top of page
Search
john87833

How FAIR Data and ALCOA+ Work Together

by


Introduction

FAIR data principles were developed to maximize the value of R&D data for downstream uses, and implicitly assume an electronic format. In contrast, ALCOA requirements focus on regulatory compliance, especially as part of drug CGMP (Current Good Manufacturing Practice). Although they originated over thirty years ago when paper records were more the norm, FDA guidance has evolved to account for electronic formats. This evolution coincides with extending the five ALCOA requirements to the nine of ALCOA+.


FAIR Data Principles (Taken from the GO FAIR Initiative)


Defined by the GO FAIR Initiative, the four principles are:

Findable:

F1. (Meta)data are assigned a globally unique and persistent identifier

F2. Data are described with rich metadata (defined by R1 below)

F3. Metadata clearly and explicitly include the identifier of the data they describe

F4. (Meta)data are registered or indexed in a searchable resource

Accessible:

A1. (Meta)data are retrievable by their identifier using a standardized communications protocol

A1.1 The protocol is open, free, and universally implementable

A1.2 The protocol allows for an authentication and authorization procedure, where necessary

A2. Metadata are accessible, even when the data are no longer available

Interoperable:

I1. (Meta)data use a formal, accessible, shared, and broadly applicable language for knowledge representation.

I2. (Meta)data use vocabularies that follow FAIR principles

I3. (Meta)data include qualified references to other (meta)data

Reusable:

R1. (Meta)data are richly described with a plurality of accurate and relevant attributes R1.1. (Meta)data are released with a clear and accessible data usage license

R1.2. (Meta)data are associated with detailed provenance

R1.3. (Meta)data meet domain-relevant community standards


ALCOA+ Requirements Applied to CGMP Electronic Systems


Attributable:

Users must be uniquely identifiable by login credentials, which also control the ability to enter or modify data. Instruments automatically uploading data must also be uniquely identifiable (e.g., by serial number).Systems must create an audit trail for each data event.

Legible:

Data must be human-readable for review by the quality unit or, in case of inspection, throughout the retention period; therefore: Automated backups and a disaster recovery plan should be implemented to maintain the data and format.

Contemporaneous:

Data must be promptly recorded when generated; therefore: Events must be timestamped according to a standard reference, and timely entry must be enforced through SOPs.

Original:

Raw data must be recorded. Reports, calculations, or other analyses are metadata traceable to raw data (through audit trails).

Accurate:

Data must be error-free, therefore: Instruments uploading data must be regularly calibrated. Systems must provide for automated accuracy and discrepancy checks, as well as for human review. Users must provide a clear explanation when making any changes. Audit trails are again essential here.


If an electronic format is assumed, the additional four requirements of ALCOA+ reinforce the previous five and add a few additional constraints.


Complete:

All recorded data and metadata are retained, including sample retests or reanalysis. Audit trails are again critical.


Consistent:

Data must be in the expected temporal sequence. This is already covered electronically by the proper use of timestamps.


Enduring:

The last two requirements together essentially extend the legibility requirement beyond the retention period, through the lifetime of the study, and beyond.


Available:

Additionally, data must be appropriately indexed or labeled for searching.


Combining FAIR and ALCOA+


Applying FAIR principles in a pharmaceutical company’s chemistry, manufacturing, and controls (CMC) group may seem to clash with its ALCOA+ requirements. The two paradigms have different motivations: FAIR is oriented toward the external use of data beyond the organization, while ALCOA+ is focused on the internal use of data by the quality unit for FDA compliance. However, they are entirely compatible, and a FAIR data culture provides a foundation for robustly meeting ALCOA+ requirements. First, a commitment to FAIR brings with it a commitment to digitization. Moving away from clunky and increasingly obsolete paper methods will streamline the process of meeting ALCOA+ requirements. (It is important to note that “paper methods” include relying on MS O ce or other programs that merely replicate a paper format as an unstructured file). If data generated by R&D is FAIR, it will provide a framework for ALCOA+ conformance. The infrastructure already set up to support FAIR can be easily adapted to serve the needs of the CMC group. Many manufacturing process elements—including raw materials, intermediates, equipment, tests, units, and specifications for test results—will already be represented in a database schema for R&D, which can be readily adapted for use in CGMP. The association of FAIR data with rich metadata can be leveraged to collect the metadata needed to fulfill ALCOA+. Reusability’s detailed provenance specification lays the groundwork for ALCOA+’s audit trails to meet their goal of reconstructing the events generating any data. More specifically, the indexing specified under Findability can be applied to fulfill Availability. Similarly, the machine readability specified under Interoperability can be extended to include human readability, helping to achieve Legibility.


Final Thoughts


If ALCOA+ data is also FAIR, it will smooth the technology transfer process as production progresses and scales up. It will also allow all data and results from the CMC group to be reused alongside research data, internally and externally. In this way, the scientific feedback loop closes, enabling information generated by CMC for efficient reuse, stimulating new research, the development of new products, and improving existing products and processes.


We hope you find this opinion article useful and thought-provoking. Steve just scratched the surface of this complicated topic, and his understanding is always evolving. Please don’t hesitate to message us for further discussion and critiques, or if there’s any more direct help we can o er to support your laboratory informatics projects.


Steve Bates is a former bench scientist who conducted research at UPenn, Stanford, and MIT. Since 2016, he has worked as a laboratory informatics business analyst. He is open to consulting engagements, contract roles, or full-time positions.

19 views0 comments

Kommentare


bottom of page