronic data capture (EDC), global collaboration, and data management capabilities that they need, the researchers chose Microsoft technologies. The resulting solution meets their needs for reliable and efficient data collection, low-cost support and training, faster application enhancement, faster knowledge discovery, and improved productivity. “We chose Microsoft technologies over LAMP because we needed 100 percent reliability and scalability, and we had very little time to get the solution into production.” Gerry Nesbitt, Director of Bioinformatics, the Epilepsy Phenome/Genome Project, University of California at San Francisco Situation Epilepsy. It’s a word no one wants to hear, but too many people hear it nonetheless. The disease afflicts more than 50 million people worldwide, according to the World Health Organization (WHO). Epilepsy can strike the young as well as the old. Symptoms can range from brief staring spells to full-blown seizures. There are treatments but no known cure. Although some patients respond well to the first antiepilepsy drug they take, others experience intractable forms of the disease that do not respond well to medications. They may require massively invasive surgery, such as the removal of part of the brain, to achieve a level of relief. In 2007, two of the largest centers for epilepsy research in the United States—the University of California at San Francisco in the west and New York University in the east—decided to team up to study one of the least understood aspects of epilepsy: its genetic basis. The National Institutes of Health granted U.S.$15 million for the research. The researchers at the University of California at San Francisco and New York University formed the Epilepsy Phenome/Genome Project (EPGP) as the centerpiece for what became an international, multi-institutional, multiyear, collaborative research effort. That effort grew to include 20 academic research centers worldwide, making it one of the largest such studies ever conducted. The broad scope of the project—which encompassed thousands of data points for each of thousands of research subjects— was necessary to give researchers a study population large enough that they could identify and link genetic information to phenotypic data. It wasn’t only the research population that needed to be large. The researchers quickly realized that they needed an informatics infrastructure and an application architecture that were as large and sophisticated as the questions they were studying. One reason was the large volumes of data that researchers would need to store and access. Another was that the researchers—who, after all, were experts in biology, not computer science— needed the ability to work easily with and analyze that computerized data. In addition, the large numbers of geographically dispersed researchers meant that the project’s five-person informatics staff couldn’t possibly provide in-depth technical support; the project therefore needed applications that were as easy for the researchers to learn and use on their own as they were easy for the informatics staff to deploy, administer, and support. That the group of researchers was large and dispersed also indicated the need for collaboration and workflow tools so that researchers halfway around the world could work together as though they were down the hall from one another. Reliability was another requirement because project leaders knew that they couldn’t afford to lose or compromise any of the research data. The project’s informatics staff, based out of the University of California at San Francisco, had a complex, multifaceted set of requirements to meet—and not much time in which to meet them. The staff had just eight months before the project was slated to enroll its first research subjects, and informatics systems had to be live by then to accept incoming data. “Enterprise-class data management was essential, and SQL Server provided that.” Gerry Nesbitt, Director of Bioinformatics, the Epilepsy Phenome/Genome Project, University of California at San Francisco Solution The first step in developing the informatics infrastructure was to identify the services and functions that the solution would have to support. Working closely with the researchers, the informatics staff developed a list of requirements that included everything from collecting and managing detailed phenotypic data, electroencephalographs (EEGs), and magnetic resonance images (MRIs) to tracking the status of participants and their blood specimens throughout the term of the study. Off-the-Shelf or Custom Development? The next step was to decide how to implement the needed functionality—that is, whether to purchase and modify an offthe-shelf solution or to develop a custom solution. The responsibility for that decision fell to Gerry Nesbitt, the project’s Director of Bioinformatics at the University of California at San Francisco. He and his team considered commercially available software for clinical trial management, such as Oracle Clinical and PhaseForward InForm, and found them wanting. Some were too expensive, others couldn’t accommodate the electronic collection of large phenotypic data sets, and most were not scalable enough. That left custom development. The most common technology platform in academia is the so-called LAMP stack—the Linux operating system, Apache Web server, MySQL database, and PHP programming language—with Apple Macintosh computers on the front end. But Nesbitt and his colleagues made a different choice. Choosing the Right Platform “We chose Microsoft technologies over LAMP because we needed 100 percent reliability and scalability, and we had very little time to get the solution into production,” says Nesbitt. “Enterprise-class data management was essential, and SQL Server provided that. SQL Server also gave us fully supported, out-of-the-box and third-party options for reporting and analysis. MySQL didn’t. Similarly, SharePoint Server 2007 gave us the features, again built-in, that we needed for both a public Web site and an internal collaboration site.” In addition to Microsoft SQL Server 2008 data management software and Microsoft Office SharePoint Server 2007, the project’s informatics solution was based on Microsoft technologies, including the Windows Server 2008 operating system, Microsoft ASP.NET, the Microsoft Visual