Best Practices for Conducting Clinical Trial EDC User Acceptance Testing
By Tommy Jackson on March 15, 2018
User Acceptance Testing (UAT) is a critical component of clinical trial development using Electronic Data Capture (EDC). It’s the responsibility of the Sponsor/CRO to perform UAT prior to using the EDC to collect data in accordance with a protocol.
Early and thorough UAT can result in studies that are easy to use and monitor, and have short submission preparation timelines. We’ve found this to be especially true when our best practices for conducting UAT are followed. Prelude EDC provides tools and features that facilitate UAT every step of the way, making it a faster and more effective process. Prelude Dynamics’ seven best practices are described in more detail below. Perhaps you have some of your own you can share with us!
- Thoroughly test every e-CRF
- Verify masking and role-dependent access
- Involve a site
- Test the electronic lab interface
- Consider risk factors
- Involve the statistician early
- Conduct Pre-UAT
Thoroughly Test Every e-CRF
This may seem obvious, but it’s surprising how often this is not done. Sometimes test personnel become complacent when most e-CRFs work as expected, and abandon test efforts early. But then during study conduct it turns out that the Investigator Signoff form – which is the typically the last subject form to be filled out – isn’t working correctly.
It’s especially important to verify that the Inclusion/Exclusion e-CRF is working correctly. Imagine the consequences to the clinical trial if subjects that should be excluded are not. Similarly, dosing calculations should be rigorously vetted.
This is a good time to check the upper and lower limits on data entry fields and to make sure all calculations are operating as intended. Prelude EDC’s embedded Field Specification displays this information for every form in the study, which facilitates this process.
Verify Masking and Role-Dependent Access
There’s almost nothing worse than unmasking a study, so don’t risk it! Identify every e-CRF where the treatment group (or other masked field) is present and verify that masked roles cannot see the value in either the e-CRF or the audit trail. Prelude EDC’s Field Specification is also useful for performing this task.
Inversely, imagine that the treatment dispenser has no access to the Dosing form. Avoiding mid-study changes to permissions reduces the probability of non-contemporaneous entry of critical study data. Prelude EDC’s Study Requirements Document (SRD) Configuration can be used to verify that permissions to view, edit, review and finalize each study form have been configured correctly.
Involve a Site
Sites are ultimately one of the most critical components of any successful clinical trial. If e-CRF layouts are clumsy or non-intuitive, sites will have trouble completing them correctly. When this is the case, site enrollment may suffer due to user frustration with the EDC.
Involving one or more trusted sites early in the UAT process can provide early insight into usability issues and allow them to be corrected before the study goes live. Prelude EDC’s Feedback function can be useful in capturing site input on usability issues. Happy sites require less monitor intervention and produce cleaner data.
Test the Electronic Lab Interface
Studies that benefit from the use of electronic lab interfaces, such as those supported by IDEXX and Marshfield, enjoy more accurate and timely lab results data in their study. It’s important to coordinate with lab company personnel to assure that data required by their LIMS is included in electronic lab requests, and in the e-CRFs if necessary. It’s typical to configure study-specific parameters used for tracking and billing in the EDC, and to test the interface in a training study (or training site) prior to going live. Prelude EDC supports fully validated electronic interfaces for both IDEXX and Marshfield Labs.
Consider Risk Factors
Every study has risks, but for every risk there is a way for the EDC to assist in detecting when a risk is becoming a reality. Many EDCs support ad-hoc query functions that generate reports based on user-entered criteria. These can be configured to generate reports pertaining to potential risks such as a high number of reactions at a drug injection site, or abnormally high blood glucose level after dosing.
Losing subjects due to lack of adequate follow up poses both a financial and a schedule risk. Tools such as the Prelude EDC Event Calendar provide ways to quickly and easily detect missed visits before a protocol deviation occurs. EDC statistics functions can also be used to view data in graphical format, which is more intuitive than rows of numbers and can be used to identify outliers.
Custom reports can be developed in the EDC that are tailored to look for specific potentially risky outcomes. These can be developed at study build time, or added later when unanticipated risks begin to arise during study conduct.
Involve the Statistician Early
Configuring the study database so that the data is organized in a way that makes it easy for the statistician to prepare the data for submission is something that is sometimes overlooked when studies are being developed. This can result in both time delays in submission and difficulty with data analysis.
We recommend entering test data into either a training study or training site, and exporting the data for analysis as part of UAT. This approach allows the statistician to take an early look at the exported data, and test it with their statistical analysis software. If changes are needed to the database to facilitate analysis, these can be made and testing prior to going live.
Conduct Pre-UAT
Pre-UAT? It’s not unusual to identify multiple study changes while testing, especially when unfamiliar with the power and flexibility of EDC. Conducting a Pre-UAT ahead of the UAT schedule can help identify and address most of these changes ahead of time. When this is done, UAT consists of verifying that all requested changes have been implemented, tested and work together in an intuitive way.
Starting study test early is generally recommended to avoid cramming test efforts into the last week prior to going live. When testing is deferred, sites have a more difficult time entering data and study conduct is more difficult and lengthy than it needs to be.
Conclusion
In our experience, following these best practices for performing UAT results in less risky, higher quality studies with excellent usability and data integrity. EDCs such as Prelude EDC provide many validated, integrated tools that facilitate both UAT and study conduct. Prelude Dynamics provides training in the use of these tools prior to the conduct of Pre-UAT and/or UAT. We are increasing receiving requests to deliver the data from training studies/sites as part of study closeout. This data serves to demonstrate that UAT has been conducted.
If you’d like more information about any of the tools and features discussed in this article, or would like to schedule a demo, please contact us.
