Currently within the Policy engine of CloudHealth you can define a policy to alert when a Service Principal moves to Critical or a Warning state.
This is useful to ensure that the Secret for the Service Principal remains current to allow CloudHealth to ingest billing data, and asset metadata.
As a best practice please set the evaluation time for a policy covering a Service Principal moving into Warning or Critical state to a daily cadence e.g.
This will ensure that the policy evaluates daily, and alerts you as quickly as possible to the Service Principal moving into a state that will block collection of billing data / asset metadata.