The co-supervision arrangement at the heart of UK doctoral clinical psychology training is one of those structures that sounds straightforward on paper. The clinical supervisor handles the clinical work in the placement service. The course/academic supervisor handles the academic and developmental side from the university. Together they hold the trainee's training year. The arrangement is required by the accrediting bodies — the BPS and the HCPC for the DClinPsy itself, the BABCP where a trainee is also pursuing CBT specialist accreditation.
In practice, the arrangement is asymmetric in ways the formal description does not capture, and most of the operational failure modes live in that asymmetry.
What each role is supposed to do
The clinical supervisor's remit is the placement. They carry case-by-case clinical responsibility for the work the trainee does in the service, sign off competence at the placement-specific level, ensure safeguarding pathways are clear, and hold the trainee within the service's standards of practice. They are, for the period of the placement, the senior clinician responsible for what happens in the room when the trainee is the therapist.
The course supervisor's remit is broader and longer-arc. They hold the developmental conversation across placements — what this placement is contributing to the trainee's overall formation, how the trainee is integrating theory with practice, what gaps the next placement might address. They advocate for the trainee when placement-level issues arise. They assemble the evidence for portfolio assessment. They are, for the three years of training, the university's representative in the trainee's development.
Described at this level the division of labour looks clean. The placement supervisor handles the immediate clinical work. The course supervisor handles the developmental arc. The trainee is held by both.
Where the asymmetry shows
The asymmetry is in cadence and proximity.
The clinical supervisor sees the trainee weekly, often for an hour or more, in a relationship that is immediate, embedded in the day-to-day life of the service, and accountable in real time. If something goes wrong in a session on Tuesday, the clinical supervisor will hear about it by Thursday at the latest, often within hours. The supervisory contact is dense, frequent, and clinical-content-rich.
The course supervisor sees the trainee periodically — monthly is typical for a full year-long arrangement, term-based for some courses, less frequent in stretches when the trainee is on a placement that the course is content with. The contact is structured around developmental review rather than case-by-case clinical work. The course supervisor's accountability is indirect: they do not carry clinical responsibility for the trainee's caseload, and their information about how the placement is actually going comes mostly through what the trainee reports.
When the trainee's experience on placement diverges between the two supervisors — when something happens that the trainee doesn't tell the course supervisor for weeks — the structure was set up to catch it, but the cadence usually doesn't. By the time the course supervisor learns that the placement has been going badly for six weeks, several of the corrective moves that could have helped are no longer available. The placement has either resolved itself, drifted further, or hardened into something that needs more than the course supervisor's quiet advocacy to address.
This is not a niche scenario. It is the predictable consequence of pairing a high-frequency supervisor with a low-frequency one and routing information between them through the person whose stake in the information is most complicated — the trainee.
The contract problem
Co-supervision typically does not have a contract that operates as a working document. There may be a generic placement agreement, signed at the start of the placement, specifying the broad terms of the trainee's presence in the service. There is rarely a triadic agreement — between trainee, clinical supervisor, and course supervisor — specifying what gets discussed where, what gets escalated and to whom, and what each supervisor can expect from the others.
This is the same structural failure that affects supervision contracts more generally (the issue takes its own treatment in the supervision contracts article), with sharper edges in the triadic case. Each supervisor has their own understanding of what the arrangement involves. The understandings overlap substantially but not completely. The places where they do not overlap are where things tend to go wrong.
Specific clauses that a triadic agreement would benefit from including, and that the standard placement agreement rarely does:
What the trainee brings to each supervisor. Clinical material to the clinical supervisor, developmental reflection to the course supervisor — but also the explicit understanding that significant clinical concerns are reported to the course supervisor in summary form, and significant placement-level concerns are reported to the clinical supervisor before they reach the course supervisor.
What gets escalated and to whom. A safeguarding concern about a client is escalated through the clinical supervisor's service channels. A safeguarding concern about the trainee themselves — their welfare, their fitness to practise during a particularly difficult period — needs an escalation pathway that the trainee has read in advance. The course supervisor is usually the safer first contact for that latter category, which the trainee may not realise without the contract spelling it out.
What the supervisors share with each other. The defaults vary across courses and services. Some assume that the two supervisors will be in regular contact; others assume that all communication flows through the trainee unless there is a specific reason for direct contact. Both defaults are workable; the failure mode is when the two supervisors have different defaults without realising it.
The cadence of direct supervisor-to-supervisor contact. Even if only once per placement, an explicit point at which the clinical and course supervisors speak — without the trainee mediating — adds a check that the structure does not otherwise build in.
