EDC – turning the corner?

pharmafile | October 22, 2003 | Feature | Research and Development |  EDC, drug development, electronic data capture, technology 

Electronic Data Capture is a way of eliminating paper in the data collection process by capturing the data directly into custom-designed electronic databases. But in 9 out of 10 clinical trials, information is recorded on paper and only transcribed to a computer system some time later.

EDC has been talked about for many years – the earliest in-house pilot studies, including one by Schering-Plough – date back to the late 1980s. The EDC market is growing respectably if not spectacularly, and has currently estimated at $140 million, up from $79 million in 1999. The value of databases and related consultancy is not accurately tracked, but will be at least the same amount again.

Hard data is hard to come by, since many of the players are small private businesses that aren't obliged to publish detailed sales data. Phase Forward claims that in the first quarter of 2003, it had 55% of market share in revenues and bookings, and at the EDC level through its InForm product, had 2002 revenues of $60 million, compared to $5 million for its nearest rival.

The scale of the opportunity is potentially enormous. Total clinical development costs for a new drug are around $800 million. And up to $300 million of that is the costs of managing data collection, review, and integration. According to Centerwatch, 7 – 8,000 clinical trials are conducted each year. Of these, only 10-15% use EDC.

In the first generation of EDC – the dialup era – EDC companies pre-configured PC's with EDC software and protocol definition, and shipped them to investigator sites. The data was entered locally and transmitted to a central database using a dialup connection. Clunky interfaces and difficulties of maintaining current data meant adoption was limited.

In the late 1990's, the Internet offered a new lease of life. The problem of technology distribution was solved: there is only one copy of the data collection software, of the study protocol and of the patient data. All the investigator needs is a standard PC with a web browser and connection.

Although EDC is starting to take off, US research by US-based Clinical Data Interchange Standards Consortium (CDISC) and CenterWatch found system performance and usability as top concerns.

• About 35% of pharma companies and CROs believed systems are still not working adequately.
• Thirty nine per cent of CROs and service providers surveyed said that pharma companies routinely requested the use of EDC, up 25% since 2000.
• Use of the technology in phase III trials has doubled since 2000, with those surveyed optimistic about its ability to reduce costs and increase trial speed and efficiency.

What are the key selling points of EDC?

Graham Bunn, vice president of e-Clinical Business at Quintiles, a leading clinical research organisation, is well placed from the customer point of view to judge this emerging market, and he sees rapid data cleaning as a key benefit for EDC.

A few examples:

• Data validity: If patients on a trial must be between 18 and 65 years old, an eCRF can quickly identify where an investigator is mistakenly recruiting an under or over aged patient and prevent this occurring.
• Range Checks: An eCRF can use simple data range checks to identify data that looks invalid. For example, a patient appears to be only 1.85 feet tall because an investigator ticked the 'feet' instead of the meters units box on their paper CRF.
• Dynamic eCRF: An eCRF can be dynamic – so that for example when an Investigator indicates that a patient is male, questions concerning pregnancy, childbirth, and lactation can instantly be removed. This prevents the occurrence of impossible data where a human investigator may have accidentally ticked an inappropriate box.

Another benefit is faster clinical decision-making, says Martin Young, professional and enterprise services director of Phase Forward. "We can now conduct a safety review which used to take weeks, in five days. Because trial management becomes more efficient, that means the trial managers can keep a closer eye on the science."

How mature is EDC as a technology?

Another sign of growing maturity is an increasing emphasis on usability, on the use of the technology, rather than the technology itself. "In the early days, technology vendors threw technology out there and expect it to work, so screens were pushed out without much thought for usability," adds Young. "Customer satisfaction rates at that time hovered around 70%, but we have done a lot of interface work, and a lot of listening to our customers, and our satisfaction levels have grown to 80-90%."

A key step along the road to technological maturity is expected to be the adoption of common standards. The Clinical Data Interchange Standards Consortium (CDISC), a non-profit group, which counts 85 member companies, is developing an XML-based standard to allow raw data to be analysed online.

The investigator interface still needs most work. The sponsor pays for EDC, but the investigator determines whether or not it succeeds. Good results have been achieved by ensuring that investigators get continuing medical education credits for learning a new database.

Graham Bunn of Quintiles says, "In our EDC projects now, we spend around 20% of our time on technology, but 80% with people and process, focused around business process re-engineering, so the technology is now pretty mature."

Probably the key test of maturity is customer acceptance, and it is noteworthy that in some of the early adopter companies – who are best placed to know – "you have to get VP sign-off for a paper trial these days", says Martin Young.

