Skip to main content

Clinical Development Resources

Browse our clinical development resources to find the latest articles, blogs, brochures, and videos from our experts. See the latest content below.

Data queries and how they can be avoided at the site

By  Anita Scudder, Clinical Data Analyst I, Parexel

Clinical trial data needs to be collected correctly so that an accurate analysis can be done on a trial’s  safety and efficacy.         

Query resolution (also called Discrepancy Management), includes the review of errors in the data, investigating possible reasons for the error, and resolving them with documentary proof or marking them as irresolvable.  Any clinical data that is gathered incorrectly or not gathered at all is required to be corrected or clarified. A small number of incorrect or missing data, might seem insignificant, but it’s the accumulation of these incorrect/missing data points, that makes it troublesome for statisticians to draw accurate conclusions.

When queries are not resolved promptly, the progress of the trial might cause the datapoint to be ignored and sometimes the subject might be lost.

The types of queries typically seen in a clinical trial are:

  • A query that will immediately report an error upon data entry, and
  • A query that is raised by Data Management after further review of the data collected.

Poor or missing data can slow down the flow of a clinical trial significantly.  Some examples include:

  • Forgotten data or assumed data
  • Misunderstanding of the question that was asked
  • Mistakes in extracting the data from the source document
  • Incorrectly performed clinical procedures
  • Incorrect calculations of values and or units

During an analysis of a phase III trial, it was determined that the time from query generation, to time of query resolution, can take between 52 days to 23 weeks.  Although such a delay very seldom impacts the outcome of a trial, it has proven to be extremely resource intensive for site staff.

What can be done to ensure that data queries don’t slow down your trial?

During the analysis, it was shown that half of the queries were related to erroneous, illegible and missing data. This can be easily addressed through online training and support tools that are available for the electronic data capture (EDC) system and the project. You should make use of the online training and data entry tools available for your clinical trial before you start with data entry for your study.

One of the most obvious ways to reduce errors, is to improve the data collection.  If you see something that doesn’t make sense in the database, report it to your Clinical Site Manager (CSM).  This will eliminate the accumulation of issues downstream.  This also includes data entry from source documentation like lab reports.

If a datapoint is dependent on other data, make sure the data entry is done holistically. Sometimes data management might query the start and stop date for adverse events based on the start and stop dates for medications taken.  If the start and stop date is updated, make sure it still makes sense for the concomitant medication.

If you are uncertain on how or what to respond for a query, talk to your CSM. If the CSM is unclear of the query, writing to the data management team to clarify before giving an uncertain answer can also reduce the frequency of queries coming in and going out for a datapoint.

Before responding to a query make sure to update the datapoint in question. Sometimes the queries are answered and indicated that the data has been changed, but the site forgot to change the datapoint.  This can cause a re-query.

And yes, it’s completely possible that a data listing is reviewed, and the site has updated the data field in question after.  Data managers do check back to the database to determine if the query from the previous review, is still valid.