| Comment Record|
Mr. Adam Sherlock ||
2002-11-27 08:56:18 |
The Synapse Partnership Ltd |
| Comments for FDA General |
1. General Comments
2. Scope, Paragraph 4:
Is it the intention that this guidance, once finalised, would augment 21 CFR Part 11 with regard to Record creation?
4.1 What does Part 11 require?
It is currently an industry strategy to retire systems that cannot be made compliant with Part 11. Based upon this guidance, is it the agency’s view that systems should be fully ‘Part 11 compliant’ before being retired to enable the record retention requirements to be satisfied? In short, is the retirement of non-compliant systems an acceptable strategy?
5.2 Factors that might affect the reliability of electronic records during the required retention period should be identified and controlled
Although the factors identified state ‘not limited to’ it may be of value to acknowledge that a key factor affecting reliability of records retrieval will be the availability of trained staff with appropriate experience in using the systems on which the records rely.
5.3 Continued availability and readability of electronic record information should be ensured
In the first sentence: rather than simply a ‘representative number of electronic records’ Synapse considers it important that this is augmented with ‘across a representative number of media types’.
5.3, Paragraph 3: ‘if you find that you are starting…’.
A strong, risk-based approach to reading of electronic records should be employed to ensure that no difficulties are encountered in reading. This guidance information on ‘starting to have difficulty’ may be true of paper, e.g. fade, but electronic records are more likely to be ‘readable’ or ‘not readable’ without any gradation. Furthermore, certain errors in electronic records are not readily detectable by simple reading. Error checking software should be employed and implemented on a pre-defined basis following a documented record retention plan.
5.3, Final Paragraph
In addition to proposing that records be kept on different physical drives it would also be prudent to utilise different types of archive media.
5.5 The ability to process an electronic records information throughout its records retention period should be preserved, paragraph 2
Is it the express intent of this section of the guidance to indicate that companies ‘mimic’ or reproduce a core of functionality from the old
records system in the system to which records are migrated or which is used as the record recovery system during the retention period? If this is
the case, should companies retain all functions relating to data processing for records required under predicate rules? Similarly, on what basis should the decision to retain functions or not be made?
6.1 The Time Capsule Approach
The agency indicates that the time capsule approach involves preserving an electronic record on the same electronic media and computer system. In the case of many systems, vendors warrant the use of a particular version of software across a range of operating systems and hardware types. In this case would it be acceptable to use the word ‘equivalent’ rather than ‘same’ as the overall effect: the ability to retain records over the retention period would be unaffected and, in fact, may be made more possible (with appropriate validation and control of changes to the system concerned)? In cases such as this would a ‘Time-Capsule/Migration Hybrid model’ term be useful?
Similarly, the time capsule approach may allow for those changes in the computer system environment that do not materially affect the records. It is generally accepted across industry that changes, such as application of patches, would not be regarded as a migration of the record.
It may be more appropriate to indicate that the system in a time capsule approach should be maintained in a manner which is consistent with its original use.
Paragraph 2: ‘routine upkeep’.
How can ‘routine upkeep’ be reconciled with the preceding paragraph requiring ‘no changes to the computing environment’? This should be clarified.
Paragraph 3: ‘limited practicality for long term maintenance’.
In considering ‘practicality’, the relative merits/demerits of the Time Capsule and Migration approach should be clarified. It is Synapse’ view that the rapid pace of technology changes supports, rather than limits, the use of a Time Capsule approach for longer record retention requirements. With rapid technology changes the cost of records migration over the entire record retention period increases in parallel with each change to the production system and cannot be accurately predicted over the retention period (as the number of system changes cannot be known for long retention periods). Using a Time Capsule-based approach the cost can be predicted for the entire retention period.
The Time Capsule approach, with appropriate controls (supplemented by migration as a strategy where a system cannot be supported in-house or by suppliers) may also be particularly viable for high cost systems that are, or were the ‘industry standard’, e.g. Clintrial. In this case, companies may choose to co-operatively retain a single example of a given system to provide a records recovery system over the retention period. This is, of course, dependent on these companies retaining the ability to support the system after the vendor has ceased to provide support.
6.2 The Electronic Records Migration Approach
To fairly balance Time Capsule versus Migration the guidance should acknowledge the technical complexity and unknown cost (over the entire record retention period) of migration if it is to meet agency expectations.
Where records are migrated to a ‘new’ system this guidance indicates that the ‘new’ system may only be ‘acceptable’ to the agency where that system retains the ability to process the old records etc. It would be of value to stress this in this section.
Paragraph 2 ‘Your ability to solve the problem…..’
This suggests that a Time Capsule may be a sound approach to couple with migration. Is this the intent?
Paragraph 4 ‘Changes in factors that affect how reliably…’
The examples of such changes should include ‘core functionality of the ‘new’ system with respect to its ability to process data’.
18.104.22.168 The Ability to process information in electronic records should be preserved
By inference this section indicates that all systems used on the entire ‘migration path’ of a given electronic record throughout its records retention period should retain ‘old’ processing capability. Whilst it is acknowledged that this is a key requirement, this is likely to present the greatest challenge to the use of new systems in new business processes. From this guidance, when deciding User Requirements for new systems the owners should (a) ‘look forwards’ to determine what the new system should be capable of for the new business process and (b) ‘look backwards’ to require the retention of existing processing functions. It may be highly challenging to reconcile these potentially conflicting requirements.
Overall, where a ‘new’ system follows an existing business process, migration may be a viable option. However, where significant changes to the business process are required in the ‘new’ system the cost of retrofitting old functionality may be prohibitively high and adversely impact performance. In such cases the Time Capsule approach may be the most viable solution as it will allow the company to move ahead without any negative legacy from the ‘old’ system. This may also be the case where particular technology ‘classes’, e.g. Relational Databases, are replaced as industry standards.
22.214.171.124 Unavoidable Differences and losses should be accounted for and explained in the migrated electronic record or new system documentation.
Final Bullet ‘The above trusted third party’
In order to be able to demonstrate continuity is a sample of verified signed records acceptable or is 100% verification required?