By Clara Team
Systemic therapy notes: relationships, hypotheses, and change
Several people can leave the same systemic therapy session with different accounts of what happened. The record does not need to choose a winning version. It does need to show who said what, how responses connected, and which context mattered.
The challenge is to describe relationships without turning a temporary position into identity. Labels such as “controlling parent” or “defiant teenager” close down a sequence that may change with the subject and moment.
Setting, participants, agreements, and systemic model all shape what belongs in the record. The structure below is one review option, not a universal template.
What makes a clinical note systemic
A note does not become systemic simply because it names an alliance or includes a genogram. Its clinical lens becomes visible when the record preserves context, relationships, and interaction rather than locating the symptom solely inside one person.
This does not mean minimising individual distress or distributing responsibility automatically. It means asking how a difficulty is organised in particular circumstances, what responses it invites, and what alternatives appear when one part of the sequence changes.
Our overview of documentation across therapeutic orientations places this lens alongside other approaches. In systemic therapy notes, multiple perspectives should remain distinct and hypotheses should remain provisional.
Context, participants, and agreements
The opening should note who attended, who was expected but absent, whether an individual meeting took place within couple or family work, and which agreements shaped the conversation. Context such as a house move, changed caring responsibilities, or contact with extended family may also alter the meaning of an interaction.
Attribute perspectives instead of collapsing them
When several people participate, the note should not create a single account as though everyone described the same experience. Attribution preserves differences:
- “E says conversations stop when they ask for reassurance.”
- “F describes pausing as an attempt to prevent the disagreement from escalating.”
- “E adds that they return to the subject hours later; F says they wait until the tension has fallen.”
This wording records convergence and disagreement without deciding which account is the truth. It also prevents the person who speaks most, or holds more authority, from becoming the voice of the whole system.
Record the interaction sequence
An interaction sequence describes what tends to happen before a response, how another person reacts, and what follows. Instead of “F withdraws and causes E's anxiety”, a record might say:
“When E asks for an immediate answer, F becomes quiet. E understands the silence as lack of interest and asks more insistently. F leaves; E reports increased anxiety and does not return to the subject. F says they wait for the tension to fall before reconnecting.”
Describing a circular sequence does not make every action equivalent. Where there is risk, violence, coercion, or a marked power difference, it should be documented directly and addressed according to the relevant clinical and legal duties. Relational language should never obscure harm.
Positions, resources, and exceptions
Relational positions can change with the subject and context. Someone who pursues contact in one conversation may seek distance in another. Labels such as “the mediator” or “the responsible one” should not be treated as permanent roles.
The record can also preserve resources: an attempt at repair, a boundary that was respected, humour, support from a wider network, cooperation under particular conditions, or an exception to the usual sequence. These are not reassuring additions. They help identify where movement may be possible.
From a closed claim to an open hypothesis
A systemic hypothesis guides enquiry and intervention. It does not prove what a family is like or establish a final cause. Tentative language and alternative explanations keep it clinically useful.
The examples below are synthetic and do not describe an actual family.
Closed claim:
“The father avoids conflict and causes his daughter's physical symptoms to keep the family together.”
Open hypothesis:
“We are exploring whether, during periods of tension, attention to the daughter's physical distress temporarily shifts the adults away from disagreement. The daughter reports that symptoms also appear before exams and when there has been no visible conflict. Her father says he usually pauses the conversation to attend to her. Her mother observes that this coordinates the adults but does not believe it explains the origin of the distress. School stress, disrupted sleep, and health factors requiring follow-up remain alternative explanations.”
The second version separates observation, interpretation, and participants' responses. It also leaves room for the hypothesis to be changed or set aside. Our guide to tracking patterns across therapy sessions offers a way to review recurrence without treating it as proof.
Intervention, responses, and movement
“Asked a circular question” says little about the effect of the intervention. The record is stronger when it names the difference being explored and how each person responded.
For example: “E was asked what F does when E raises their voice; F was asked what E does after F looks away. E said F becomes silent; F noticed that silence leads E to repeat the question more intensely. E recognised that pressing increases F's withdrawal; F identified that withdrawing intensifies E's pursuit of an answer.” The circular question makes the feedback loop visible without deciding who started it.
Movement can also take the form of a less predictable response, inclusion of an absent perspective, recognition of an exception, a clearer boundary, or a different description of the problem. If no movement is visible, the responses still help the therapist reconsider the intervention, timing, or hypothesis.
The formal record and supervision
The formal record should retain observations, hypotheses that informed decisions, and clinically relevant responses. Raw therapist reactions, highly speculative relational ideas, or material needing further reflection may be better explored in supervision. Keeping it outside the clinical record does not remove data-protection responsibilities. If it contains personal data, it may still be subject, depending on context, to confidentiality, security, retention, and applicable access obligations. It should be handled with restraint and within the relevant professional and legal framework.
Confidentiality when more than one person is involved
In couple, family, and wider-system work, confidentiality and access rules may vary by each person's role, setting, jurisdiction, and the agreements made with participants.
One participant's authorisation does not automatically permit disclosure of another person's information. In both formal records and supervision material, third-party details should be limited to what is clinically necessary and handled consistently with applicable access rights, boundaries, and obligations.
A review structure for systemic therapy notes
Before closing the record, review six areas:
- Context: Is it clear who participated, what the frame was, and which circumstance mattered?
- Perspectives: Is each account attributed without merging different voices?
- Sequence: Can the interaction be followed without assigning one person as the sole cause?
- Positions and resources: Are positions described as variable, with relevant capacities, support, or exceptions included?
- Hypothesis and intervention: Does the hypothesis remain open, include alternatives, and make the clinical decision understandable?
- Movement: Does the record show each person's response, what shifted, and what should be observed next?
Not every area needs equal space in each session. The structure checks whether enough relational reasoning remains to resume the work. Over time, clinical documentation designed for continuity can show how sequences, perspectives, and hypotheses develop without fixing them too early.
With several voices in the room, Clara preserves attribution, the feedback sequence, and the question that remains open for the next session.