When identifying and reassessing cases in response to a system-wide change, typically the requirements for processing fall into one of these categories:
All cases affected by the change must be identified and reassessed (i.e. all reassessments which are sufficient to make those cases consistent with the change). It is not acceptable for any case affected by the change to be "missed" during batch processing.
Only cases affected by the change should be identified and reassessed (i.e. only those reassessments which are necessary should occur). It can be wasteful to spend system time either reassessing cases which are not affected and/or reassessing any particular case more than once.
It must be possible to reassess cases in parallel batch streams (to promote scalability). It must be possible to reassess cases in batch while the online application is being used (to avoid the necessity for system downtime for online users).
You may have your own business-specific requirements to control reassessment of cases and/or perform additional processing, e.g.:
In practice, there are trade-offs to be made to meet these requirements. Driving the Identification of Affected Cases describes the various approaches available for handling a bulk change, i.e. for identifying the cases identified so that they can each be reassessed.