YouTube Facebook LinkedIn Google+ Twitter Xingrss  

By Allison Proffitt

May 4, 2010
| The Dana-Farber/Harvard Cancer Center, the largest cancer center in the United States, is a consortium made up of seven Harvard University affiliated institutions running about 800 cancer clinical trials per year, in more than a dozen disease areas, with more than 14,000 patients enrolled. And yet until 2000, all of those trials were tracked on paper, and researchers were still grappling with data gaps common to paper records, says Marina Nillni, EDC program manager at Dana-Farber Cancer Institute.

Most of the trials are Phase II trials, with 82% being Phase II, II-III, or III. 84% of the studies are principle investigator-initiated (PII) clinical trials, a number that tripled between 2001 and 2005. These are the most complex trials, Nillni explained, which makes them higher risk. These are also the trials that PIs plan to publish—having their data lost in reams of paper was slowing the publishing process and impeding grants.
When Nillni and her team first started looking at a suitable EDC program, they had a short list of objectives. Nillni wanted a solution that could:

• Eliminate paper case report forms (CRFs);
• Reduce data entry workload through system integration;
• Improve quality of data through up front edit checks;
• Improve turnaround time for study analysis and safety reporting; and
• Increase efficiency by reducing the number of queries to the study team.

With those objectives in mind, the Dana-Farber team came up with a list of 173 functional requirements. For most of 2004—1500 hours—the team investigated options, reviewed RFPs, participated in demos, performed reference site visits and technical assessments, analyzed risks and costs, and finally winnowed the list of EDC options to one: InForm by Phase Forward. It was an “exhaustive process,” Nillni remembered.

After signing the contract in February 2005, the team began the process of bringing the solution in house. “We didn’t want to use a hosted model,” Nillni said. “Given the volume of our studies, that really wasn’t financially a good idea for us.” Training, hardware and software installation, and library construction took about five months, and in October that year, Dana-Farber launched its first EDC study.

A Few Good Trials
With the huge volume of trials underway at Dana-Farber, EDC implementation could never have been universal. With only a couple of exceptions (a very long-running bone marrow study and a relatively new pediatric acute lymphoblastic leukemia study), no studies were “grandfathered in”, said Nillni. Only new studies were entered into EDC.

Studies were entered into EDC by disease group. Disease groups were chosen preferentially based on the number of paper forms processed in a year, the number of subjects,  the number of new trials each year, and the length of those trials. Standard case report forms were developed for disease group as much as possible. Disease programs are most likely to be eligible for EDC if they include multi center studies, long trial duration, and high or fast subject accrual.

Nillni encountered the challenges expected of a very large cancer center with many trials. There was a lack of standardization across (and sometimes within) disease programs. Time was needed from clinical, biostatistic, and technical teams. Large and long-running studies need to reflect changes in EDC over time.

In the four-and-a-half years since the first EDC study went live, Dana-Farber has launched 117 EDC studies including more than 7500 subjects. More than one million paper forms have been or will be saved with the current portfolio and hundreds of phone and email queries have been eliminated. That said, Nillni acknowledges that Dana Farber still has many more trials that have not been moved to EDC yet.

Asked if she would do it again the same way, Nillni said yes, the research was worth it and the in house model has worked well for Dana Farber. Indeed, there have been unexpected bonuses. “By going through the process of EDC development—looking at the case report forms and data points and when they need to be collected—protocols became more robust and of a higher quality.” 

Bio-IT World Web Symposia Series: “Evolving EDC: Finding eClinical Solutions.” November 2009.

For reprints and/or copyright permission, please contact  Jay Mulhern, (781) 972-1359,