Cleaning duplicate files before applying the clinics system

Duplicate files are a problem that silently accumulates: a client was registered twice with two different numbers, his history is divided, your reports are prepared by two people, and his entitlements are distributed. Migrating them as is transfers the clutter to a new system and makes it harder to clean up later.

Cleaning before migration is more efficient and accurate. But it requires rules, not diligence, and a documented merging decision, not silent deletion.

Scope of work before starting

First select your approved ID after migration: ID number, mobile number, or internal ID. Then determine who owns the merging decision — one or two people with specific authority, not everyone working on the data.

And decide on a principle before starting: merging does not delete. What is merged remains in the history, and it is still possible to know that these two files were separate, who merged them, and when.

Steps and responsibilities

• Extract filters (supplied): List of similar files according to agreed upon rules, not automatic merging.
• Review (you): An authorized member of your team reviews each candidate pair and decides: a merge, or two truly different files.
• Define the base file (you): Which one remains a master identifier, and what data is transferred to it from the other.
• Implementation (provider): Consolidation with transfer of appointments, sessions and dues, and saving the process history.
• Verification (you): Open the merged files and ensure their dates and dues are complete.

What needs to be tested and certified?

Rules that detect repetition, tested on a sample before generalizing:

• ID number matches: The strongest rule, but it is not enough alone because many old records are without identity.
• Matches mobile phone with name difference: Be careful here — families often share one number, especially with children. The shared mobile phone is not evidence of repetition, but rather evidence of a relationship.
• Similarity of name and similarity of date of birth: A strong filter that always needs human review.
• Quad name match: Is a candidate, but the similarity of names is common and is not enough alone.

A general rule was adopted: no automatic merging, regardless of the degree of matching. Incorrect merging mixes the history of two people, and this is a serious professional problem in psychological services that is more severe than keeping two separate files.

Risks and Exclusions

• Single Family: A mother and her children with the same mobile number and similar names. These are independent files linked by a relationship, not duplicate files.
• Name changed: The difference in name writing between two records does not mean two people.
• Irreversible Merger: Before executing, ensure that the log can be reviewed, and that the process is documented.
• Active cases in the two files: If each file has an ongoing episode of care with a different specialist, merging is a professional, not administrative, decision—review with supervisor.

And allocate realistic time for human review. In a center with two thousand files, dozens of filters may appear, and each pair requires a minute or two from someone who knows the data. This is work that cannot be accelerated by automation, and estimating it in the plan prevents the project from being stalled at this stage — which is one of the stages where projects are most delayed because it is estimated at less than it is worth.

The next step is to prepare the implementation plan

Ask to extract the filter list early — in the sample phase rather than after the full migration — and take time to review it. And review Files and cases page To know how the system prevents a new occurrence after booting.