The 3 tenets for designing a clinical data management system

Abstract:
This post reviews the importance of 1) proper study design, 2) good data modeling and 3) realistic estimation of project timetables. The article concludes with a discussion of eSource and attempts to dispel some of the myths including how DIY EDC study build save time (they don’t).

Enjoy!

The trend of DIY: good for EDC vendors, less good for sponsors

The trend for small studies/IIS (investigator-initiated studies) is to use cloud EDC applications
that enable end-users to build eCRF and edit checks using a graphical user interface. This so-called DIY (do-it-yourself) approach is used by most cloud EDC vendors such as Medrio and Clincapture as a way of lowering their barriers to entry to the market.

However – what is good for vendors (lowered barriers to entry) is not necessarily good for sponsors (faster time to market of their innovative drug or medical device).

There are several downsides to DIY study-builds, in particular in the first multi-center study that a sponsor builds.

1. Most sponsors (and this is particularly true for startup biotech and connected medical device companies) do not have the expertise to design and implement and EDC system. Most sponsors do not have specific expertise in data modeling and may not grasp the difference between the GUI and the actual data model itself.

2. Lack of expertise extends to Usability where the novice sponsors doing DIY study build often create overly complex forms with difficult to use sequences of edit checks. As a rule of thumb, a eCRF should have up to 6-7 edit checks. Less than that, and you run the risk of not validating the data properly and more than that you run the risk of an unsatisfied CRC and slow direct data entry into the eSource (the EDC eCRF becomes the electronic source with direct data entry).

3. Validation and verification? If you DIY your own study, you will have to use a third-party for V&V which adds more time and money and further negates the economic advantage of GUI form builders.

4. In addition – many biotech/biomed executives are running startups with small
teams that need to get to market fast with constrained budgets.Constrained budgets  and study build done by a student will not get you there faster.

5. Startups running new studies will grossly underestimate the amount of time needed for the study build. As we show below, the actual coding of the eCRF is less than 10% of the total time. Consider the following work breakdown:

– 2 days to review, finalize and reconcile the protocol with the eCRF
– 2 days to build the eCRF using a GUI form builder,
– 3 days to design the edit checks and document in a spec; add another 3-5 days for review and design iterations
– 3 days to code the edit checks
– 10 days for user acceptance testing (3 rounds x 5 days/round)
– 10 days for V&V

A total of 30 working days or 6 work weeks. Note how the eCRF form building is less than 10% of the entire project.
Even if you had a GUI builder that was 10X more productive – you would have negligible impact on the project schedule.

To place the notion of DIY in even better perspective – consider a 12 month study where you saved 2 days in the study build, by using a DIY approach and incurring the risk of a poorly-designed data model
(making it difficult to almost impossible for the study statistician to analyze the data for the statistical report).

Usability

While almost every EDC vendor touts their features and price points, there is
little value for sites, data managers, and study subjects if the software is not
simple and highly-usable for the study’s purposes.

If you are reading this post, you are either still using paper data capture,
cycling through free trials of EDC software in hopes of finding the right match
for your sponsor needs, or unsatisfied with the current electronic data
capture system in place at your study site(s) and considering another vendor
when it comes time to renew your EDC contract.

Professional study design as a critical success factor in reducing risk
In clinical studies, trial efficiency is the critical path to product release to
market. In the United States alone, the biopharmaceutical industry spends
billions of dollars and 10+ years on developing a single new drug. While
successful drugs lead to a better quality of life for the ill, their research
and development is titanically costly. Thus, study efficiency is revered as gold
in the medical research industry.

Build a standard data management platform for a series of studies

A planned and professional study build process enables CRCs,
data managers and study monitors to collaborate on an efficient process of collecting eSource,
performing RBM (remote risk-based monitoring) and focusing attention on the most critical success factor
for study success – namely – patient compliance.

A planned and professional study build process enables and EDC study build to serve as a standard data model and eSource platform for a series
of studies of any scale from Phase I, II, IIB, III and IV.

Using eSource to save time and effort

Writing this post, I found that I was totally flummoxed by the eSource story.
Why are all these companies developing special purpose applications to collect
data when the guidance clearly says that the data can be entered directly to the
EDC.

According to FDA Guidance, data can be entered into the eCRF either manually or electronically as described below.

a. Direct Entry of Data Into the eCRF
Many data elements (e.g., blood pressure, weight, temperature, pill count, resolution of a
symptom or sign) in a clinical investigation can be obtained at a study visit and can be entered directly into the eCRF by an authorized data originator. This direct entry of data can eliminate errors by not using a paper transcription step before entry into the eCRF.

For these data elements, the eCRF is the source.

So if we have a modern EDC with good online edit checks – there is no need to duplicate all the EDC functionality in another application.

With your teams you can just use the EDC to do direct data entry on mobile devices.
Unfortunately many sponsors still don’t get it.. You can propose entering data directly into Medidata but many sites still want to use paper.

Although while using paper, study coordinators usually write notes on the paper, as a matter of fact, it is just as easy to enter annotations and discrepancy notes directly into the EDC.

Remote monitoring and eSource

A sponsor does not need a special “eSource” system to do remote monitoring although an eSource system may have some attractive user-satisfaction features. For example clinical inks esource has writing functionality.. the EDC is just like paper but on tablet

Typically in a medical device study,the patients show up for a screening visit – they fill in ICF, IE etc directly into the EDC. RBM is important for monitoring patient compliance and it can be done using the data that was entered directly into the EDC, collected from ePRO or from connected medical devices.

The process of designing a good RBM plan starts with identifying critical data for the patient compliance to the protocol. If critical data lies in F/U visits for example, and the RBM module detects deviations, the site will need to go back and verify the correct values in the eSource.

For the ePro data, one can use computer data science algorithms to identify outliers or unusual trends

In summary

1) The FDA Guidance specifically says “direct data entry to eCRF”.
2) Changes in data due to RBM or data management review can be annotated in the eCRF annotations/discrepancy notes that are an integral part of the EDC anyhow.
3) The CRA does not have to physically visit the site to do this (although site visits have their own value)
4) There is no justification for a special eSource app if you use RBM and a modern EDC

I appreciate you taking the time to read my post!

1 Comment

Leave a Reply