Clean Up Your Data – Effective Edit Checks

By Editorial Team on July 7, 2022

A lot of the time, it’s okay to make mistakes. In fact, it’s encouraged! But not so much in clinical trials. Not entering a visit date or accidentally entering a subject’s weight incorrectly could have dire consequences. That’s where constraints (more commonly known as edit checks) come in to save the day.  

What Are Edit Checks?

Constraints, or edit checks, are used to check data to make sure it meets standards/requirements and alerts the user that data entry fields have or may have errors, such as: 

  • Required data is missing 
  • Entered data is not formatted correctly 
  • Entered data does not meet range or order requirements  

This allows data to be “cleaned” as it is entered, enhancing data integrity for a study. Try doing that with paper!   

Edit Checks do play a critical role in collecting cleaner data, but keep in mind that too many can prolong the data entry process. Balance is key. So, what type of edit checks should I add to my forms?  

Firstly, know that a constraint can be a warning or an error. A warning indicates a soft error. Soft errors do not prevent a form from being placed in review. An error indicates a hard error that must be fixed before a form can be placed in review. The following constraints cover both types. 

Guide the Entry of Information

Use warnings or errors on all fields on a form to guide the entry of the information. For example, if a physical exam form is saved with the subject’s vitals but no visit date, the first error might be generated by an error constraint on the blank visit date field. Or, consider the “Other” option with a “Description” textbox – an error constraint on the “Description” field could be used to require more information if “Other” was selected and “Description” was left blank. 


With dates, check chronology. Make sure that users are not entering dates in the future or that specified dates are in the right order (e.g., the end date of an adverse event cannot come before its start date). Or, maybe a subject was supposed to come back for a follow-up visit 7 days after the last visit, +/- 1 day – issue a warning if the follow-up visit is out of range. 


Issue warnings when data is outside the expected bounds (e.g., for temperature, weight, blood pressure high/low ranges). Use warnings, instead of errors, to allow out-of-range data to be entered – which can happen in clinical settings.  


When dealing with tables, sometimes you will want to check that at least one field was entered. Take, for example, the question of whether the subject experienced an Adverse Event – yes or no. You pose the question and present a table allowing the user to enter data about the Adverse Event(s). If the answer to your questions was “yes,” use an error constraint to ensure that at least one adverse event is entered (e.g., a clinical sign). Let’s take this one step further. If a clinical sign was entered, consider what other data in that AE’s row should be entered too – a start date, outcome, treatment, etc.  

While organizations require edit check practices that are tailored to their work, these tips can help optimize productivity and reduce time spent entering data. Edit checks can help you avoid mistakes for cleaner data. Happy Building! 

Interested in building a study with Prelude?

Request a Demo