Care management
A treatment plan is a list of things that have not happened yet
The plan is agreed in the room and written into a note as prose: eight weeks of therapy, a recheck at twelve, a referral, a medication, a work restriction reviewed at six. From that moment it is inert. It is not a structure anything can query, nothing counts down against it, and the question of whether any of it actually happened has no owner. It is usually answered months later by a person reading backwards through a chart.
The plan becomes objects that can be counted
Each element of a plan is recorded as an expectation: a thing due, a window it is due in, a definition of what closes it, and a party responsible for the next move. A course of therapy is not one expectation but a count with a cadence. A repeat test is an expectation that names the test, not a reminder that names a date.
This is deliberately not a task list. A task is something a person ticks, which means the record shows that somebody clicked rather than that something occurred. An expectation is closed by evidence — an encounter that took place, a result that arrived, a referral that was accepted — and where no such evidence exists it stays open and says so.
The diagnostic loop closes itself
This is the part that only works if results are structured, and it is the reason the capability sits on this platform rather than beside it. “Recheck in twelve weeks” is an expectation that a discrete, LOINC-coded result closes by itself when it arrives from any connected laboratory, in or out of the practice, with the right code inside the right window. Nobody reconciles it and nobody ticks anything.
A result filed as a PDF cannot close anything, which is the same failure the rest of this site is about, arriving in a different place.
| Expectation | Closed by | Left open, it means |
|---|---|---|
| Repeat or interval testing | A discrete result with the expected code, inside the window | A monitoring plan that stopped being monitored |
| Therapy course, by session count | Encounters recorded against the course | A course that was prescribed and abandoned partway |
| Specialist referral | Acceptance by the receiving organisation, then a returned consult note | A referral that was made and never landed anywhere |
| Imaging ordered | The study performed and its report returned | An order sitting unscheduled or unreturned |
| Critical or malignant result | Acknowledgement by the ordering clinician, then a documented next action | The most dangerous open item in the system |
| Medication started or changed | Fill or administration evidence where the practice has access to it | A change nobody confirmed was made |
| Work restriction review | A review encounter on or after the review date | A restriction running past the point anyone assessed it |
| Post-discharge or post-procedure contact | A completed contact of the type specified | The window in which deterioration is most likely, unobserved |
Gaps surface, they are not discovered
An expectation that passes its window becomes a visible condition the day it passes, on the same worklist that carries outstanding orders, unacknowledged criticals, pending send-outs, and unmatched results. It is not a monthly report somebody runs, because a monthly report is a way of finding out about a six-week gap in week six.
Conditions are ranked by consequence rather than by age. An unacknowledged malignant result outranks a missed therapy session, and no amount of accumulated small items pushes it down a list.
Whose follow-through is being measured
Adherence tooling almost always assumes the patient is the variable: the plan was sound, the clinic executed, and the person did not turn up. That is sometimes what happened. Very often what happened is that the order was never sent, the referral was never actioned, the recall was never run, or the result came back to a fax tray and no one told the patient there was anything to attend.
So each open expectation records where it is stuck, and the two are reported separately. A practice can see how much of its own gap rate belongs to it. This is less flattering than an adherence percentage, which is the argument for it: a number that can only indict the patient is not a measurement, it is an explanation prepared in advance.
Which is the position the rest of this site takes
The patient is a party with a legal right to their record, not a compliance risk to be managed. Tracking exists so that the people who undertook to do something can be held to it, and so that a person is told what is outstanding rather than asked why they failed. Nothing here scores a patient, and nothing here is a reason to withhold anything from one.
Gaps in a lien or personal injury book
In a personal injury matter a gap in treatment is used, routinely and effectively, to argue that the injury resolved or was never serious. That is a fact about how these cases are litigated and there is no benefit in being coy about it: an unexplained eight-week absence from a therapy course will be put to the treating clinician, and a chart that cannot say why will not help.
What the record can do is carry the reason. Gaps have causes — transport, work, childcare, cost, a flare, a second injury, or genuine improvement — and a reason recorded at the time by the person who took the call is worth considerably more than one reconstructed at deposition. Where a gap was the practice's, the record says that too.
Treatment is not scheduled to serve a case
The clinical plan is the clinician's, made on clinical grounds, and nothing in this capability proposes visits, extends a course, or flags a gap because a matter is open. The system records what was planned, what happened, and why. Building care around litigation value is both a clinical failure and the precise allegation opposing counsel is most eager to make, and a system that helped do it would be evidence for them.
What this is not
It is not clinical decision support and it does not propose plans. It is not a messaging product: patient-facing contact runs through the consented channels a practice already operates, scoped to that practice's own patients, and a person who asks to stop receiving it stops receiving it without that becoming a flag on their record. It is not a surveillance surface for a payer, a lienholder, or an employer — adherence detail is not in any of their release scopes, and the release model is where that is enforced rather than promised.
Which stakeholder is currently asking you for data you cannot easily give them?
That is the useful first conversation, and it is a short one. Bring the laboratories, the practices, and the thing that breaks today.
service@meta.clinic Read the release model first