How NOT to Design eCRFs
By Tommy Jackson on February 18, 2020
There are so many ways to design electronic CRFs. Perhaps by looking at how NOT to design them, we can come up with some fresh ideas for how to make your next study a smashing success.
Fail to consider how well the forms will function on handheld devices.
The sun us setting on big screens and desktop workstations. Soon a new day will dawn, and the sun will shine solely on small screens. They provide mobility and greater access across the board. They are the key to conducting business in the information age.
Take, for example, form layout. Regardless of whether you are viewing it in landscape or portrait, a long list of fields is not conducive to the thought-processes of the human brain. Fields need to be grouped together in a way that makes sense to the person filling it out. And news flash: the person filling out the forms for your study is guaranteed to be relying on a human brain.
Put simply: don’t shortchange the design process of the study build in favor of a quick turnaround. You will thank yourself later when you are parsing through the raw data for usable statistics.
Require users to navigate extensively to do their jobs.
Do not bury important fields at the bottom of the form. Don’t force your sites to have to dig deep to get what you need. Cater the form configurations to quick and easy navigation.
Employ checkboxes instead of radios.
Simplify the investigators’ jobs before they even begin. Give them less options when possible. Dummy-proof your study to the point where the patients themselves could fill out their forms.
Don’t provide options for “Neither,” “N/A,” “Not Done,” or “Other.”
Bottlenecks happen when a field is lacking necessary options. The investigator will simply leave it blank if they do not see the option they need. Then the monitor goes in and queries the blank field. The investigator responds that they didn’t perform that particular test. The issue gets shunted up the line until it gets to Prelude and a mid-study update is requested to add the pertinent option to the field in question. This takes time, causes confusion, and costs extra support hours, depending on the complexity of the request. There will always be mid-study updates. It’s a normal part of the process, but some things can be prevented. Do your best to predict issues before they happen, and your study will run much more efficiently. Just don’t overthink it…
Assume that the data will be entered in a certain order.
After taking into account all of the above, we’ll leave you with this: You can’t predict everything. Put as much thought into your eCRF’s as you can, but first you have to learn the difference between what you can control and what you cannot. Providing a radio in place of a list of checkboxes is all well and god, but don’t assume they will fill out the first item on the form first. Don’t go overboard with field suppressions and calculations. Try to think the way a clinical would. That’s all you can do.