Can the return on investment be proven?

Metrics in EDC are few and far between. Measurement can't happen unless there are EDC projects to measure, and only 10-15% of trials use EDC. Metrics are often quoted showing incremental improvements, but if they are based on pilot studies, they make no allowance for economies of scale, where costs can be spread across many trials. Martin Young of Phase Forward says that up to 50% of costs can be saved when more than 10 trials use EDC. Between 60% and 80% of queries can be carried out online.

But cost saving and speed, he argues, isn't the whole story. "You can achieve all the often-quoted metrics like data management savings, faster cycle time and earlier database lock with paper if you throw resources at it."

A key driver in the next few years will greater clinical trials productivity, and because EDC is scalable, it will become more and more attractive, says Martin Young. "Three to four times as many NDEs are hitting preclinical compared with three years ago, and there are simply not three times as many resources around. This means a major capacity issue in the early phase trials in three years time, and in the later phase trials 2-3 years later."

The other key difference is that, like many new technologies, EDC initially enables people to do existing processes faster, and later enables them to do new things altogether, says Graham Bunn: "If expensive EDC technology is layered onto a standard paper-based clinical trial process the outcome is high quality, faster but more expensive clinical trials.

"EDC offers more than just rapid collection of high quality data. This data becomes the foundation for trials that react sooner to what is actually occurring at the clinical sites, more focused on outcomes, and more efficient because many tasks simply disappear."

Incorporating this change involves business process re-engineering across the whole organisation (Quintiles calls this eClinical Trials), but when these changes are implemented, it becomes clear that EDC-based eClinical trials become faster, higher quality and lower cost compared to equivalent paper based trials.

Adds Graham Bunn, "The ROI is now very clear for each and every trial that is conducted with the new trial process, but getting there may be a slow and painful process for many companies, and the ROI for initial EDC studies is sometimes difficult to prove."

The real secret, adds Colin Urquhart of Datatrial, is to get and keep the investigator using the system, which is why they built a portal around the user interface, offering for example, disease data and news.

How does EDC cross over with other technologies?

EDC is subject to a curious catch-22 – investigators will not add data to a system on a computer unless they derive benefit from it, and they will not get a benefit from it until they enter the data.

The key benefit will be when the investigator can access current date whenever he needs to. And that means closer links with the clinical data management system.

The various tools used in the clinical trials process, such as patient diaries, the clinical trial system and the IVR system will hold subsets of the same data, containing similar data and so currently need to be integrated manually, and automatic integration is becoming a major priority.

On the hardware side, new products such as tablet PCs or PDAs have a greater impact in Phase I clinical trials, where healthy volunteers are brought together at fixed times and tested together.

Portability, the great selling point of these devices, is an issue where a number of patients have to be seen at the same time, but less so in later-stage trials where patients are seen individually.

Evolving regulatory authority acceptance of EDC

In many ways, regulators are leading the adoption of EDC. The FDA is encouraging electronic data submissions, preferring them to the truckloads of paper that were once standard practice.

Says Graham Bunn of Quintiles, "In my opinion, 'encouraging, supportive but evolutionary' is the phrase that sums up regulatory acceptance of EDC. Regulatory authorities are very keen to reduce the volume of paper they receive for each and every clinical trial submission. They are also trying to implement faster and safer ways of reviewing and approving submissions so that the patient population has better access to New Drug Entities (NDEs).

"EDC and associated technologies really help to drive these regulatory initiatives forward so they have been very encouraging and supportive of EDC. At the same time the regulatory authorities are undergoing similar business process reengineering steps required to fully take advantage of all the benefits that the new technology brings. This is always an evolutionary process."

What are the key questions to ask about your clinical trial to find out if you need EDC, and what to look for in a supplier?

1. Does the EDC company understand the clinical trial, process, or are they purely IT specialists? Do they know how a clinical trial manager will resolve issues, and anticipate them?

2. How important is training on their list of deliverables? A big investment in training can lead to fewer problems down the line, says Colin Urquhart. Datatrak, for example, provides e-training through its eTRAIN product, which provides scalability, or the opportunity to move from, say, 10 to 100 sites without having to increase resources at the sponsor end, and to meet regulatory requirements that all users are trained before implementing EDC.

3. How long is the lead-time to set up? Lead-time is a major consideration. Some EDC systems take several months to set-up. Formedix technologies, with auto- generation of both EDC databases and electronic forms have such a short lead-time that they are applicable for phase 1 where the specification and rollout period is, claims founder Mark Wheeldon, just a few weeks.

