Changing your hospital system without losing patient history
The first question every hospital asks before replacing its system is simple: what happens to our patients’ files? Here is a practical plan, and an honest account of what Nabd imports today.
Hospitals rarely change their information system because they want to. They change because the old system no longer fits: it cannot handle insurance properly, the accounts never match the cashier, or the vendor has stopped answering. Yet the fear of losing years of patient history keeps many teams on a system they have outgrown. That fear is reasonable, and it is also manageable. A migration fails when it is treated as one big copy on the last weekend. It succeeds when it is planned as a series of small, checked steps.
Start with an inventory, not with a file export
Before anyone exports a single row, list every place where patient and operational data lives today. In most Egyptian hospitals that list is longer than expected:
- the current HIS or clinic software (patient register, visits, invoices);
- Excel sheets kept by reception, the lab or the insurance office;
- the pharmacy and stores program, often separate, with its own item codes;
- the accounting package with receivables, payables and opening balances;
- paper files, scanned reports and PDF results stored on a shared drive.
For each source, write down who owns it, how current it is and what you actually need on day one. Not everything has to move. Ten years of visit notes can stay readable in the old system; what reception needs on the first morning is the patient, the file number and the contact details.
Decide what moves, and what stays readable elsewhere
A useful rule is to split data into three groups. Master data must move: patients, items, insurance payers. Opening figures must move once and be reconciled: stock on hand per batch and expiry, and account balances on the go-live date. Historical clinical documents are a separate decision.
For that last group, be realistic. Converting old free-text notes and scanned reports into structured data is slow, expensive and rarely accurate. The common and honest approach is to keep the old system available in read-only mode for reference, or to attach important documents (discharge summaries, operative reports, imaging reports) to the patient file in the new system when the patient returns. Nabd does not convert old clinical notes automatically, and we tell hospitals so at the first meeting.
Map the columns and clean the duplicates first
Most problems in a migration come from the data, not from the software. Before the first import, agree on a mapping: which column in the old export becomes which field in the new system, and how coded values are written. Then clean the patient register, because duplicates are where history gets lost.
- Search for patients sharing the same 14-digit national ID, or the same name and mobile number.
- Decide which record is the surviving one and keep the other file numbers in a note.
- Fix obviously wrong national IDs and phone numbers; the new system will validate them.
- Agree on one spelling convention for Arabic names where the old system stored variations.
Nabd helps here in two ways. Registration validates the Egyptian national ID and mobile numbers, and the patient register detects possible duplicates and supports an audited merge after go-live, so a duplicate missed during cleaning can still be corrected later without losing either history.
How the import works in Nabd today
Nabd has a dedicated data import screen, held by the system administrator role by default. It reads Excel (.xlsx) and CSV files and imports, in a suggested order: items, insurance payers, patients, opening stock and opening account balances. You can see the screens on the features page.
The old file number stays with the patient
Each imported patient gets a new medical record number, but the old file number is stored with the record, stays unique, is searchable from reception and appears on the patient banner. A patient who arrives holding an old card is found immediately. Rows whose old file number was already imported are skipped, so a large export can be split into several files and re-run safely.
The whole file is checked before anything is written
Every run starts as a check of the whole file. Errors point at the exact row and column, with the offending value and a reason: a missing required column, an invalid national ID, a duplicate inside the file, a payer code that does not exist. Applying the file writes nothing while any row still has an error, and then writes everything in one transaction. There is no half-imported state to clean up.
Arabic exports are read as they are
Old Egyptian systems export files with Arabic column headers, Arabic-Indic digits and day-first dates. The import recognises the usual Arabic headers such as رقم الملف and الرقم القومي, reads Arabic-Indic digits and dates like 21/03/1985, and accepts common words for coded values (ذكر / أنثى, blood groups). CSV files saved by old Arabic Windows programs are read as well.
Every import leaves a trace
Each imported patient gets one audit entry, so you can show later where a record came from. Opening stock and opening balances post accounting entries on the chosen opening date, and the same file cannot be posted twice: the system recognises it and refuses it already at the preview.
Rehearse on a copy, then run in parallel
- Import a first sample (a few hundred patients) into a test installation and let reception search for real cases.
- Import the full files into the test installation and compare counts with the old system: patients, items, stock value, trial balance.
- Train each role on the test installation with the imported data, so staff recognise their own patients and items.
- Fix the source data, not the result, and repeat until a full run has no errors.
- Choose a cut-over date, ideally the start of a month for the accounts, and freeze changes in the old system for master data.
- Run the final import into production, check the counts again and open the doors.
Many hospitals keep the old system open read-only for a few weeks after go-live. Staff look up old notes there while every new visit, order and invoice is made in the new system only. Running two live systems in parallel for new transactions usually creates more errors than it prevents.
What is not imported yet
Honest status: Nabd imports patients, items, payers, opening stock and opening account balances today. Importing patient insurance cards, allergies and problem lists, staff and user accounts, and open patient or payer invoices as billing documents is planned for the next stage. Until then, those are entered during implementation or at the patient’s next visit.
One detail matters for finance teams: patient opening balances imported through the opening-balances file reach the general ledger with the patient as party, but they are not billing invoices, so they do not appear on the cashier screen. Agree with your accounts team how open patient balances will be collected after go-live.
Your data stays yours after the move
Migration has a second side that hospitals often forget to ask about: leaving. In Nabd, an administrator can export all data at any time, as one archive with every table as CSV plus the stored files, and the licence never deletes, hides or locks patient data. Ask any vendor the same question before you sign. Our guide on choosing a hospital system lists the other questions worth asking.
A short checklist before you start
- Inventory of every data source, with an owner for each.
- A written mapping of columns and coded values.
- A cleaned patient register with duplicates resolved.
- A test installation, a rehearsal import and reconciled counts.
- A decision on old clinical documents: read-only system or attachments.
- A cut-over date agreed with finance, and training done on real imported data.
If you are planning a move, we can walk through your current exports in a demo session and show the import screens on fictional data. You can also read how Nabd fits a complete hospital management system project.