ePRO and EDC: Perfect Together

For a growing number of trial sponsors, the clinical research process involves the use of a mix of stand-alone electronic systems designed to automate and streamline various aspects of the development process. Members of the sponsor's clinical research team, clinical sites, and contract research organizations (CROs) are likely to interact with multiple eClinical systems throughout the life of a trial, including tools that handle activities ranging from study build to data capture to trial management. As these systems become modus operandi in clinical trials, getting them to work together and easily share data can allow sponsors to realize additional benefits and further improve trial efficiencies and data management.

Photography: Getty Images

The next step

With the increasingly widespread adoption of eClinical tools, sponsors are now looking to integrate systems so that data captured through various technologies can be viewed together or compared for monitoring and decision making. Clearly the industry has much to gain from a more holistic, end-to-end process in which clinical and operational data are now accessible earlier in the clinical trial, without the effort of manual and ad hoc batch transfers. This availability of combined data promises to facilitate efficient and effective decision-making and enhance the overall drug development process.

A vivid example of the significant potential gains from integrating clinical systems is the opportunity to link two separate and important sources of subject data: patient reported outcome (PRO) data, which subjects personally record, and case report form (CRF) data, subject-related data that clinicians enter into electronic data capture (EDC) systems.

Recently, more and more trial sponsors are asking subjects to record self-observations and measurements to better understand drug effectiveness and safety. For example, PRO systems are often used to capture subjects' pain-related data, enlisting them to record their pain level several times a day. Such PRO data has become an important component of drug development.

At the same time, the potential unreliability of traditional, paper-based PRO data has become well known, as sponsors note subjects' tendency to complete paper forms, for example, just before a doctor visit rather than on the schedule given by the trial protocol. To bring more rigor to subject data capture, sponsors are increasingly turning to compliance improving electronic PRO (ePRO) methods. In either a device-based (eDiary) or phone-based (IVR) ePRO system, subject entered data is electronically time and date stamped and uploaded to secure Web-based applications that enable sponsors and investigative site personnel to monitor subject and trial progress.

Figure 1. Medidata Rave provides users with an integrated view of both CRF and PRO data.

Meanwhile, EDC systems provide computer-based clinician entry and can streamline the clinical trial process to deliver significant time and cost savings to clinical trial sponsors. EDC systems typically combine data capture and data cleaning in one process, enabling sponsors to reduce risks and improve efficiencies.

While EDC systems have enabled sponsors to more efficiently capture and manage CRF data (physiological measurements, lab values, etc.), ePRO systems are often employed separately to collect and report on subject data. Currently, there are combined EDC/ePRO systems on the market with varying degrees of tools and flexibility. Users will need to investigate and decide which tools will best integrate with their other eClinical systems.

Why integrate?

For sponsors and site personnel who must enter study data on a regular basis, the multisystem approach requires study users to log in to two or more system interfaces to view both sets of data. Bouncing between systems to see a complete view of a given subject's data adds unnecessary steps and no doubt frustration to the site workflow. Ideally, researchers should be able to access an integrated view of subject data that includes both CRF and PRO data.

From a data management standpoint, the two streams of data must eventually be combined, which raises potential data matching and synchronization issues that traditionally bring risk of trial delays. Without integration, discrepancies between common data in the two systems, such as subject identification or visit dates, must be resolved near the end of the trial when pressure to lock the database is greatest and monitors and investigators must be guided to deal with the conflicts. The resulting added pressure of managing discrepancies in the two isolated datasets almost invariably adds to the time and expense of clinical trials.

When sponsors do consider linking ePRO and EDC systems, they may be daunted by the perceived time and effort required to design, develop, and implement an integration, which may seem to introduce new delays into their trial timelines due to the need for custom programming. Despite the expected gains of integration, sponsors are generally wary of extending trial set-up time to accommodate extensive custom integration work.

What's needed?

To overcome the challenges of developing an integration between systems, it is helpful to look for an integration approach that can be leveraged beyond just meeting specific requirements of a particular study. It should be one that can be reused and quickly deployed when rolling out each new study. This can avoid the need for time-consuming, expensive, and complex custom solutions. Additionally, reusing good practice designs allows research teams to avoid redundancies in integration customizations and learn from prior experience without reinventing the wheel each time.

Thanks to the efforts of the Clinical Data Interchange Standards Consortium (CDISC), the industry now has ubiquitous, basic standards that support eClinical integrations. These include the Operational Data Model (ODM) standard for the acquisition, exchange, reporting, and submission of clinical research data. The ODM standard provides a flexible XML-based format that can accommodate all data values and greatly helps facilitate data sharing across ePRO and EDC systems. In order to realize these benefits, sponsors must choose clinical systems that support industry standards.