Studio 2005 development system, and, later, Visual Studio Team System 2008 Team Suite. Applications and a Portal The informatics team worked for eight months, and its solution went live in January 2008. That solution consisted of applications accessed through a secure intranet portal. All components of the solution were Web based, which addressed the requirement that the solution be easily accessible both to global users and to informatics maintenance and support personnel. Those components included the following: A public Web site to provide information about the project and to solicit the participation of volunteers as research subjects. A private intranet portal for researchers, which provided access to the applications, to the study’s documents, and to collaboration tools. Applications for tracking participant activity, which made it possible for each clinical site to track its participants’ visits, manage events and issues, and complete phenotypic data collection. “By using Microsoft technologies, we’ve been able to give our researchers and coordinators a level of informatics sophistication rarely seen on projects funded by the National Institutes of Health.” Gerry Nesbitt, Director of Bioinformatics, the Epilepsy Phenome/Genome Project, University of California at San Francisco A specimen-tracking system to manage the life cycle of blood specimens from the moment a blood sample was drawn to its arrival at the specimen repository. A case-report form designer and datacollection tools. Applications for EEG data collection and reviews by neurophysiology experts. Applications for MRI data collection and reviews by imaging experts. Secure file transfer for EEG and MRI files. A central database repository providing data storage, validation and integration, and robust security. A data warehouse through which researchers query the data using Microsoft or third-party tools, including Microsoft Office Excel spreadsheet software, Office Access database software, SQL Server Reporting Services, Office PerformancePoint Server 2007 (a part of Office SharePoint Server 2007), and Visokio Omniscope. coordinators a level of informatics sophistication rarely seen on projects funded by the National Institutes of Health,” says Nesbitt. “We’re seeing more reliable and efficient data collection, fewer resources needed for support and training, faster completion of application enhancements, faster knowledge discovery, and higher productivity overall.” Cuts Development Time by 50 Percent To build the large and sophisticated informatics solution on an aggressive project schedule, the developers turned to an Agile development strategy, which they implemented through the use of Visual Studio 2005 and, later, Visual Studio Team System 2008 Team Suite. Nesbitt estimates that because the team used the Visual Studio tools, development time was about 50 percent less than it would have been using the alternatives he might have chosen. These applications are hosted in a West Coast data center and mirrored for backup and disaster recovery in an East Coast data center. Each data center hosts a pair of databases: one is a transactional database, with which researchers interact through the various applications for data collection, tracking, and transmittal. The other is a reporting database that provides the basis for analysis, data visualization, and reports in response to queries from the researchers’ business intelligence tools. “The Visual Studio tools allowed our team to get a real jump-start on development,” says Nesbitt. “We took advantage of the tools’ strong interoperability with our foundation applications, such as SQL Server, and of their features for collaboration among developers—such as check-in and check-out of code and strict version control. That helped us to achieve smooth collaboration between our domestic team and our outsourced developers in Chennai, India.” Benefits For example, the Microsoft tools supported a 24-hour development and quality assurance (QA) cycle in which local developers created code and performed QA before the end of day, and then passed their work on to the developers in Chennai to address bugs and other issues. The outsourced developers finished their work just in time for Nesbitt’s local developers to The massive research project was up and running in just eight months and now gives researchers easy and dependable access to applications and data. In addition, it minimizes the support burden on the lean informatics staff. “By using Microsoft technologies, we’ve been able to give our researchers and pick up again with the new code. That shortened development considerably. The Microsoft technologies also helped to speed development by providing components—such as business intelligence in SQL Server and collaboration tools in Office SharePoint Server—that eliminated the need to build or buy this functionality and then integrate it with the rest of the system. The informatics team used Microsoft Office SharePoint Server 2007 to create a public Web site that doesn’t look like a “typical SharePoint site.” Master pages in SharePoint Server made it easy for the developers to add logos, colors, and typefaces into the public Web site, and then to make the use of those visual elements consistent throughout that site. Nesbitt proudly notes that his team got the public Web site and intranet portal up and running in less than one week. Further, the same capabilities that supported the fast development of the solution are now used by the project developers to speed application enhancements and extensions. Provides Exceptional Ease of Use The EPGP informatics applications had to be as easy for researchers to use as they were easy for the informatics team to develop. The Microsoft technologies also helped the informatics team to meet that requirement, according to Nesbitt. Researchers benefit from the use of Office SharePoint Server for the intranet portal, which provides a convenient, central place to find all of the EPGP applications, all of the research data and tools, and all of the reports and other study documentation. “Because the applications are based on Microsoft technologies, we can train new coordinators in one hour, and they are fully equipped with the skills and confidence to use the applications immediately.” Gerry Nesbitt, Director of Bioinformatics, the