Jeff Vannest of LabVantage recently addressed the issues of new temporary data rules and data inegrity concerns in a recent guest commentary for DDNews titled "Should you be concerned about temporary data?" and, in the wake of that commentary, we wanted to follow up with a few related questions to continue the conversation. (We highly recommend you head back to the link in the previous sentence and read the commentary first, if you haven't already done so.)
Q: In the recent guest commentary from LabVantage for DDNews, temporary data was an area brought up as something that those in pharma and biotech R&D should be concerned about, given some recent rumblings by regulatory and health agencies globally. How much of a surprise do you think this new focus on temporary data was for the pharma world? And why or why not?
A: Given that analysts know keeping data on a sticky note is bad policy, it is logical to anticipate regulators wanting to eliminate the digital equivalent, hence tracking data in temporary memory. In addition, there have been occasions where users were cited for falsifying results and such violations—whether in temporary memory or not—may have led regulators to look at mechanisms for enforcement beyond SOP training that may not have been fully effective. So, no, it should not be a surprise to industry; that said, there are some who will wait until the guidance is formalized before taking action to preserve temporary data.
Q: What, in your mind, were the propelling forces behind agencies in the United States and elsewhere deciding that they needed to “crackdown” on temporary data?
A: In a word, risk. While data is in temporary memory, it is subject to modification that cannot be easily detected. Since one purpose of routine and periodic data review is to look for data integrity failures, any practice that obscures that failure must be avoided. In retrospect, the previous position that the record began when data was stored to durable media seems quite lax. So the new guidance is not so much crackdown as it is an evolution of position to protect the integrity of important data.
Q: Now, as the recent LabVantage commentary also notes by way of introduction, it’s only been a couple years since the WHO announced good data and record management practices and year since the FDA and MHRA issued draft guidance on data integrity. Give us the highlights, if you would, on other critical issues that pharma R&D folks need to pay more attention to with regard to data integrity and handling—and why?
A: The importance of data integrity in GxP has a long history, much longer than these updated documents. What has changed is the interpretation and application of the core data integrity principles. From a merely tactical perspective, the new guidance seems to be very challenging. For vendors that use a transactional database to store data, the idea that the integrity of data must be monitored prior to the user hitting the Save button is pretty difficult to grasp. But if you look at the core principles of data integrity, it’s a natural position. Labs should examine their processes for areas where data can be manipulated or discarded without accountability; for example, instrument files stored on unprotected file systems before processing and software that allows data to be manipulated into specification without traceability.
Q: Finally, what do you think are future areas—other than temporary data—in data management and handling that might become high priorities for regulatory agencies and the like, and that pharma R&D operations might want to get ahead of now?
A: It is natural to wonder how data integrity will be enforced within emerging technologies. For example, labs are asking more frequently for voice-activated data input, but anyone who’s used Siri to dictate an email knows how error-prone that can be. What steps are taken to ensure what was said is what gets recorded? Thinking about AI [artificial intelligence], how does a lab document the inputs from which a data point was deduced? While we probably don’t need to worry about these things today, exercising the core data integrity principles against these future problems may help us explore the boundaries of our existing systems and processes, where the risks to patient safety and product quality may be greatest.
Perhaps less futuristic than AI, but a need is the electronic capture of all information about the execution of test methods. LES —laboratory execution system—worksheets are becoming more prevalent, both to eliminate paper and drive more complete test-method execution. Most labs, even with a LIMS, still have some paper worksheets that are audited documents which need to be archived and maintained. Eliminating that paper and electronically capturing all the execution information, as well as the results, is a rising trend.