Managing Risk, Safeguarding and Accountability When Working With ICBs in Mental Health
When mental health services are delivered across system boundaries, safeguarding and risk are where accountability is tested. ICBs want confidence that providers can recognise deterioration early, escalate promptly, coordinate safeguarding effectively, and evidence proportional decision-making — including least restrictive practice where restrictions are used. The biggest operational risk is “shared responsibility” becoming “no responsibility”, with drift at interfaces and unclear escalation. This article sits within working with ICBs in mental health and connects to mental health service models and pathways, focusing on practical controls, governance rhythms and evidence trails that commissioners and CQC recognise as robust.
What “shared risk” means in practice
Shared risk is not a strategic concept; it shows up in day-to-day operational situations:
- Uncertainty about who leads when a person’s presentation escalates.
- Different threshold interpretations between providers and Trust teams.
- Safeguarding actions split across agencies with unclear ownership and timelines.
- Informal restrictions introduced in response to risk, then left in place without review.
- Information gaps at handover that lead to unsafe assumptions.
Providers cannot control every partner response, but they can control their own recognition, escalation, recording, follow-through and governance. That is what commissioners and inspectors test.
Core operational controls for risk and safeguarding in system working
1) Early warning recognition and documented escalation thresholds
Providers should define early warning indicators for high-risk cohorts and ensure these are recorded routinely. Escalation should be triggered by thresholds that staff can apply confidently, with a standard way of recording what changed and what actions were taken. In practice, this is best embedded into contact notes and review templates rather than added as a separate process.
2) Escalation ladders that do not rely on individuals
Escalation must be repeatable: front-line escalation, manager escalation, senior escalation. Each step should be recorded with timestamps and outcomes. This protects both people using services and staff, because it evidences that risk was recognised and acted upon within expected timescales.
3) Safeguarding action ownership and “what happens next” clarity
Effective safeguarding in system working depends on action ownership: who is doing what, by when, and what escalation route applies if actions are not completed. A safeguarding action tracker (even a simple one) often provides more control than lengthy narrative, because it makes drift visible.
4) Least restrictive practice and step-down expectations
Where restrictions are used (formal or informal), providers should treat them as time-limited measures with explicit review points and step-down planning. Commissioners and CQC are increasingly alert to restrictions that become default practice without evidence of proportionality, alternatives considered, and review decisions.
5) Governance verification (not just governance discussion)
Governance must verify that controls are working. This typically means sampling high-risk cases, reviewing escalation logs, auditing safeguarding action completion, and re-checking whether restrictions are being reviewed and stepped down.
Operational examples (how risk and safeguarding controls work day to day)
Example 1: Escalation control for a repeat-crisis cohort
Context: A provider supports a cohort with frequent crisis escalation. The ICB is concerned about late escalation, inconsistent documentation and repeated urgent presentations that increase system pressure.
Support approach: The provider introduces a repeat-crisis control plan: early warning indicators recorded at every contact, a standard escalation summary, and mandatory manager review for any second escalation within a defined period. Escalation timelines are recorded so learning is possible.
Day-to-day delivery detail: Staff record early warning indicators and actions taken (not just discussion). When thresholds are met, they complete the standard escalation summary through the agreed secure route. Managers review repeat escalations weekly, check whether plans were updated after escalation, and ensure follow-up actions are completed (for example, safeguarding checks, medication review request via appropriate routes where relevant, increased contact frequency). Governance reviews cohort trends monthly and identifies patterns: time-of-day access issues, information gaps, or threshold confusion.
How effectiveness/change is evidenced: Reduced late escalations, improved timeliness of step-up actions, clearer decision trails, and better post-crisis plan updates. Evidence includes escalation logs, sampled records, and governance minutes showing actions and verification.
Example 2: Safeguarding coordination for exploitation and self-neglect
Context: Exploitation concerns involve multiple partners. Some actions drift because ownership is unclear and response times vary. The ICB wants assurance that safeguarding is timely and that providers escalate when partner response is slow.
Support approach: The provider implements a safeguarding action tracker with owner/deadline/escalation fields and runs weekly safeguarding huddles for active cases. A short staff guide clarifies thresholds and “what to do today” actions, aligned to multi-agency safeguarding expectations.
Day-to-day delivery detail: Each safeguarding concern is recorded with immediate actions (safety planning, information gathering, partner notifications) and a clear next-step plan. In the weekly huddle, team leads review each action and confirm completion, documenting partner contacts and outcomes. If partner response is delayed, the escalation ladder is used and recorded. Supervision uses real cases to test decision-making and proportionality, ensuring staff confidence is consistent across teams.
How effectiveness/change is evidenced: Higher safeguarding action completion, improved timeliness, clearer multi-agency evidence in files, and fewer repeat concerns without documented learning. Evidence includes tracker outputs, huddle records and audit sampling.
Example 3: Restrictive practice controlled through review and step-down
Context: Following incidents, teams introduce informal restrictions (monitoring, limiting access, limiting community activity). Restrictions persist without review and become embedded, creating rights risks and potential CQC concern.
Support approach: The provider introduces a restrictions standard: rationale, least restrictive alternatives considered, time limit, review date, and step-down plan. Restrictions are tracked on a register reviewed monthly, with quarterly senior sampling.
Day-to-day delivery detail: Supervisors require explicit review decisions: continue, modify, or step down, with rationale. If a restriction continues, staff must document why less restrictive options are not sufficient and what is being done to reduce risk. During handovers across partners, the minimum dataset includes restrictions and review dates so accountability is not lost. Governance reviews long-running restrictions and ensures that safeguarding actions and positive risk-taking plans remain active.
How effectiveness/change is evidenced: Reduced duration of restrictions, improved proportionality documentation, and visible step-down decisions over time. Evidence includes the register, sampled files and governance action tracking with re-checks.
Explicit expectations that must be met
Commissioner expectation
ICBs expect risk and safeguarding controls that are auditable and consistent across teams. They will look for documented escalation thresholds, evidence of timely safeguarding action, ownership and completion of safeguarding tasks, and clear management of interface risks. They also expect verification: sampling, audit and trend review demonstrating that controls are working and improving.
Regulator / Inspector expectation (e.g. CQC)
CQC expects safe care, effective safeguarding and protection of rights through least restrictive practice. Inspectors will test whether staff recognise deterioration and escalate appropriately, whether safeguarding responses are timely and recorded, and whether restrictions are proportionate, reviewed and stepped down. They will triangulate leadership oversight, staff understanding and record evidence.
What to show in assurance conversations
The most persuasive assurance evidence is a small set of “risk traces”: anonymised examples that show early warning recognition, escalation steps, safeguarding actions with completion evidence, review decisions, and learning embedded. Combined with trend measures (repeat escalations, safeguarding action completion, restriction duration), these traces demonstrate control rather than intentions.