Multi-agency integration breaks down less from lack of goodwill and more from unresolved conflict: eligibility rules that don’t align, unclear risk ownership, and funding boundaries that incentivize delay. The systems that perform best treat disagreement as a predictable operating condition and design decision-rights and dispute resolution to keep pathways moving. This guide builds on system integration and partnership delivery models and anchors to commissioner expectations for accountability in integrated care, focusing on how to prevent stalled cases becoming safety incidents.
The Commissioning, Funding & System Design Knowledge Hub supports clearer thinking about rates, contracts, provider markets, and system sustainability.
Why “who decides?” is the hidden backbone of integration
In community-based care, integrated pathways often span different legal authorities and payer rules. One agency may control service authorization, another may carry clinical responsibility, a third may manage housing supports, and a fourth may be the crisis responder. When decision-rights are unclear, the system slows down in the exact cases where speed matters: people with escalating behavioral health needs, unstable housing, caregiver breakdown, or repeated ED presentations.
Stalls usually show up as “admin delay” but they create real-world harm: missed follow-up after crisis, unsafe gaps in medication oversight, loss of engagement, or preventable placement breakdown. Because multiple agencies touched the case, no one can clearly explain why action didn’t happen. That is exactly what commissioners and oversight bodies are trying to eliminate.
What commissioners and oversight bodies expect to see
Expectation 1: A defined escalation ladder with time limits and named authority
Commissioners increasingly require escalation that is time-bound, not discretionary. When agencies disagree, the system must show: who is contacted first, when it escalates to a supervisor, when it triggers a cross-agency decision meeting, and who has final authority if consensus cannot be reached. Oversight teams look for evidence that escalation is used early enough to prevent risk, not after the person has already deteriorated.
Expectation 2: A “no wrong door” stance operationalized as provisional support
Where eligibility disputes take time, systems are expected to prevent gaps by using provisional supports (time-limited, lower-intensity, safety-stabilizing actions) while payer/program fit is resolved. Oversight expects clear rules for what can be provided provisionally, how it is documented, and how it transitions to the correct funded pathway. The goal is not to blur funding lines—it is to stop delay-driven harm.
Core components of decision-rights that actually work
Decision-rights map. A practical map specifies which decisions sit where: eligibility determinations, clinical risk decisions, crisis response triggers, service reductions, restrictive intervention thresholds (where relevant), and discharge/transfer approvals. It also states when a decision must be reviewed (for example, when risk increases, when a person refuses services, or after a critical incident).
Evidence standard for decisions. Dispute resolution fails when decisions are based on informal opinion. Strong systems require a minimum evidence set: recent contact history, risk screen outcomes, current service plan summary, and documented attempts to engage. This prevents both under-serving (“not eligible”) and over-serving (“keep them because we’re worried”) without defensible rationale.
Separate “who pays” from “what must happen now.” Operationally, these must be different conversations. If funding dispute blocks immediate action, the pathway will fail under real-world conditions.
Operational Example 1: A time-bound escalation ladder for eligibility disputes
What happens in day-to-day delivery. When frontline teams cannot agree eligibility/program fit within a set window (often 24–72 hours depending on risk), the case is escalated using a defined ladder. Step 1: supervisor-to-supervisor call with a structured dispute form (what program is proposed, why, what rule conflicts, what evidence supports the position). Step 2: cross-agency “rapid resolution” slot (a short, scheduled meeting with decision-makers, not a general partnership meeting). Step 3: final determination by a named authority (for example, a county program manager, MCO clinical lead, or designated integration lead) with a documented rationale and review date. The decision and rationale are logged in the shared pathway record, and the receiving team confirms acceptance with a start plan.
Why the practice exists (failure mode it addresses). Eligibility disputes often become “infinite loops”: referrals bounce, teams ask for more documentation, and responsibility is quietly avoided. A time-bound ladder prevents the failure mode where the system protects itself from financial risk by delaying action, leaving the person exposed to safety risk.
What goes wrong if it is absent. Without a ladder, disputes become informal and personality-driven. Cases stall until someone “has time,” or they escalate only after crisis. The system generates poor records: no clear decision point, no accountable owner, and no rationale that can withstand commissioner review. Providers also absorb hidden cost from repeated coordination attempts and staff burnout from unresolved conflict.
What observable outcome it produces. With a ladder, systems can evidence shorter time from disputed referral to accepted pathway, fewer repeat referrals for the same person, fewer “returned referrals” due to unresolved eligibility questions, and fewer crisis escalations attributed to service delays. Evidence includes dispute logs, time-to-resolution reports, and audit samples demonstrating rationale quality.
