Supported decision-making (SDM) is most vulnerable precisely when it matters most: during escalation, incident response, and restrictive-practice decisions. In these moments, teams are under time pressure, risk feels immediate, and the system’s instinct is to control. Rights-safe providers design SDM “under pressure” workflows so autonomy is not erased when safety concerns rise. This article sits in Supported Decision-Making, Rights & Autonomy in Practice and connects to operational service pathways in Service Models & Support Pathways.
What SDM means during escalation (and what it does not mean)
SDM during crises does not mean allowing any choice without limits. It means: (1) using the person’s communication and regulation supports before imposing control, (2) offering choices that are real within safety boundaries, (3) documenting the person’s preference and how it was supported, and (4) using governance to ensure any restriction is proportional, time-limited, monitored, and reviewed toward reduction. The operational standard is “least restrictive while safe,” evidenced through a traceable decision process.
Providers should separate two concepts: immediate safety actions (time-limited, reactive, and documented) versus ongoing restrictions (planned, authorized, reviewed). Many services get into trouble when an emergency action becomes an ongoing restriction by drift—without review, authorization, or a reduction plan.
Build an SDM escalation pathway that staff can follow
Step 1: Regulation supports first
Staff should have a “regulation menu” for the person: sensory tools, quiet space, movement breaks, music, reduced demands, trusted staff. This should be pre-written and practiced, not invented during crisis.
Step 2: Offer bounded choices
Even during escalation, choices can exist: where to take space, which staff member supports, whether to use headphones or weighted blanket, whether to talk now or later, whether to walk outside with staff or sit inside. Bounded choices reduce power struggles and support autonomy.
Step 3: Document preference and supports used
Use a brief SDM-in-incident note: what the person signaled, what choices were offered, what supports were used, and what the person selected. This is different from an incident report narrative; it is evidence of autonomy-support attempts.
Step 4: Trigger a restrictive decision review when thresholds are met
If a restriction is being considered (limits on community access, locked storage, increased supervision), trigger a time-bound review chaired by a manager and, when applicable, a clinician/BCBA. The review must consider alternatives and specify time limits and monitoring metrics.
Operational Example 1: SDM during in-the-moment escalation in a supported living setting
What happens in day-to-day delivery: A person begins escalating after a schedule change. Staff follow the person’s pre-agreed “regulation and choice” card kept in the home: first, reduce demands and offer the regulation menu (quiet room, music, outside walk). Staff present two options visually and allow processing time. The person indicates a preference (points to outside walk). Staff document, in a brief SDM-in-incident note, the choices offered, supports used (visual prompt, processing time, reduced language), and the person’s selection. If the person refuses all options, staff shift to safety positioning and call the on-call supervisor per protocol, while continuing to offer a single bounded choice (“quiet room or sit with me here”).
Why the practice exists (failure mode it addresses): Escalations often become control battles because staff move quickly to directives (“stop,” “go to your room”) rather than offering real options. That increases distress and can lead to avoidable restraint, police contact, or ED utilization.
What goes wrong if it is absent: Staff respond inconsistently: one staff member offers choices; another uses threats or withdrawal of privileges. The person learns escalation is the only way to exert control, and incidents increase in frequency and intensity. Documentation becomes purely behavioral (“aggressive,” “noncompliant”) with no evidence that autonomy-support strategies were tried.
What observable outcome it produces: Providers can show a consistent pattern of autonomy-support attempts during escalation through SDM-in-incident notes. Teams often see fewer restraints, reduced incident duration, fewer repeat escalations linked to schedule disruptions, and better stability indicators (engagement, willingness to re-enter activities after decompression).
Operational Example 2: Preventing “restriction drift” after an incident
What happens in day-to-day delivery: After an elopement incident, staff immediately implement time-limited safety steps (enhanced supervision for 24–72 hours) and log them as emergency actions with a review date. A restrictive decision review is scheduled within 72 hours with the person present using SDM supports (visual agenda, simplified options, trusted supporter). The team maps alternatives: travel training refresh, check-in plan, safer routes, staff accompaniment at specific times, or technology the person agrees to. If a limitation is chosen (e.g., supervised community access for two weeks), it is paired with measurable milestones (route rehearsal completed, check-in reliability) and a step-down plan.
Why the practice exists (failure mode it addresses): Services frequently impose restrictions after incidents and then forget to review them. Restriction drift is a major driver of rights violations and can worsen behavior by reducing autonomy and community participation.
What goes wrong if it is absent: The provider defaults to blanket limits (“no outings,” “door alarms always on,” “must be 1:1”) without time limits or skill-building. The person experiences reduced quality of life and may escalate further. If oversight reviews the case, the provider cannot evidence proportionality, least-restrictive alternatives, or a plan to reduce restrictions—creating high governance risk.
What observable outcome it produces: Restrictions become time-limited and measurable, with documented reduction steps. Providers can evidence governance compliance (reviews, alternatives considered, monitoring) and often see fewer repeat incidents because the focus shifts from control to skill-building and safer independence.
Operational Example 3: SDM-informed behavior support plan updates (including complex needs)
What happens in day-to-day delivery: When repeated incidents occur, the team updates the behavior support plan using an SDM structure. Staff gather “what mattered to the person” data: triggers linked to choice denial, sensory overload, communication breakdown, or unmet goals. The person participates using their preferred communication supports (storyboards, role-play, choice cards). The updated plan includes: proactive autonomy supports (predictable choices, advance notice of changes), bounded choice scripts for early escalation, and a clear list of non-restrictive de-escalation options. Any proposed restrictive element triggers governance review and requires a time limit, monitoring, and a reduction plan.
Why the practice exists (failure mode it addresses): Behavior planning can become staff-centric (“manage behavior”) rather than person-centric (“support needs and autonomy”). Without SDM structure, plans focus on compliance and control, which often increases distress and crisis recurrence.
What goes wrong if it is absent: Plans rely on reactive measures, staff apply inconsistent strategies, and the person’s preferences are not captured in actionable terms. Restrictions are added as a substitute for better communication supports or environmental changes. This can increase safeguarding risk and create poor outcomes such as repeated ED visits, police involvement, or placement instability.
What observable outcome it produces: Providers can evidence that plan updates were informed by the person’s voice and that autonomy-support strategies were prioritized. Over time, services see reduced repeat crisis utilization, fewer restrictive interventions, improved engagement, and better continuity because the plan works across staff teams and settings.
Oversight expectations you should explicitly design for
Expectation 1 (quality/safety oversight): Providers are typically expected to have clear restrictive-practice governance: defined authorization, proportionality, monitoring, time limits, review, and movement toward less restrictive supports. SDM under pressure must connect directly to those controls so restrictions do not become informal, permanent, or undocumented.
Expectation 2 (system/funder): Funders and system partners commonly expect services to be person-centered and rights-respecting while maintaining safety through competent incident management. In practice, that means your record should show both: what you did to reduce immediate risk, and what you did to preserve autonomy (choices offered, supports used, least-restrictive analysis, step-down planning).
Implementation tip: add a monthly “restriction drift” audit. Pull all active restrictions, confirm each has authorization, a monitoring metric, and a review date, and verify that the person’s preference and SDM supports are documented. This single routine prevents many long-term rights and compliance failures.