Patient-data preflight checklist
- Inventory the clinical database, attachments, scanned referrals, pathology/imaging interfaces, appointment and billing data, shared drives, email exports, local desktops, backup sets and portable media.
- Identify the owner, sensitivity, location, volume, retention requirement and access permissions of each copy. Include old exports and vendor-held data, not only the main server.
- Record application versions, database engines, licences, service identities, certificates, scheduled tasks, scanners, printers and remote access.
- Ask clinical vendors to confirm supported operating systems, database versions, hosting architecture, backup methods and migration responsibilities in writing.
- Map dependencies such as Medicare Online, HICAPS, My Health Record, secure messaging and pathology feeds. Confirm their actual use rather than assuming every clinic has every integration.
- Agree who may access patient data, how transfer is encrypted, where temporary copies will be stored and who verifies their removal.
Use a restricted inventory. Do not attach patient records or database copies to an ordinary email ticket to “show the provider what we have”.
Cloud, local replacement or hybrid?
Moving files into SharePoint or email into Microsoft 365 does not migrate a clinical application. A vendor-hosted clinical service, a hosted server and a new on-site server have different dependencies and support responsibilities. Do not put a live clinical database in a consumer synchronisation folder.
Compare vendor support, latency from consulting rooms, internet dependency, outage options, remote access, capacity, backup/export capability and total ongoing costs. A hybrid arrangement can keep a supported clinical application local while moving suitable documents elsewhere. Cloud hosting alone does not establish security or compliance; access controls, contracts, governance, recovery and data handling still matter.
Ask where primary data, backups and support access are located, who the subcontractors are, and what happens on exit. Australian hosting is not a blanket privacy guarantee. Obtain appropriate advice about overseas handling and record-retention requirements rather than relying on a “compliant cloud” label.
Prove recovery and pilot before scheduling the move
Have the vendor and technical lead create application-consistent backups using supported methods. Test a restore in an isolated authorised environment, checking database integrity and readable attachments with an authorised practice representative. Record the recovery point, elapsed time, failures and corrective action. A completed backup job is not proof of recovery.
Pilot representative workflows with test data where possible. Where real records are necessary, authorise and minimise their use and protect the test environment. Validate logins, approved roles, document scanning, printing and relevant integrations; avoid transmitting test claims or prescriptions into live services. Agree controlled test methods with each vendor.
Define acceptable downtime and data loss, the final go/no-go time and the people who can approve either decision. Recovery objectives are planning aims unless expressly contracted, not promises that this article creates.
Cutover and rollback checklist
- Before the window: confirm tested backups, a usable recovery environment, vendor availability, staff instructions, contact tree and the approved downtime workflow for patient care.
- Freeze: agree when users stop making changes, prevent conflicting writes and capture the final supported backup or synchronisation. Record the final transaction point.
- Transfer: use approved encrypted channels and named, limited administrator access; log actions and protect temporary staging data.
- Validate: confirm record counts or vendor-approved integrity checks, attachments, permissions, appointments, billing and interfaces before general access resumes.
- Go/no-go: pause if data integrity, access or critical workflows fail, or if the agreed deadline cannot be met. The named approver decides with clinical and technical advice.
- Rollback: use the vendor-approved return procedure and record which system is authoritative. Do not let old and new databases both accept changes.
- Reconcile: if the new system has accepted records, the vendor must plan safe reconciliation before reverting. A blind restore could discard care recorded after cutover.
- Communicate: tell staff which system to use, record downtime work securely and reconcile it through approved clinical procedures.
This is a hypothetical planning checklist, not a report of a Netluma IT clinic migration. No downtime duration or outcome is implied.
Acceptance, monitoring and eventual disposal
Get practice and vendor acceptance against written criteria. Monitor backup success, performance and integration queues after go-live, and resolve outstanding exceptions. Revoke temporary access and document where retained copies remain.
Keep the old server isolated and protected only for an approved recovery/retention purpose. Do not leave it as an unmanaged second live system. Before disposal, obtain the practice’s retention and legal-hold decision, then arrange suitable sanitisation or destruction and evidence. Disposal must cover disks and staging copies as well as the server chassis.
Provider questions and relevant guidance
Who owns each application migration and integration test? What evidence proves recovery works? Who authorises rollback and reconciles new entries? Which costs cover vendor work, after-hours assistance, follow-up and disposal? What can we export if we later change provider?
Discuss separately scoped cloud migration and backup and recovery. Use the permissions matrix before recreating access.
Official sources checked 26 October 2026: RACGP F8, backup testing and OAIC APP 11. F8 is published sixth-edition material; assessment transition guidance is separate, so confirm the applicable edition with your accreditation agency.
Responsibilities and next steps
Use these steps with the practice’s authorised decision-maker and relevant vendors. They are evaluation advice, not confirmed Netluma IT inclusions, prices, testing schedules or response commitments. See healthcare IT support and the RACGP F8 and APP 11 responsibilities guide to agree owners, approvals and evidence.
Netluma IT does not certify compliance or confer accreditation. Buying IT support does not guarantee accreditation or legal compliance. Practice governance and legal duties are not transferred to the provider. This is general information, not legal advice or an exhaustive compliance checklist.
Need IT Built for Healthcare?
Discuss practical IT support for your clinic, including agreed access, backup, connectivity and vendor responsibilities. IT support does not guarantee legal compliance or accreditation; practice governance and legal duties remain with the practice.
Related Services
IT SupportManaged ITCybersecurityPriority-based remote responseLocal Gold Coast team



