Feed
Everything published here, in date order, including the policy changes.
Most sites syndicate their blog and leave their legal documents to change quietly. A revision to a privacy policy is the publication a laboratory actually needs to know about, so it is in the same stream as the writing.
Six entries is the honest number. A feed padded with entries nobody wrote is worse than a short one.
Subscribe with any reader.
This page is for reading. The file below is the same content in the format a feed reader understands, and it is generated in the same build from the same list, so the two cannot disagree.
Paste it into a reader. Selecting the box selects the whole address, so it works without the button too.
All entries, newest first
All eight legal documents rewritten
Every legal page replaced, in one pass, with documents that state what LabFlow does and what it does not do rather than with a template. Three things changed that are worth knowing about even if you do not read them: the word used about HIPAA is now conscious everywhere and never compliant, the four retention windows are published with what each is anchored to, and every third-party company that sees anything is named with what it actually receives.
The terms and the privacy policy also carry a new clause about using non-clinical content from free accounts to improve features, with patient records, orders, results and all clinical content explicitly carved out on every plan.
Westgard rules, without the mnemonics
Six rules, what each one is actually looking for, and the arithmetic that explains why running all of them on every analyte produces more false rejections than real ones. Includes the plot, and the reason the one-two-s rule is a warning rather than a rejection.
About 11 minutes.
HL7 v2 or FHIR: which one to build first
The honest answer is v2, and the reason is not technical merit. It is that the systems you have to talk to this year speak v2, and the pipe that carries it is a separate problem from the message inside it.
About 9 minutes.
LIS or LIMS: the distinction that decides your data model
One is organised around a patient and the other around a sample, and the two produce different tables, different permissions and different reports. Buying the wrong one is usually discovered at the first accreditation assessment.
About 8 minutes.
Tenant isolation in a multi-tenant LIMS
Why a list query has to prove its own tenant from its own filters instead of trusting the caller, why a policy that reads a field the query never filters on is a leak that returns a clean success, and what a shared patient record does to the whole argument.
About 12 minutes.
Offline-first phlebotomy, and the conflicts nobody plans for
A home visit and a hospital basement are the two places a phlebotomist reliably loses signal, and they are also the two places where guessing which write wins is unacceptable. What queues, what cannot queue, and why a conflict has to reach a person.
About 10 minutes.
Four things a feed usually carries, which this one does not.
Each of them is absent for a reason worth stating, because a short feed with no explanation reads as neglect rather than as a decision.
The product is being rebuilt and nothing has shipped since that started, so a changelog would have nothing true in it. It joins this feed with the first release, and the entries will name what changed rather than say "improvements and bug fixes".
They carry no publication date, and a chronological stream needs one. Rather than invent dates so the feed looks fuller, they stay on the comparisons page, which is where they are meant to be read.
There is nothing to announce, no customers to announce it to, and no email list. If one ever exists it will be opt-in and it will say what it sends before you give it an address.
No tracking pixel, no per-subscriber link rewriting, no open counting. A feed reader fetching the file is not identified, and the addresses in it are the same addresses on this page.
A policy that changes quietly has changed without you.
The reason the legal documents are in this stream at all is that a laboratory evaluating a vendor has to be able to answer one question later: what did this say when we signed, and what has moved since. A feed answers it. A page carrying only a date does not.