DB design:
Example:
Feed - p1 (create), p2 (create), p3 (create), p1 (updated occupation), p1 (updated nid2), p2 (voided, merged-with p3), h4 (create), p5 (create). pn represents HID.
All have same NID, nid1 but different catchments, c1, c2, c3, c4, c5. Also, h4, h5 have same phone #.
Rules: same nid and same phone # are dedups. Dedups cannot be resolved for just one rule, but for all.
| Scenarios | Action | Event in Feed | Scheduled job runs | Dupes in C1 | Dupes in C2 | Dupes in C3 | Dupes in C4 | Dupes in C5 |
| 1 | Patient created (P1) | created:p1 | No dupes found. | |||||
| 2 | Patient created (P2) | created:p2 | Finds P1 with same NID | P2-P1 (NID) | P2-P1(NID) | |||
| 3 | Patient created (P3) | created:p3 | Finds that P3 shares NID with P1 and P2 | P2-P1(NID) P3-P1(NID) + | P2-P1(NID) P3-P2(NID) + | P3-P1(NID) + P3-P2(NID) + | ||
| 4 | Patient updated (P1) | updated:p1 (occupation) | FInds matching NID with P3 and P2. But those combinations are already stored in dedup tables, so can we let it be? No adding new entries to tables. | P2-P1(NID) P3-P1(NID) | P2-P1(NID) P3-P2(NID) | P3-P1(NID) P3-P2(NID) | ||
| 5 | Patient updated (P1) | updated:p1 (NID) | Now then job runs, no dupes wil be found for P1. But we must also search the dedup tables to remove any pairs with P1 (reason NID match) that were earlier created. | Empty | P3-P2(NID) | P3-P2(NID) | ||
| 6 | Resolution. P2-P3 conflict for all rules. p2 - voided and merged with p3 | voided:P2 (merged with p3) | Since P2 does not exist anymore, we must remove all pairs with P2 from all dedup tables. | Empty | Empty | Empty | ||
| 7 | updated:P3 | Run deduplication rule anyway. Found no duplicates | Empty | Empty | Empty | |||
| 8 | Patient created (P4) | created: P4 | P4 matches P3 in NID. P1's NID has changed in earlier step. | P4-P3(NID) + | P4-P3(NID) + | |||
| 9 | Patient created (P5) | created:P5 | P5 is detected for NID duplication, matching with P3, P4. And Also with P4 for Ph. num | P4-P3(NID) P5-P3 (NID) + | P4-P3(NID) P5-P4 (NID+Ph.No) + | P5-P3(NID) + P5-P4 (NID+Ph.No) + | ||
| 10 | P3-P4 resolved for all rules, both are unique. | ignore-duplicate: P3, P4 | As soon as approver says these must remain separate, we must go back and take them out of all catchments | P5-P3 (NID) | P5-P4 (NID+Ph.No) | P5-P3(NID) P5-P4 (NID+Ph.No) |
Scenario 1:
Scheduler picks up p1 from feed.
Searches all patients by nid1. No matching patient. No further action taken.
Searches all patients by phone # ph1. No matching patient. No further action taken.
Scenario 2:
Scheduler picks up p2 from feed.
Searches all patients by nid1. Find p1. Creates the following db entry:
Add Comment