Helix IM and Surround SCM Provide Exceptional Value for CureSearch
CureSearch implemented Helix IM and Surround SCM to replace custom programs and manual processes, reducing study implementation time from over four months to less than two weeks.
What value did CureSearch get
Time and cost savings
Brilliant branching model
More elaborate products took months of effort and still did not work.
AT A GLANCE
Organization: CureSearch Children’s Oncology Group
Web Site: www.childrensoncologygroup.org
Headquarters: Monrovia, CA
Description: CureSearch implemented Helix IM to gain a fully functional issue management system customized for their needs.
The search for a cure to pediatric cancer requires the dedicated efforts of researchers and multi- disciplinary cancer specialists around the globe. The CureSearch Children’s Oncology Group (COG), a National Cancer Institute-supported clinical trials cooperative group, was established from the merger of the major pediatric clinical trials groups based in North America to maximize the benefits from this work.
As a united team, COG develops and coordinates clinical trials for cancer treatments at more than 235 member institutions, which include cancer centers at major universities and teaching hospitals throughout the United States and Canada, as well as at sites in Europe and Australia.
COG’s clinical trials target all major types of childhood cancer, including leukemia, Hodgkin’s disease, lymphoma, neuroblastoma, sarcomas, and various types of tumors (central nervous system, kidney, germ cell, and liver). COG typically has 100 ongoing trials that enroll approximately 5,000 new patients each year. Tracking the progress of these patients, along with 35,000 childhood cancer survivors in active follow-up, requires collecting massive amounts of information from its geographically dispersed member institutions.
Streamlining a Cumbersome Process
Historically, the collection of this information was facilitated by the group’s Remote Data Entry system (RDEs), COG’s clinical-trials data- collection system. This workhorse allowed researchers to submit updates on the progress of study participants from remote locations. RDEs was a collection of individual custom programs written for each new study. Supporting new studies entailed the expansion of its databases, web interfaces for data collection, and validation scripts to ensure data quality. Consequently, the period of time to bring each new study from development to release was more than 18 weeks.
To streamline this cumbersome process, COG’s technical team—led by Ken Ingram, CIO—began development on the Enhanced Remote Data Entry System (eRDES), an entirely new set of tools that promised to reduce study implementation time to less than two weeks. Partway into this effort, the team needed to upgrade its tools to match the demands of an increasingly complex project. The first task was to put a formal issue tracking system in place.
“We started with a combination of email and a spreadsheet to track bugs,” said Ingram, “but that approach did not scale well—it was too time-consuming. To improve the situation, we tried MetaQuest’s Census issue-management system. However, after four months of effort, the team abandoned this project because we were unable to make it meet our needs.” “With Helix Issue Management (IM), our team implemented a complete issue management solution in days. More elaborate products took months of effort and still did not work.”
Customized, Fully Functional Issue Management
“Not only do we rely on Helix IM for bug tracking, but we also use it to log customer support issues,” noted Ingram. “This is achieved by configuring Helix IM to automatically create new issues when emails are sent to our support address. The flexibility built into Helix saves us both time and money because it eliminates the need to purchase and implement a separate help desk tool. These savings are especially important to a nonprofit organization.”
During their discussions with Perforce, COG discovered Surround SCM, a change management solution fully integrated with Helix ALM. This was fortunate, because version control was another thorny issue for COG.
“We were using Microsoft’s Visual SourceSafe, and the eRDES code base was increasingly difficult to manage,” Ingram said. “Most problematic was Visual SourceSafe’s poor support for sharing classes between projects. The eRDES comprises a Windows desktop client that allows Research Coordinators to create study-specific forms and a web interface that delivers those forms to trial participants.”
To maximize code reuse, the software is designed to let the client and web application share common classes. Since updates to these applications are released on separate schedules, COG needed a version control tool that supported multiple development streams. “In our case, releasing software through Visual SourceSafe became an overwhelming task because it was difficult to have separate builds within a single project.”
Ingram’s team found that Surround SCM was designed explicitly to overcome these issues. Not only does Surround offer a flexible branching model that lets COG efficiently manage the eRDES code base, but it also makes it easy to specify which files are shared between projects and which are not.
In addition to solving the code-sharing problem, Surround SCM delivered other unexpected benefits. “Before Surround SCM, the deployment of eRDES from development to quality assurance and then to production was tedious,” Ingram said. “Preparing a build package took at least two days. We now use Surround SCM to manage deployments because it ensures that no file is left behind—files from a tested branch are simply retrieved from the web server’s root directory. Our developers also found that Surround SCM is faster than Visual SourceSafe. Getting the latest files now takes less time than before. All of this means increased productivity and shorter cycle times—we now can manage multiple builds per week.”
Another pleasant surprise for Ingram’s team was the ability to run Helix IM and Surround SCM in an accessible environment using Windows clustering services. COG has two servers that run the Perforce tools: one is configured as the active server and the other as the passive. Both nodes access application data on a shared SAN device. When a failover event occurs, the Microsoft cluster manager switches to the passive server, which brings the tools back online. “Although Perforce does not formally support this environment,” Ingram said, “we got it working in a matter of hours. With the Census product, which is supposed to support failover configurations, we never were able to get it to work.”
Perforce Provides an Exceptional Value
“Perforce tools are an exceptional value and are delivered by a company that stands behind what it sells,” Ingram concluded. “The responsive support we receive illustrates how much the company cares that its products address our needs.”