For clinicians
A Formulation Short Enough to Actually Use
There are two kinds of couples formulation in practice, and neither of them works.
The first is three lines in a chart: high conflict, communication difficulties, working on de-escalation. It satisfies the record and guides nothing. The second is the twelve-page conceptualisation written for a training placement, which is genuinely thorough and which nobody — including its author, six weeks later —
ever reads again.
What is actually needed sits between them: one page, written so that a colleague covering your caseload could read it in ninety seconds and know what to do on Thursday.
What has to be on it
9. Screening status. Both partners screened individually, date, finding, disposition, re-screen due. This goes first, not last. If it is not on the page, nothing else on the page is safe to act on.
10. The formulation in three sentences. Neither partner as the villain, and both would recognise themselves.
11. Each partner’s protective strategy, and what it costs this relationship. One line each. Not a label —
a description of what they do and what it produces.
12. The cycle, in one sentence. Behaviour only, both directions.
13. The attachment injury, if you have located one. The event and the date. If you have not, write that you have looked.
14. Stage and sequence. Safety, grief, or skills — and the evidence for why you think you are there.
15. Contraindications and cautions. Substance use, untreated trauma, a live legal matter, anything that changes what you can do.
16. What would tell you this is working. Specific and observable.
Eight items. It fits on a page because each one is a sentence rather than a paragraph, and the discipline of a sentence is what makes it usable.
The three-sentence test
Item two is the one people find hardest, so here is the test that tells you it is finished.
You could say it out loud to both partners, neither of them is the villain in it, and both of them would recognise themselves.
If one partner would experience it as a verdict, it is not a formulation. It is a diagnosis of the person you have found more difficult, and you will act on it without meaning to.
A worked example. Not: she is anxiously attached and pursues; he is avoidant and withdraws. That is a pair of labels and one of them is going to feel accused.
“He learned very early that needing someone out loud produced nothing, so under pressure he goes quiet and manages it alone. She learned that people leave without warning, so under pressure she moves toward him fast to check. What each of them does to stay safe is the exact thing that frightens the other.”
Same content. Nobody is the problem. Both are describable to their faces, which is the point — a formulation you cannot say out loud in the room is one you are keeping from your clients, and you should ask yourself why.
Item eight, and why it is not optional
What would tell you this is working is the item most often skipped and the one that prevents the largest failure in this work: a course of therapy that feels productive for eleven months and changes nothing at home.
Feeling better is not an answer. Sessions going well is not an answer. Answers look like: they had the money conversation without either of them leaving the room. She asked him directly for something instead of testing. He came back within the hour after the Thursday argument rather than the next morning.
Write two. Date them. Look at them at session twelve. If neither has happened, the question is not whether to persevere — it is what you have been treating.
On unvalidated instruments in a chart
If you use structured tools that are not validated instruments — and most useful clinical sorting devices are not — the chart must say what they are.
A score from a sorting device is a structured clinical impression. It is not a test result, it has no norming sample, and writing it in a way that implies otherwise creates a problem for you and a false impression for anyone reading the file later. Write structured clinical impression and the tool’s name, and you have said something true and defensible.
When to rewrite it
Formulations should be revisited, not preserved. Three triggers:
• Session eight or thereabouts. Enough has happened that your first version was written with a fraction of what you now know.
• Any re-screen finding. New safety information invalidates everything downstream of it.
• Whenever you notice yourself surprised. Surprise is the reliable signal that the model in your head does not fit the couple in the room. It is the most useful diagnostic event available and clinicians routinely let it pass.
And write the reason for any departure from your plan at the time you depart. Reconstructed reasons are always tidier than real ones, and the tidiness is what makes them useless when you look back for the actual decision point.
The point of the constraint
One page is not a compromise for the sake of brevity. It is what forces the thinking.
A twelve-page conceptualisation can contain three contradictory hypotheses without their author noticing, because there is room for all of them. One page cannot. You have to decide what you actually think, which is the entire clinical value of the exercise — and it is why the page takes forty minutes rather than the ten its length suggests.
Writing it so a colleague can act on it
The test for the whole page is not whether it is accurate. It is whether a competent colleague covering your caseload, reading it cold on a Thursday morning with the couple arriving at ten, would know what to do.
That test rules out a surprising amount of what gets written in charts. Working on communication does not survive it. Attachment issues does not survive it. Sequencing into grief work; she can now stay in the room for about four minutes before she needs a break, and he does not yet initiate the return survives it easily, and takes the same number of seconds to write.
Specificity is not thoroughness. It is usually shorter than the vague version, because a vague sentence has to gesture at everything while a specific one only has to say one thing.
What to leave off
Three things clinicians include that make the page less usable.
• Extended developmental history. It belongs in the history sheets, not the summary. The summary needs the one line from that history that explains the loop.
• Your hypotheses about the marriage’s future. Not clinical information, and once written it will influence whoever reads it, including you.
• Anything from an individual session that the partner has not agreed can be in the shared record. This is a confidentiality question before it is a formatting one, and it needs settling at the time of the individual session rather than at the time of writing.
From the book: The Clinical Workbook builds this as two sheets: Who Is in the Room, which develops the formulation, and the one-page Formulation Summary written to be readable at speed. Both are reproducible and both cross-reference the manual chapters behind them.
Approx. 1162 words in the body above.