Teams that coordinate travel for patients abroad handle a lot of sensitive paper: passports, referral letters, scans, itineraries, payment confirmations. Whoever holds that material is a natural target, and so the question is not whether your information systems face pressure from attackers but whether you have a calm, repeatable way of tracking it. ThreatOps is such a routine. This post turns it into an audit checklist you can run in one sitting, with a pass, partial or fail mark for each point. It deals only with administration and information security; nothing in it touches medical decisions or evaluates any provider.
The checklist mirrors the structure of a long Ops Atlas reference on ThreatOps. That reference notes that the term is not formally standardised and that the underlying disciplines, intelligence and hunting, are the established part. Where I quote a number, it comes from that entry’s worked examples.
How to use the checklist
Bring whoever manages your IT, whoever owns data protection and one operations lead. Read each item aloud, agree a mark and write one sentence of evidence beside it. Every partial or fail turns into a task that names a person and a deadline. Plan on ninety minutes.
The ten checks
1. Is there a named owner?
Pass means one person is accountable for the routine, who has time set aside for it, even if that is only a half-day each week. Fail means the work is “everyone’s job,” which in practice means no one’s.
2. Are the standing questions written down?
Look for a register of five to eight questions, each tied to one decision. A question like “Which attacks are aimed at travel-document portals like ours this quarter?” qualifies. A topic like “ransomware” does not. Check that the register was reviewed within the last three months.
3. Is every source rated?
Open the source list. Each entry should carry a reliability letter and each report a credibility number, in the FIRST style. If the list is just a set of links, mark it partial and add the ratings.
4. Does every indicator expire?
Ask to see the oldest item in your blocklist or watchlist. Because the validity-end field in STIX is optional, entries often have no lifespan at all. A pass means a written policy exists with different lifetimes per type, and the oldest entry falls within its limit.
5. Are blocks gated and reversible?
Pass needs three things on paper: criteria for blocking (top-tier rating, measured precision over ninety days, not on an allowlist), a rule that anyone on call may undo a block, and a review within the hour. Without them, mark fail; one wrong block could strand an entire booking system on a busy day.
6. Is there a hunting calendar?
Look for a fortnightly cadence with one or two behaviours per hunt. Ask to see the last three. Each should have produced something filed: a request for new detection logic, a ticket about missing data, an escalation note for management, or a record stating the activity would have been visible.
7. Could a stranger rerun the last hunt?
Hand the write-up to someone who was not involved. If they can reproduce it, pass. If they have to ask the author how, partial.
8. Are sharing rules explicit?
Check for labelling under the Traffic Light Protocol, version 2.0, and a note on which classes of information may go out without approval. The Atlas points out that material marked for a restricted circle should not travel past it. It also mentions that European law, in a clause of the NIS2 directive, allows voluntary sharing, and that a set of United States protections is due to lapse on 11 December 2026, which makes it sensible to verify where things stand before leaning on them.
9. Do you know your own telemetry?
Ask how many days of logs exist for your main systems. If the answer is under ninety, the decision tree in the Atlas puts data repair ahead of everything else, since hunting through thin records breeds false comfort. Check whether every site or system keeps the same retention, as gaps tend to hide in the smaller ones.
10. Are the metrics honest?
Look for a short quarterly scorecard. The Atlas worked example uses coverage of questions (83 percent, five of six), the share of intelligence that changed an action (35 percent, 14 of 40), the share of hunts that led to an artefact (42 percent, five of twelve) and gaps closed (44 percent, four of nine). Your numbers will differ; the point is that they exist and measure results. Mark fail if the scorecard only counts items ingested or hunts started.
Scoring
Give a pass two points, partial one and fail zero. Sixteen or more suggests a routine that is working; eight to fifteen means the foundations exist but need tightening; below eight, start with checks one, two and four, because they cost almost nothing and unlock the rest.
Related reading on the site
Because coordination teams constantly ask and answer checklists of their own, two articles may be handy beside this one: a list of questions people are encouraged to ask before travelling for surgery and a guide to verifying hospital accreditation before going abroad. They are about different questions, but the habit of asking before acting is the same.
After the audit
Take the three most valuable gaps and schedule them over the next quarter. Do not try to fix everything. Repeat the audit in three months and compare. A routine that scores modestly but improves is far healthier than one that scored well once and was never looked at again. This checklist is general information about security administration and does not replace advice from your own data-protection and security professionals.