However, even with the great flexibility of data standards, sponsors can encounter challenges given the different data requirements of the two systems and different studies. Here, some extra effort is required to define conventions within the ODM framework so that a specification template can be reused for each new study. In a given study an example of this arises from the different

frequencies for data collection in ePRO and EDC. While ePRO data may be collected several times a day, CRF data is collected during less frequent subject visits. This difference begs the question of how to best fit PRO data within the visit structure typically defined within the EDC system.

In the case of site-based PRO, in which subjects provide self-assessments during the office visit, it may make more sense to include the PRO data in the visit structure. With home-based PRO, often this is resolved by parking the PRO data in a dedicated diary folder within the subject's electronic file in the EDC system. Agreeing up front on a standard approach to handling ePRO data within the context of the ODM eliminates the need to revisit this design decision for each study.

Since the objectives of integration are to provide a single view of the combined data and resolve discrepancies earlier in the trial process, thus enhancing decision making, sponsors should seek an integration designed to support transfers from the ePRO system to the EDC system in close to real time. Transferring data in weekly or monthly batches leads to outdated views of the data in the EDC system and potential confusion about the real, current state of the data. Most ePRO systems feature a central database that stores the data for all trial subjects. To properly accomodate an ideal integration, whenever the central ePRO database is updated, new data should immediately be transferred to the EDC system.

Defining a clear strategy for system integration can also reduce the risk of a sponsor being locked into one vendor's systems and allows for sponsors to select systems that meet their needs as well as to leverage their IT investments.

Benefits of integration

The case for integrating ePRO and EDC systems lies in the access to a comprehensive, close-to-real-time view of global trial progress. Site personnel can enjoy the convenience of logging in to one system to access an EDC dashboard that contains both subject assessment and subject visit data, including subject compliance information. Site investigators can then proactively drive subject behavior to ensure compliance with the study. For example, the electronic time stamps in subject diaries, now viewable in the EDC system, help investigators understand subject diary entry habits and encourage compliant behavior.

Benefits abound for data managers as well. As with eCRF data, data managers can run real-time edit checks against subject diary data in the EDC system. Moreover, the inclusion of ePRO data in EDC systems supports ad hoc reporting of eCRF and ePRO data together in real time. The ePRO–EDC integration also enables sponsors to resolve discrepancies much earlier in the trial process. Data teams can compare and contrast subject reported data as well as administrative data, such as checking consistency of subject IDs across data streams on a more real-time basis. This allows sponsors to identify and resolve conflicts in subject and site-reported data faster, yielding to a faster time to database lock. With errors flagged within the integrated system for easy identification, exception notices are generated earlier within trials, eliminating the time and cost of manual, time-intensive identification of these discrepancies. Finally, in situations where eligible and successful randomization relies on a combination of both PRO and CRF data, sponsors are able to automatically calculate, execute, and review more accurate and comprehensive randomization criteria in a single system, translating into more informed randomization decision making.

And for IT teams, choosing ePRO and EDC systems that offer rapid data sharing is key to an efficient study start. Vendors that offer tools that comply with CDISC and other industry standards are the most appealing, positioning sponsors to easily set up integrations for near real-time sharing of clinical and operational data across systems. For the selection of ePRO and EDC systems specifically, compliance with ODM standards is a key criteria.

Clearly, the benefits of well-designed ePRO–EDC integration offer a significant impact on a trial's execution. With a broader view and context of subject health within the trial, along with the ability to make more informed decisions throughout the life of the trial, integration leads to more efficient handling of subject compliance issues and helps optimize site, subject, and trial performance, ultimately offering a streamlined trial with reduced risk and enhanced cost savings.

The future of integration

While sponsors have already begun to see the benefits of using ePRO data within EDC systems, there's also potential to benefit from mining ePRO metadata in trial analysis. For example, as more experience is gained with viewing combined subject data, sponsors can also include more ePRO metadata, including assessment completion rates, missed alarms, and data transfer rates, which will help them track, evaluate, and drive overall subject compliance for each study. In this brave new world of ePRO–EDC data sharing, the industry may also welcome "dynamic CRFs," in which EDC systems are able to add questions to the eCRF on the fly based on data imported from the ePRO system.

With a focus on bringing real-time access to trial data unified from various clinical systems and across clinical activities, the future of clinical data management will feature a significant uptake of ePRO–EDC integration. Those who undertake these integrations by leveraging industry standards and best practices will maintain a competitive advantage with the ability to drive enhanced decision making and more efficient trials.

Jon McClelland* is vice president of product development at invivodata, Inc., 5615 Scotts Valley Drive, Suite 150, Scotts Valley, CA 95066, email: jmcclelland@invivodata.com.

Andrew Newbigging is vice president of integrations development at Medidata Solutions Worldwide.

*To whom all correspondence should be addressed.