The communication problem
Clinical and course supervisors rarely speak directly. The trainee mediates. This is not always inappropriate — the trainee is, after all, the subject of the supervision, and direct supervisor-to-supervisor contact at every turn would infantilise the trainee and reduce the developmental value of the placement. The trouble is the cases where direct contact would help and the trainee is the worst-placed person to broker it.
Two categories recur. The first is safeguarding concerns about the trainee themselves — a trainee whose mental health is deteriorating, whose personal circumstances are eroding their capacity in placement, whose performance is slipping in ways that worry the clinical supervisor. They may not relay that accurately to the course supervisor. They may not relay it at all. The clinical supervisor's choice becomes whether to go around the trainee directly, which feels like a breach of the supervisory relationship, or to leave the course supervisor under-informed, which is its own kind of failure.
The second is service-level issues that affect placement quality — short-staffing, slipping supervision standards, a context where the trainee is being asked to do work above their level because there is no one else. The trainee, who depends on the placement and on their working relationship with the clinical supervisor, is in the worst position to raise it. The standard structure has no clean channel for the clinical supervisor to raise it either, particularly when the clinical supervisor is themselves part of the service that is creating the problem. In both categories, the issue is not bad faith. It is that the cadence and proximity asymmetry, plus the absence of a direct supervisor-to-supervisor channel, leaves the most important information stranded.
The drift implication
The therapist drift literature documents how qualified therapists move away from evidence-based practice when fidelity-supporting structures are not in place. Trainees, by virtue of being in active training, are in some respects better protected — they are observed more, scrutinised more, supervised more. The protection depends on the supervisory structure functioning as designed. A co-supervision arrangement that is functioning only at the clinical-supervisor level, with the course supervisor receiving filtered and delayed information, gives the trainee a single supervisory perspective rather than two. The structural argument for triadic co-supervision is precisely that two supervisors with different angles catch what one alone would not. When the arrangement collapses into a single working supervisor and a periodic check-in, the protective effect of two perspectives collapses with it.
What good co-supervision looks like operationally
An agreed cadence for clinical/course supervisor direct contact, specified at the start of each placement and held to. Once a placement is the minimum; mid-placement and end-of-placement is standard; more frequently when the placement is complex or the trainee is in their first year.
A triadic agreement at the start of each placement, specifying scope, escalation pathways, communication norms, and the cadence of direct supervisor-to-supervisor contact. Signed by all three parties. Reviewed at mid-placement.
Explicit escalation pathways. Safeguarding concerns about clients go through the clinical supervisor and the service's safeguarding lead; safeguarding concerns about the trainee themselves go through the course supervisor in the first instance; service-level concerns about placement quality can be raised through either route but will be communicated between supervisors directly once raised. The pathways are documented, not improvised at the moment of crisis.
A shared view of the trainee's developmental priorities for the year. Both supervisors know what the trainee is working on, what last placement's review identified as gaps, and what this placement is expected to contribute. The information lives somewhere both supervisors can see.
The course supervisor as advocate when service-level issues affect placement quality. This is the use case the formal structure is least equipped to support and where direct supervisor-to-supervisor contact pays off most — the course supervisor's institutional position outside the placement service gives them leverage the clinical supervisor often lacks.
None of these are enormous asks. They are what any reasonable co-supervision arrangement would have built into it from the start, given a structure that prompted for it rather than relying on the supervisors to assemble it themselves.
Supervisia's supervision pathway supports triadic co-supervision arrangements.
The supervision pathway carries the triadic agreement — DClinPsy trainee, clinical supervisor, course supervisor — with explicit escalation pathways, agreed cadence for direct supervisor-to-supervisor contact, and a shared developmental record visible to all three parties. The asymmetry between the high-frequency clinical relationship and the longer-arc course relationship stops being a structural blind spot, because the structure itself surfaces what each supervisor needs to see.
References
- BPS. Standards for the accreditation of Doctoral programmes in clinical psychology. British Psychological Society, current edition.
- HCPC. Standards of education and training. Health and Care Professions Council.
- BABCP. Criteria for Accreditation as a CBT Therapist (trainee and post-qualification pathways). British Association for Behavioural and Cognitive Psychotherapies.
- Waller, G. & Turner, H. (2016). Therapist drift redux: Why well-meaning clinicians fail to deliver evidence-based therapy, and how to get back on track. Behaviour Research and Therapy, 77, 129–137. DOI: 10.1016/j.brat.2016.01.007. PubMed: 26752326.
Last updated: May 2026