Operational Example 2: Provisional support packages to prevent delay-driven harm
What happens in day-to-day delivery. When a person’s need is urgent but program fit is unclear, the system triggers a provisional support package for a defined period (for example, 7–14 days). The package is not a full service authorization; it is a safety-and-stability bridge. It may include: rapid welfare check, short-term care coordination contacts, medication reconciliation support, crisis/safety planning refresh, and scheduling of priority assessments. A named coordinator tracks the package, documents every contact attempt, and sets a hard deadline for final pathway decision. At the deadline, the system either transitions the person into the correct funded program or escalates again with evidence.
Why the practice exists (failure mode it addresses). Integrated systems often fail by treating eligibility as a prerequisite to any action. Provisional support exists to prevent the failure mode where “paperwork time” becomes “risk time,” particularly after hospitalization, crisis events, or sudden caregiver breakdown.
What goes wrong if it is absent. Without provisional support, urgent cases wait for decisions and deteriorate. The person may end up in ED, inpatient, or in crisis services that are far more costly and disruptive than short-term stabilization. The system then faces a reputational and contractual problem: it cannot show it acted proportionately to known risk.
What observable outcome it produces. Provisional packages produce measurable stability indicators: fewer crisis re-contacts in the first days after referral, improved engagement rates (because someone actually reaches the person), and clearer transitions into longer-term services. Evidence includes provisional package triggers, contact logs, safety plan updates, and conversion rates from provisional to funded pathway.
Operational Example 3: Decision-rights for risk ownership and “who holds the plan”
What happens in day-to-day delivery. For individuals supported by multiple agencies, the system assigns a “plan-holding” role for a defined period. The plan-holder is responsible for maintaining a single current risk picture: recent incidents, crisis contacts, safeguarding concerns, and protective factors. They convene brief case synchronization when triggers occur (missed visits, medication non-adherence risks, housing disruption, escalating behaviors). The plan-holder does not replace other agencies’ responsibilities; instead, they ensure the system has one current version of “what is happening” and one accountable owner for cross-agency coordination until stability improves.
Why the practice exists (failure mode it addresses). In multi-agency care, risk can become “everyone’s concern and no one’s responsibility.” Plan-holding exists to prevent the failure mode where agencies assume another team is monitoring deterioration, leading to missed escalation and avoidable harm.
What goes wrong if it is absent. Without plan-holding, agencies maintain separate notes and different assumptions about risk. One team may reduce contact because they believe another team increased it; another may assume medication oversight is happening elsewhere. Failures surface as serious incidents, safeguarding escalations, and post-event reviews that show fragmented information and unclear decision-making.
What observable outcome it produces. Plan-holding supports improved escalation timeliness, fewer contradictory plans, and clearer accountability in audits. Evidence includes documented plan-holder assignment, trigger-based synchronization notes, updated risk summaries, and incident review findings showing improved information continuity.
Governance and assurance mechanisms that make dispute resolution defensible
Standardize dispute documentation. A short, consistent dispute form (problem statement, evidence summary, proposed pathway, rule conflict, interim risk actions, requested decision) reduces ambiguity and prevents disputes being driven by informal persuasion rather than evidence.
Track “stalled days” as a quality metric. Many systems track time-to-service but not time lost to inter-agency disagreement. Commissioners value transparency: how many cases stalled, why, and what was changed to reduce stalls.
Use post-resolution learning loops. If the same dispute type repeats (for example, behavioral health eligibility thresholds, housing supports boundaries), the partnership should update rules, templates, or escalation triggers. Oversight bodies expect continuous improvement, not recurring friction accepted as normal.
Common pitfalls
- Consensus-by-meeting: monthly partnership meetings do not resolve urgent disputes. Build rapid-resolution slots with decision-makers.
- Risk dumped into “not eligible” language: separate immediate safety actions from longer-term program placement decisions.
- Unclear final authority: without a named final decider, disputes become delays.
What “good” looks like in practice
High-performing integrated systems can explain, plainly, what happens when agencies disagree: who decides, how fast, what evidence is required, what interim supports prevent harm, and how the decision is recorded. They can show that delay is treated as a risk signal, not an administrative inconvenience. That is the operational difference between an integration model that looks good on paper and one that withstands real-world pressure, oversight review, and commissioner scrutiny.