Curaeon Help Centre / KB-064
Open in the Help Centre →  ·  All topics
KB-064Security & privacyHow-to
Draft. This article is awaiting technical review and may change — if anything here conflicts with advice from our team, follow the team.

Sign off the weekly Security review, and act on alerts and audit-log check warnings

Ten minutes each Monday that gives your practice accreditation evidence, and a clear list of which warnings on the audit-log strip mean "contact support".

Before you start

You need to be an administrator. Settings → Security review is ready every Monday with the week before (and each whole week missed while the server was off, up to a year back). Nothing in it accuses anyone — an after-hours open may be the on-call doctor. It puts the log in front of someone who knows the practice.

Sign off the week

  1. Open Settings → Security review.
  2. Read each section — failed sign-ins and lockouts; sign-ins from a new address or outside hours; exports and backup downloads; documents sent out of the practice; records opened after hours or with no purpose given (KB-062 — Why Curaeon asks your reason for opening a record); one patient's trail looked up in the audit log; emergency access and refusals (KB-061 — Restrict a patient's record, and open one in an emergency); records of possible relatives; a hundred or more patients opened in a day; permission changes; rare changes such as a record destroyed or a legal hold released; and nightly audit-log checks that failed or did not run.
  3. A name or patient opens the Audit log on that week only, if you need to look further.
  4. Sign it off with a note of anything you followed up. A week with something in it asks for that note before it will sign — a signature with nothing beside it records only that somebody clicked, and it is what an accreditor is shown. A quiet week signs off with nothing to say.
Settings → Security review: alerts to acknowledge at the top, then each week waiting "To review" with a one-line summary of what it holds.
Settings → Security review: alerts to acknowledge at the top, then each week waiting "To review" with a one-line summary of what it holds.

If you appear in the week yourself, you may still sign it — with one administrator there is nobody else — and the page and the audit log both say so. Print gives a kept week on the letterhead; Show older weeks reaches back as far as the log goes.

Act on an alert

Alerts come at once, not on Monday: a burst of lockouts, an export by someone who is not an administrator, emergency access to a restricted record, or an audit log that no longer verifies or could not be checked. While one is unacknowledged, every administrator's sidebar carries Security alerts with the number on it; Curaeon sends nobody a message.

  1. Open the alert and find out what happened.
  2. Acknowledge it with what you found.
  3. If personal information may have gone where it should not, start an assessment from the alert (KB-065 — Assess a possible data breach: the 30-day NDB and 12-hour Services Australia clocks).

Your dashboard shows the last seven days as tiles; any alert or review waiting comes first.

The strip on the Audit log

The strip at the top of Settings → Audit log says when the chain was last checked (nightly at 02:30) and whether it holds, with the last record sealed and the start of its fingerprint. Write those two down on the weekly sign-off — an anchor nobody with access to the server can change. Check now runs the check at once.

The strip at the top of Settings → Audit log: when the chain was last checked and whether it holds, the last record sealed with the start of its fingerprint, and Check now.
The strip at the top of Settings → Audit log: when the chain was last checked and whether it holds, the last record sealed with the start of its fingerprint, and Check now.
The strip says What to do
The chain does not hold, naming the record Someone changed the database directly. Keep the backups as they are and contact support. It stays on the strip until its alert is acknowledged with what was found.
A backup was restored, and at which record Expected after a restore. Acknowledge the alert with who restored it and why. If nobody restored a backup, contact support.
A copied file is missing, unreadable or does not check out The records are fine; the copy outside the database is not. See to the backup folder (KB-067 — Set up nightly backups, a second copy and a monthly restore drill).
The nightly check is not running More than 26 hours since the last check. Press Check now; the nightly check retries every quarter hour.
The server runs as the login that owns the tables An older installation or a test server. Tell whoever runs the server.

Still stuck? Raise a ticket at support.curaeon.com.au or call 1300 XXX XXX. If your clinic can't see patients right now, call and choose option 1. Support is staffed Monday to Friday, 8:00–18:00 Sydney time; outside those hours a call or text to the same number is answered on a best-effort basis. - KB-127 — Connect an outside system: create, scope, rotate and revoke API keys and webhooks in Settings → API access — Connect an outside system: create, scope, rotate and revoke API keys and webhooks in Settings → API access