4. Does the trial need EDC? Graham Bunn of Quintiles says "EDC can benefit nearly every clinical trial, but there is an appropriate technology to use for each trial. Sometimes this will be EDC, and sometimes this will be paper. If EDC is selected, then the appropriate solution may involve an eCRF, but at other times fax or IVRS. The industry has got smarter at choosing the right technology to use for each trial rather than applying an eCRF in each and every case.

5. Can they prove customer satisfaction? Obtain references from customers (sponsors, CROs, site staff) who are happy about their EDC experience with this technology/service.

What is the range of service offerings?

There is a large range of EDC technology systems – from the computerised fax collection of paper based CRFs, to telephone-based IVRS through to eCRFs based on PCs, PDAs like Palm Pilots and electronic patient diary systems based on PDAs, IVRS or mobile phones.

The very latest technology has focused on capturing simple data (such as blood pressure, sleeping patterns, or cardiogram information) directly from the patient using a wide range of technologies.

Around each of these technologies new services have arisen to design and build the eCRF, ePD or IVRS system, and to train and support the people and the technology through the clinical trial.

The leader in the phone-based market, ClinPhone, has become Aventis' preferred provider for data management of it clinical trials worldwide. ClinPhone will transfer information direct into the pharmaceutical company's clinical trial database, allowing real-time access and eliminating data duplication.

Pen-based technology is an elegant attempt to bypass investigator reluctance to enter data on computers. ReadSoft, Anoto and CGE&Y have developed a digital pen and paper solution, which allows users to fill in data on a paper form with a digital pen. Handwritten text is sent via bluetooth technology and a mobile phone to data capture software where it is recognized and interpreted, and can start exchanging information with the back-end system.

What other questions remain about EDC?

EDC can be applied to many areas within a clinical trial. The collection of electronic drug safety data (Serious Adverse Events) is one area that Quintiles is moving forward for one particular customer.

Adds Bill Byrom, product development director at ClinPhone, best known for its interactive voice response technology, "Integration with the back end clinical database is crucial. You can't build tons of data checks and edit checks into the thin client data collection systems at the front end, because they will slow down screen refresh rates and limit the utility of the application."

Once you've achieved the ability to push data directly from EDC to the back end database you need to feed queries raised by the more rigorous logic and edit checks that can be applied, back into the EDC application. The quicker you ask the question, says Byrom, the more likely you are to get a high-quality answer.

In addition, front-end integration is important to the efficient application of EDC. Numerous eClinical applications share common data and functionality, such as interactive voice response systems, patient diary applications and clinical trial management systems.

The principle of front-end integration is "do something once only, and in the application that best fits the study workflow." This eliminates duplication, and data reconciliation issues.

Adds Graham Bunn, "EDC is raising many questions about the way clinical trials have been conducted in the past and the way that they should be conducted in the future. It has helped to start the process of questioning each and every task in the process."

Colin Urquhart of Datatrial argues that the key to success, for individual companies and for the EDC industry as a whole, is "To listen to what the investigators want, after all they are the 'real' end users."

So far, integration of all the data capture new technologies has been very poor. Technologies have acted as data silos, and companies have not employed consistent data standards across their organisation to encourage or allow data integration.

Says Graham Bunn, "With the implementation of technology standards like XML, and data standards like CDISC (Clinical Data Interchange Standards Consortium) that run across pharma companies, technology vendors and service organisations like Quintiles, barriers to integration are now being removed. I fully expect EDC to be firmly integrated with other technologies within clinical research in the next few years."

Practical differences between Europe and the US?

EDC first started in the US, but Europe has since caught up, and there are now no real differences in approach between these regions.

"There are still differences in suitability – but these generally occur at the site rather than country or region level and in the US and Europe at least, most of these are easily addressed," says Graham Bunn.

"We are now seeing a lot of interest in conducting EDC trials outside of the US and Europe. Once a country has the general communications infrastructure required to run an EDC trial, the site situation continues to be the most important factor."

Conclusions

In the past five years, EDC has moved from being a theoretically compelling but practically troublesome technology into a valuable part of drug development, enabling scalability in one of the least scalable of activities, as pharma pushes its cornucopia of new drug entities into development. It's been a long time coming, but EDC seems to be turning a corner, making clinical trials amenable to IT at last. The proof of the pudding seems to be that, in Quintiles' experience, at least, 85-90% of investigators don't want to go back to paper once they have used EDC.

John Hosken is principal consultant at Information Advisors, which provides consultancy on new media strategy for healthcare. He is also a qualified business coach.