Reply To: Stability Data Reporting Requirements?

[John O’Neill]

#25031

(Response posted in behalf of Laura Pack, a well-respected industry stability statistician)

As to the data formatting, here are my thoughts / typical practices (feel free to post if you like):
– I don’t know of any guidance on stability table format specifically
– My general principle is to provide the least amount of information possible – from a lifecycle approach, this makes it easier to maintain tables when updated in case information changes throughout the course of the study (e.g. specification acceptance criteria)
– In data tables, we typically provide:
o Lot number
o Formulation / presentation
o Storage condition
o Method, attribute
o Acceptance criteria (ONLY for clinical filings and ONLY if regulatory affairs requires). Otherwise, I point to current specs in S.4.1 / P.5.1 and footnote attributes removed / limits changed for studies that don’t meet current AC
o Data formatted to specification reporting precision
– YES, I have used JMP in S.7.1 / P.8.1, but usually only in commercial applications:
o Provide regression analysis for expiry, including plots and a top-line conclusion that each attribute’s regression supports your requested shelf life
o Use ‘recorded values’ for stats analysis – the unrounded, recorded results with more decimal places are used for data analysis. Note that these recorded values are not provided in S.7.3 / P.8.3
o The raw data files for stats can be provided in ts.xpt format, similar to the way FDA expects clinical trial data to be submitted.
 FDA did request our raw stability data to be submitted in one of my recent commercial filings (it was fast-tracked and part of a few pilot programs)
 See https://www.fda.gov/industry/study-data-standards-resources/study-data-submission-cder-and-cber for data submission standards (applies to clinical trial data, but can be used for stability data files too)