There is a chronic misconception that the only significant difference between EDC and ePRO is that the former is used by Investigators, and the latter is used by patients. I’ve discussed this topic recently with a number of people and there has been interesting perspectives. Personally, I’ve been slightly shocked how little many people with mostly an EDC background actually know about ePRO. Sure, there are similarities, we’re collecting data from patients for the purposes of clinical trials, but once you start looking at things with a bit more detail, you’ll start noticing some obvious differences.
In no particular order:
Difference #1 – Scientific Matters
There is really almost no clinical outcomes science involved in EDC. eCRF pages do not need to be scientifically validated to demonstrate content validity, reliability and ability to detect change. This is a difficult one to even try to explain to data managers or EDC experts. They just give you a blank look or mix the issue with data standards or part 11 or other system validation requirements. I suppose it is understandable, EDC is often just a mechanism to make existing paper data electronic and usable by different stakeholders in the study – you don’t need to worry so much about where the data comes from, which leads me to my second difference …
Difference #2 – Users
Users in EDC are mainly study coordinators, CRAs and data managers. Patients (almost) never touch these systems themselves and it’s clear to see that already from the user interface. They’re clearly designed by data managers to collect data efficiently, without having to worry too much about ease-of-use and workflow features. The users are professionals that are paid and trained to use the systems and can deal with a bit of complexity. Patients, on the other hand, come from all walks of life, many with no previous computer experience (not to even mention mobile devices) and have to deal with often complex requirements for frequent data collection without getting paid. This puts completely different design requirements for the two systems and this is something that is quite difficult to grasp. I also started my career in this industry building EDC systems and eCRF pages and I was pretty happy with my geeky and efficient designs. When we moved into ePRO, it was quite difficult to learn to understand my new users and learn to adapt. I’ve managed people building ePRO for many years and I’ve witnessed the same thing many times, people with no ePRO background can have really hard time learning this. After 12 years, I’m probably pretty good at this now, but I can honestly say that I’m still learning …
Difference #3 – Data
Another significant difference is the source data. I still get the question “How do you perform SDV?” in ePRO. The answer is that with ePRO, you don’t. Your ePRO data IS the source. This answer usually leads to a few seconds of thoughtful silence, but it certainly computes in the EDC person’s world as well as it is likely something they’d probably rather avoid with all EDC data and for many, this is their dream come true. For me, the paper that’s still so often involved in the EDC world is one of it’s biggest weaknesses and also the reason why I think ePRO actually adds a lot more in terms of efficiency than EDC, since ePRO completely eliminates paper, while EDC basically just makes electronic copies of it …
Difference #4 – Edit Checks
EDC systems are designed to detect data discrepancies. A study coordinator might key a value that is illogical compared with another entry. In this instance, an EDC system will flag up a query, and ask the user to make a correction. Edit checking or ‘rules’ in ePRO is a bit different and more focused on guiding the patients with alarms, performing automated calculations and making sure patients don’t miss a question. Data is primarily standalone and rarely needs to be cross-referenced with other PRO data. Referential integrity is much less of an issue in ePRO. As a result, the focus of implementation is far less on the logical checking of data, and far more in ensuring the patient is compliant with the protocol.
Differences #5 – Functionality
PROs / ePROs are usually very important in studies, often they’re the primary or at least key secondary endpoints and due to the regulations in place, studies often need a reliable electronic audit trail for this data and use of paper is not really a realistic option. This elevates the importance of ePRO in a study to a very high level and a large proportion of resources and time is spent in designing the perfect tool, but in terms of volume of functionality, ePRO is often very simple compared to EDC. To be serious in EDC, you will need randomization, coding, lab data processing, adaptive roles / workflow, query management, hierarchical status management, etc etc. So while there might be less science in EDC, there is a lot more functional complexity.
Of course, after all, the true source for most of the data collected is the patient. There are just different pathways for that data to get into the study database. If the process for collecting this data is inspected above the ‘ePRO’ or ‘EDC’ silos, perhaps by an engineer with no experience in either system, or someone who’s has an in-depth understanding of both systems, then this process probably would look a whole lot different than it does in most studies today. It truly amazes me how difficult it seems to be to see beyond your own particular silo, even when they’re so close together. There is a lot of potential for clinical trials to be faster, more efficient and cheaper by rethinking how the data collection process works. Sure, this also requires the tools used for this to be adapted to support it, but most leading eClinical products already offer a lot of flexibility in terms of configurability and connectivity and a lot can be done with those already today. Without doubt, in the future these systems will be closer to each other. Some vendors and sponsors will lead the progress, while others will follow. Where are you going to be?