Epilepsy Phenome/Genome Project, University of California at San Francisco The intranet portal also gives researchers a broad choice of analysis tools. Researchers can work through familiar tools, such as Microsoft Office Excel; visual dashboards based on Office PerformancePoint Server; and research-specific, third-party tools, such as Omniscope. Because the informatics solution uses Web-based tools for tracking participants and phenotypic data collection, researchers and study coordinators gain real-time access to subject enrollment statistics, study metrics, and research data. Researchers also find that their work is expedited—and that the impact of having colleagues spread among various research centers is mitigated—by collaboration tools, especially those available in Office SharePoint Server. For example, they use discussion boards for ongoing virtual conversations, while wikis provide a way to share best practices—such as how to boost subject recruitment—among study coordinators. Even nonresearch functions are made easier for project researchers through the use of Microsoft technologies. The public Web site, for example, is maintained not by dedicated informatics staff, but rather by two researchers whose work on the site is one of their many duties. That work is made easier both by the publishing tools in Office SharePoint Server and by the software’s workflow tools, from which Nesbitt and his colleagues created an automated workflow that routes proposed content to authors, editors, and reviewers for approval prior to posting. And researchers aren’t the only ones who can work effectively because of the Microsoft technology underlying the solution. The easy-to-use applications and tools make it possible for study coordinators to complete their tasks quickly, contributing to the widespread acceptance of those applications, higher productivity among coordinators, and reduced turnover in staff. “The applications had to be especially easy to use for two reasons,” says Nesbitt. “First, anything that distracts our researchers from doing their real work is a problem. Second, research projects generally have a high turnover of study coordinators—which requires us to continually train new coordinators. Because the applications are based on Microsoft technologies, we can train new coordinators in one hour, and they are fully equipped with the skills and confidence to use the applications immediately.” Delivers Near 100 Percent Uptime, Minimizing Support Needs In choosing a technology set for their applications, Nesbitt and his colleagues were also looking for a platform that would minimize ongoing support demands. The Microsoft technologies meet that requirement, too, and do so in a variety of ways. First, the project solution is fully scalable, able to support the hundreds of broadly dispersed, simultaneous users who contribute to the project. Should the population of researchers ever grow beyond the limits of the current architecture, the informatics team can add Web and database computers costeffectively, using commodity hardware. In addition, because all the informatics applications are Web based, the Epilepsy Phenome/Genome Project spends less on deployment than it would to deploy traditional client-server applications and also saves on the maintenance costs of keeping servers and workstations updated and trouble-free. The system is also highly reliable. Built-in validation rules and checks in the applications and databases help to ensure that the collected data is correct, that it is verifiable, and that it follows the study’s protocols. Should problems occur with the databases, SQL Server data mirroring between the East and West Coast data centers helps ensure that data is available for backup or data recovery. Not that Nesbitt is anticipating such problems—to date, the system has experienced negligible unscheduled downtime, providing the equivalent of near 100 percent uptime. Even scheduled downtime for maintenance is just two to three hours a month, 50 percent of the downtime that Nesbitt says he would have seen with other platforms. All this means that researchers can use the applications and data as they need to, without having to consider limits to system availability. Nor does the IT staff need to devote significant resources to system support. Help-desk calls are running at the rate of one or two a day—a very small number for an endeavor the size of the Epilepsy Phenome/Genome Project, according to Nesbitt. “Using Microsoft technologies, we’ve built a major platform that delivers enterpriselevel scalability and reliability, along with the functionality and ease of use that our researchers need, all of which is supported by an informatics administrative and development staff of five,” says Nesbitt. “This is such an advance for the field of clinical research informatics that we’re looking at making components of the architecture available to others in the research community.” For More Information Microsoft Server Product Portfolio For more information about Microsoft products and services, call the Microsoft Sales Information Center at (800) 4269400. In Canada, call the Microsoft Canada Information Centre at (877) 5682495. Customers in the United States and Canada who are deaf or hard-of-hearing can reach Microsoft text telephone (TTY/TDD) services at (800) 892-5234. Outside the 50 United States and Canada, please contact your local Microsoft subsidiary. To access information using the World Wide Web, go to: www.microsoft.com For more information about the Microsoft server product portfolio, go to: www.microsoft.com/servers/default.mspx For more information about the University of California at San Francisco, call (415) 476-9000 or visit the Web site at: www.ucsf.edu Software and Services Microsoft Visual Studio Microsoft Server Product Portfolio − Microsoft Visual Studio Team System − Microsoft SQL Server 2008 Enterprise 2008 Team Suite − Windows Server 2008 Enterprise Technologies Microsoft Office − Microsoft ASP.NET − Microsoft Office Access 2007 − Microsoft SQL Server 2008 Reporting − Microsoft Office Excel 2007 Services − Microsoft Office PerformancePoint Server 2007 − Microsoft Office SharePoint Server 2007 This case study is for informational purposes only. MICROSOFT MAKES NO WARRANTIES, EXPRESS OR IMPLIED, IN THIS SUMMARY. Document published May 2010