How Mental Health Providers Should Work With NHS Trusts in Integrated Systems
Working with NHS Trusts in integrated mental health systems is most challenging at the operational edges: referrals that do not meet thresholds, delayed step-up decisions, incomplete information at handover, and uncertainty about who is leading when risk rises. Commissioners are less interested in partnership statements than in whether these interfaces work safely and consistently. This article sits within working with ICBs in mental health and links to mental health service models and pathways, setting out the practical operating model providers need when delivering alongside Trust teams across integrated pathways.
Start with clarity: boundaries, thresholds and accountability
Before collaboration can work, the system needs shared clarity on three things:
- Boundaries: what each service is responsible for (including what is explicitly out of scope).
- Thresholds: how decisions are made on acceptance, step-up, crisis escalation and safeguarding.
- Accountability: who is leading at any point in time, and how leadership changes during escalation.
Where these are unclear, providers fall into two unhelpful patterns: either they take on work that is out of scope (creating hidden risk and performance drift), or they escalate repeatedly without clear evidence of work attempted (which looks like risk dumping). Commissioners and Trust partners will challenge both.
Operational practices that make Trust working credible
1) Standardise the “handover minimum dataset”
Trust working becomes safer when information quality is predictable. A minimum dataset for referrals, step-up requests and handovers should include: current presentation, key risks, what has changed, actions already taken, current plan, and what is being requested with time sensitivity. A standard format reduces delay and dispute because it supports rapid decision-making.
2) Use agreed escalation ladders and record them
Escalation should not depend on personal relationships. Providers should use an agreed ladder (front-line → manager → senior) and record each step: when escalation happened, what information was shared, what response was received, and what was done next. This creates an audit trail and supports learning when interface delays occur.
3) Build “interface learning” into governance
Interface incidents (late crisis step-up, referral bounce, unclear accountability) should be reviewed as learning themes. Good governance does not treat these as partner blame; it treats them as system design problems that can be improved through clearer thresholds, better information standards, and stronger escalation routes.
4) Maintain consistency across localities and teams
In integrated systems, Trust interfaces vary across localities. Commissioners expect providers to recognise and manage this variation: documenting differences, explaining impacts, and agreeing improvement actions rather than allowing each team to reinvent its own approach.
Operational examples (day-to-day Trust collaboration)
Example 1: Stabilising referrals and acceptance decisions
Context: A provider receives high volumes of referrals from a Trust service, with frequent missing information and mismatch to scope. People experience delays and staff spend time clarifying basics.
Support approach: The provider works with the Trust to introduce a referral checklist and clear threshold guidance, plus a weekly 30-minute triage alignment call for edge cases. Declined referrals are logged with structured reasons to evidence patterns.
Day-to-day delivery detail: Triage staff use the checklist consistently and respond with clear acceptance/decline rationale. The weekly call reviews the top decline themes and updates guidance. The provider trains staff to use consistent language so Trust colleagues receive predictable messages. Governance tracks decline reasons and time-to-first-contact for accepted referrals.
How effectiveness/change is evidenced: Reduced incomplete referrals, fewer avoidable declines, and improved time-to-start. Evidence includes referral logs, call notes and trend reporting.
Example 2: Crisis escalation interface that improves timeliness
Context: A cohort experiences repeated crises. Staff report uncertainty about when to escalate and frustration about inconsistent crisis access routes.
Support approach: The provider introduces early warning indicator recording for the cohort, a standard escalation summary, and manager review for repeat escalations. Interface delays are documented and escalated through agreed routes rather than informal chasing.
Day-to-day delivery detail: At each contact, staff record early warning indicators and actions taken. When escalation is required, the standard summary is sent so Trust crisis teams receive decision-ready information quickly. Managers review repeat escalations weekly, confirm plans are updated with learning, and log interface delays with outcomes. Where access issues persist, the provider escalates formally and tracks agreed improvements.
How effectiveness/change is evidenced: Improved time-to-action from early warning indicators, fewer late-stage crises, and clearer documentation of escalation decisions. Evidence includes escalation logs, file sampling and governance action tracking.
Example 3: Shared safeguarding work without default restriction
Context: Exploitation concerns involve multiple partners. Staff become risk-averse and apply informal restrictions that are not always time-limited or reviewed, while safeguarding actions drift due to unclear ownership.
Support approach: The provider introduces a safeguarding action tracker (owner, deadline, escalation route) and a restrictions standard (rationale, least restrictive alternatives, time limit, review date, step-down plan). Weekly safeguarding huddles track progress.
Day-to-day delivery detail: Team leads ensure safeguarding actions are completed and that partner contact is recorded. Supervision tests staff decision-making on thresholds and proportionality. Where restrictions are in place, review decisions are recorded explicitly and step-down is expected when risks change. Governance reviews long-running restrictions and ensures safeguarding actions remain active and outcome-focused.
How effectiveness/change is evidenced: Higher safeguarding action completion, stronger multi-agency evidence in files, and reduced long-running restrictions. Evidence includes huddle notes, audit samples and review decisions.
Explicit expectations that must be met
Commissioner expectation
Commissioners expect Trust working to reduce friction and improve reliability at interfaces. They will look for clear thresholds, documented escalation routes, evidence of shared risk management, and trend improvement in recurrent interface problems. They also expect performance data to reconcile to operational records and case files.
Regulator / Inspector expectation (e.g. CQC)
CQC expects safe coordination, continuity and learning across organisational boundaries. Inspectors will test staff understanding of escalation and safeguarding routes, check that risk and restrictive practice decisions are proportionate and reviewed, and triangulate governance claims against file evidence and staff practice.
What to show in a commissioner or Trust review meeting
Providers perform best in reviews when they bring practical evidence: a short set of anonymised “interface traces” (referral acceptance, escalation, safeguarding coordination) with clear timelines and decision records; a small number of trends (referral bounce reasons, escalation timeliness, safeguarding action completion); and a documented improvement cycle (issue → action → verification). This keeps the discussion grounded in reality and shows mature, controllable partnership working.