Interview: Steve Friedman on NCI's trial group overhaul

Steven Friedman, chief at the Operations and Informatics Branch for NCI's Cancer Therapy Evaluation Program, is leading the IT charge in the National Cancer Institute's overhaul of the Clinical Trials Cooperative Group Program.

The transformation aims to speed up cancer drug development, as reported earlier, by revamping the 55-year-old program that serves 14,000 investigators nationwide. A report by the Institute of Medicine, which cites complexity and inefficiency, calls for consolidation of back-office administration and data management operations in conjunction with the overhaul.

How do you begin thinking about an IT overhaul for such an established program?

Friedman: Our goal is to ensure that IT systems are supportive of the revised business processes of the cooperative group program. And not just our IT systems; also the local systems of the groups. They submit data to us. Among them, we look for IT gaps that we can address. 

For our IT systems, we think about concept design. We have to consider and analyze the impact of changes on existing processes and systems, and make sure we have the resources.

We already have some ongoing projects in support of the cooperative groups. Some are aligned with the transformation effort, some aren't. The question becomes, do we alter an existing program because of the transformation? As we hit milestones, we have to make sure that all activities are keeping pace.

When did you get involved with the project?

Friedman: I've been involved in conceptual discussions over the last year as we began the transition to a new director [Harold Varmus became NCI director last July]. Ideas were highly conceptual then; the IOM report gave the concepts some teeth.

Since then, our team has developed a Reports Timeline module so group investigators, ops personnel and group chairs can see how their protocols are developing against timelines.

Who are the program constituents, and what do they need?

Friedman: Investigators are the frontline constituents, conducting trials and generating data. But the NCI team is our major constituent.

Investigator needs get funneled up to the cooperative group ops level and then to the group chairs for policy-making. We work with the chairs and ops folks. Many chairs are investigators, so we get that perspective first-hand.

Groups, institutions and investigators may have varying degrees of IT savvy or systems setup. But we have longstanding processes that they use for submitting data to us. Here we rely on the support of the ops centers.

Our systems use older technologies and we're migrating to a service-oriented architecture.

What are some of the technology topics you're likely to consider?

Friedman: Commercial off-the-shelf technology versus custom applications: We've relied on custom applications in the past. But we can use more of the COTS that's out there now. We're seeing a shift to COTS technology, integrated according to our needs and those of the groups and cancer centers. It's a collaborative process more than ever.

The NCI has procured a COTS product from Medidata for implementation of a clinical data management system, both internally to the institute and externally with the groups. It will provide standard platforms for data submission to the NCI. We're already working with the groups on implementation; subcommittees are established to address implementation activities.

Data management: We're trying to capitalize on co-locations for data storage. By leveraging the existing support that's out there, we can reduce the burden on us. Robust disaster recovery is another area.

Mobile technology: There's definitely more telework. We're seeing the shift and are mindful of it. There's now an iPhone app that the oncology community uses in the field to track clinical toxicity-related adverse events. But we have to wait for policy, at the NCI or even the HHS level, before we can address broad-based use.

Data security: Protecting patient data is of high concern. All of the patient data we get is de-identified. We make sure it's secured. 

Also, we have to make sure that the right people have the right access to the right apps and data at the right time. We'll use rolling credentialing to make sure that investigators in the cooperative groups access only the projects they're working on, while group administrators can access all of the projects in their groups.