Turning 'Calm and Safe' into Measurable Objectives on a Behavioral Health Unit
A 24-bed inpatient pediatric behavioral health unit is entering schematic design. The steering committee's stated goal is 'a calm, safe environment that reduces seclusion and restraint.' Members disagree sharply about what that means physically. Nursing wants centralized observation with full visibility into every room. Psychiatrists want doors that close, for de-escalation and dignity. Families want open visiting and overnight space. Facilities wants ligature-resistant, low-maintenance, durable fixtures. The unit currently records 6.1 seclusion/restraint events per 1,000 patient days and 14 staff assaults in the past year. Design development starts in four weeks and there are no agreed success measures.
1. Convert the committee's goal into three or four measurable objectives with operational definitions, data sources, and target direction.
Read the worked answer
Each objective needs a construct, a metric, a source, and a direction or target.
- Reduce coercive events. Metric: seclusion/restraint episodes per 1,000 patient days, counted per the hospital's existing CMS-aligned definition, plus mean duration. Source: incident system, already collected. Target: 25% reduction from the 6.1 baseline within 12 months of occupancy.
- Reduce staff injury. Metric: OSHA-recordable assaults per 100 FTE, and lost days. Source: employee health. Target: 30% reduction from the 14-event baseline.
- Improve perceived safety and dignity. Metric: patient and staff safety perception items on an existing instrument, plus family experience items. Source: survey with a stated minimum response plan. Direction: improvement, with a pre-registered threshold.
- Preserve therapeutic engagement. Metric: proportion of awake hours spent in programmed activity off the bedroom, sampled by observation. This is the counterweight that prevents 'safety' from being satisfied by confinement.
Vagueness is expensive later; a POE cannot answer a question that was never operationalized.
2. Two objectives genuinely conflict - full sightline visibility versus closable, private rooms. How do you adjudicate rather than split the difference?
Read the worked answer
Make the conflict explicit and test it against evidence and mechanism, not seniority.
- Restate both positions as hypotheses about mechanism: visibility → faster staff response → fewer escalations. Privacy/closure → lower stimulus and perceived coercion → fewer escalations. Both routes end at the same outcome, which is why they can be compared.
- Ask what the literature actually supports and where it is thin, then identify the design features that serve both: a door with a vision panel and a curtained alcove, observation at the room entry rather than inside the room, a quiet de-escalation room adjacent to the nurses' station.
- Where a true trade-off remains, tie it to the primary outcome and to risk: ligature risk and line-of-sight to a suicidal patient are safety-critical and non-negotiable; privacy of the bathroom and changing area is a dignity and legal requirement. Adjudicate on which failure mode is more harmful and less recoverable.
- Document the decision, the evidence relied on, and the residual uncertainty, so the trade-off can be revisited at POE.
3. What baseline and measurement infrastructure must exist before move-in for these objectives to be evaluable afterward?
Read the worked answer
Retrospective baselines are the usual failure point. Before occupancy, secure:
- 24 months of seclusion/restraint and assault data at event level, not just monthly rates, with the counting rules written down - so a definitional change cannot masquerade as an improvement.
- A fixed denominator plan (patient days, FTE) and agreement on who pulls it.
- A survey administration protocol: instrument, timing (e.g., day 3 and discharge), mode, and a response-rate contingency, since families of minors are hard to reach.
- A short observational audit protocol for activity engagement, with an inter-rater check, run for two weeks pre-move to establish a comparison.
- A confound log started now: staffing model, milieu program changes, admissions acuity - the things most likely to move the outcome alongside the building.
Budget and assign ownership; an unowned measure will not be collected.