Data Integrity Matters | The Human Factors

By August 25, 2017

3 signs that you need to create a culture of compliance

As mentioned in my first post on data integrity, people trust that companies create safe, quality products and rely on data to support that belief. But most product data, especially laboratory testing, should be designed to catch studies, batches, or test samples that fail to meet specifications. A test is no use if it cannot raise an alarm when something is not right.

Of course any company is driven to create good quality products – but the purpose of testing is to ensure a consistent level of quality, not simply to create data that “proves quality.” Data that shows perfect answers all the time can also raise concern of trust: “Are you sure? It passes all the time? Can I believe it?”

Quality departments are often seen as a necessary cost to a business, something that has to be done to meet requirements. Their true value to the business’s reputation and brand in preventing costly or dangerous errors is often missed. Similarly, test results that could indicate poor product or product performance should be celebrated, but in order to meet company expectations – they’re often denigrated.

Does your laboratory show these early warning signs of Quality problems?

⚠ Common reactions to an individual creating results that fail:
• “What did you do wrong?”
• “Go back and try again.”

An individual analyst may feel pressure to avoid presenting “failed tests” and be tempted to try again to get the right result. But when the behavior becomes the modus operandi of a Quality Control department or a clinical study director, public safety is seriously put at risk.

In many regulatory observations, the analyst staff may not even be aware that “fudging the result into specification” is a serious quality issue. Oftentimes a responding comment is that this practice “is normal.” If everyone who trained the analyst and who works around them thinks it is expected, then it becomes the norm.

⚠ Performance measurements can indicate a problem:
• Increasing the number of batches passed by the laboratory
• Reducing the number of OOS results
• A 100% pass rate
• Reducing the turnaround time of tests

On the other hand, analysts may also be pressured to follow pointless rules.

⚠ Arbitrary laboratory rules should also raise warning flags:
• “No manual integration”
• “No single-test injections permitted before you start an HPLC sequence”
• “You must get QA approval to reintegrate a chromatogram”
• “If you need to repeat a run, you must complete this 40-page report first”
• “You must print out every method, every chromatogram, every integration and all the audit trails to submit to QA”

Raising these kinds of barriers to productivity simply drives analysts to find creative way to obscure their own errors from view.

And finally, we should acknowledge the individual pressures that management and supervisors may put on their teams to pretend those results never happened. In the major Ranbaxy fraud case (2013), senior directors who were aware and uncomfortable with “less than ideal study data” resigned from the company rather than stand up to the executives who were financially motivated to ensure new products were approved.

Much has been written about the need for a Culture of Quality but equally important is nurturing a Culture of Courage, where staff can confidently raise concerns about how common laboratory testing practices can be improved or may be being misused, with immunity and without fear of retribution.

Educating employees about the “Why?” of data integrity as well as simply enforcing it should encourage those who are deep in the details of a process – whether laboratory testing or manufacturing – to be involved in continually improving the quality and integrity of the data they produce.


More resources:

  • Live WebinarBalancing Technical Controls, Tools, Transparency, and Trust for a Culture of Data Integrity
    (September 22, 2017  | 10am EDT  | 9am CDT  |  3pm BST  | 4pm CEST )
  • On-demand webinar: Why is Electronic CDS Data a Major Data Integrity Concern for Regulators?
  • What Does Compliance Mean for Laboratory Computerized Systems? White paper  | Infographic
  • Learn more about ensuring Quality through compliance with Waters

< Previous
Next >

Categories: Data Integrity