Issue link: https://www.e-digitaleditions.com/i/1302985
36 Pharmaceutical Technology Regulatory Sourcebook October 2020 P h a r mTe c h . c o m Data Integrity and with separation of roles to avoid conf licts of interest, then the protected electronic records and associated reports ensure data integrity. These features are not usually found in hybrid and pa- per-based records. It is the analytical methods and records shown in yellow in Figure 3 that are the central to the authors' question: "Do pharmacopoeias inadver- tently facilitate data integrity violations?" These analytical methods and records vary from simple observation (e.g., color, appearance, odor, classical titration) to instrumental analysis data recorded on paper (e.g., LOD). While the main emphasis of the regulatory guidance documents listed in the introduction have focused on computerized systems, it is the simpler analytical methods that have a high risk of data integrity violations, be- cause they are mandated in the pharmacopoeias, where they lack objective evidence of testing and an effective second person review. For example, in many laboratories, the second person review of these tests merely checks if the manual observations have been recorded or cal- culated as required by the applicable procedure. In the case of classical pharmacopeial methods, such as LOD, specified in the monograph, the manual method outlined in USP <731> (16) has significant data integrity vulnerabilities. Data integrity Some of the data integrity issues surrounding pharmacopeial test results that are based on cap- turing the observed result by manual recording are as follows: • What is the risk and impact of the test result? Many of these results are release tests, but what is the risk weighting of appearance compared with, for example, the impurity profile of an API or a potency test of the finished dosage form? Therefore, how much effort should be placed on ensuring the integrity of the result? • Visual acuity. For color tests, is color blind- ness testing for all analysts conducted to en- sure consistency of results? Qualifying ana- lysts should be the basis for objective testing of color tests. • Do we need objective evidence such as pho- tographic records for some tests, and is color corrected software required? Some melting point instruments can record a video of each test so that there is objective rather than sub- jective evidence of testing that ensure the in- tegrity of the result. • Classical methods that do not generate elec- tronic records (e.g., LOD) rely on printouts, which requires that the analytical balance is controlled. • Classical methods that generate hybrid records are primarily paper based but may have elec- tronic records (e.g., auto-titrator procedures linked to a network with manual weighing and sample preparation). • Uncontrolled blank forms used for recording analytical results and have the potential to allow testing into compliance must be con- trolled and reconciled. This practice is a re- quirement in several data integrity guidance documents (4, 5) and has been discussed in publications by the authors (19, 20). Data are required to be attributable, legible, con- temporaneous, original, and accurate (ALCOA) as well as complete, consistent, enduring, and avail- able (ALCOA+). These requirements need to be considered throughout any analytical procedure