Step 2
Search & Filter Stage
Purpose:
Generate a first‑pass list of possible programme candidates using what the EMR
currently contains.
When this is appropriate:
This step is most appropriate where Step 1 shows reasonable concordance between
recorded activity and what would be expected from national averages.
What is done:
• Build EMR search cohorts using:
– Existing diagnostic coding
– Structured data fields
– Prescribing-pattern cohorts
Generate three working lists:
– Possible CDM candidates
– Possible PP andidates
– Possible OCF candidatesLimitations:
• Lists will include some false positives and miss some true positives
• Patients may be missed if diagnoses exist only in correspondence or free-text
notes and were never coded
Output:
• Three working "possible candidate" lists
• Commentary on how reliable these lists are likely to be in this specific
database
