In case you missed it, the first part of this series can be found here.

Now that we reviewed the top-level data integrity (DI) trends, we next get into the specifics of what the FDA is actually citing. This post will walk through our analysis and tagging methodologies, and we will share our preliminary results.

(NOTE: Looking for an easier way to research observations by specific keyword? Get your FREE 483 Observation Report and you can quickly determine who has been hit with the same observations. Learn more here!)

Get your 483 Observation Report

Our Process

We started by categorizing DI keywords into the ALCOA Plus grouping. ALCOA Plus is a common and well-understood framework that helped us make meaning of what the FDA was really doing.

ALCOA Plus includes (Figure 1):

  • Attributable
  • Legible
  • Contemporaneous
  • Original
  • Accurate
  • Complete
  • Consistent
  • Enduring
  • Available
ALCOA Plus includes

We used our database of 212 distinct keywords and phrases on data integrity (i.e. manipulation, access controls, privileges, etc.) and mapped them on top of the ALCOA Plus concept categories (Figure 2).

We used our database of 212 distinct keywords and phrases on data integrity (i.e. manipulation, access controls, privileges, etc.) and mapped them on top of the ALCOA Plus concept categories

Interestingly, the term “data integrity” itself is rarely cited by the FDA, especially in 483s, which is why researching all of the terms that encompass DI is necessary to understand it’s citation trends.  

As you might imagine, certain keyword terms align to multiple ALCOA Plus categories. For example, if there is a citation for “legible,” then it must have two categories of LEGIBLE and ACCURATE. Since if you cannot read it, then you cannot verify that it is accurate. (Figure 3 and Figure 4).

As you might imagine, certain keyword terms align to multiple ALCOA Plus categories.  For example, if there is a citation for “legible”, then it must have two categories of LEGIBLE and ACCURATE.  Since if you cannot read it, then you cannot verify that it is accurate.
As you might imagine, certain keyword terms align to multiple ALCOA Plus categories.  For example, if there is a citation for “legible”, then it must have two categories of LEGIBLE and ACCURATE.  Since if you cannot read it, then you cannot verify that it is accurate.

Preliminary Results for 483s

In 483s, the most cited ALCOA Plus topic was “Accurate” (see Figure 5 for all included terms). The FDA devotes more words in more 483 Observations describing how pharmaceutical firms do not have accurate data. This is not surprising, given that “accuracy” overlaps with so many other concepts.

Also, when you read the 483 Observation text, you will see how much effort the FDA takes to describe the nature of the inaccuracies. For example, buried on the 30th line of the Observation could be statements like “Your firm has no scientific rationale or supporting data to demonstrate equivalency.” In other words, it should come as no surprise that “Accuracy” is the top item!

In 483s, the most cited ALCOA Plus topic was “Accurate” (see Figure 5 for all included terms).

In the last part of this series, we’ll look at how DI citation trends change when it comes to warning letters and review takeaways from all the DI citation trends that we covered. Continue reading here.

Is there another analysis you’d like to see? Please leave a comment below.

(NOTE: Looking for an easier way to research observations by specific keyword? Get your FREE 483 Observation Report and you can quickly determine who has been hit with the same observations. Learn more here!)

Get your 483 Observation Report

Get a Demo

We’ll can show you insights into any of your key suppliers, FDA investigators, inspection trends, and much more.

Request a Demo