Notice. Illustration and research prototype. Not clinically validated, not a medical device, not for patient care. Example data are fictional.
Live demonstration. The sliders and test cases work — every score recomputes in your browser. The data is generated, no patient is involved, and nothing here has been verified by a clinician.

sedation.yaml

Configuration · sedation.yaml · 171 lines · editable in place

SEDATION DEPTH ============================================================================ Idea 39. Deterministic: is this patient at their target depth, was a target ever set, when were they last scored, and is the drug the one the guideline recommends.

WHY IT MATTERS MORE THAN IT SOUNDS ---------------------------------- Over-sedation is endemic and invisible without scoring. It drives delirium, ventilator days, and cost. The ABC trial paired awakening with breathing trials and cut ventilator days by 3.1 and one-year mortality by 14% — that is a larger effect than most drugs in critical care, from a protocol.

The failure is not disagreement. Almost nobody argues for deep sedation. The failure is that a target is never set, or is set and never checked, and a patient sits three points below it for a week while everyone assumes someone else is watching.

Sources: SCCM PADIS guidelines and the 2025 focused update. Richmond Agitation-Sedation Scale (Sessler et al.). ABC trial (Girard et al., Lancet 2008).

meta
version0.1.0
ownerUNASSIGNED — set to the responsible clinician
verified bynot set
rass
scale
- {score4, label: "Combative", desc: "Overtly combative, violent, danger to staff"}
- {score3, label: "Very agitated", desc: "Pulls at tubes or catheters, aggressive"}
- {score2, label: "Agitated", desc: "Frequent non-purposeful movement"}
- {score1, label: "Restless", desc: "Anxious, movements not aggressive"}
- {score0, label: "Alert and calm", desc: ""}
- {score-1, label: "Drowsy", desc: "Not fully alert, sustained waking to voice over 10s"}
- {score-2, label: "Light sedation", desc: "Briefly wakes to voice, under 10s"}
- {score-3, label: "Moderate sedation", desc: "Movement to voice, no eye contact"}
- {score-4, label: "Deep sedation", desc: "No response to voice, movement to physical stimulus"}
- {score-5, label: "Unrousable", desc: "No response to voice or physical stimulus"}
assess every hours4
note
Ten points, -5 to +4. Scoring takes under a minute and is the only way
over-sedation becomes visible.
target
default range[-2, 0]
default labelLight sedation
tolerance0
deep sedation indications
"Neuromuscular blockade in use"
"Raised intracranial pressure"
"Severe ventilator dyssynchrony not otherwise controllable"
"Refractory status epilepticus"
"Severe ARDS in the early phase, where prone or paralysis is planned"
require indication below-3
indication note
Below RASS -3 an indication must be recorded. Without one the patient is
deeply sedated by default rather than by decision.
drugs
preferred
- namepropofol
noteShort acting, but causes hypotension and carries infusion syndrome risk.
- namedexmedetomidine
noteLess delirium; bradycardia and hypotension are the limits.
avoid
- namemidazolam
why
Benzodiazepine. PADIS suggests propofol or dexmedetomidine over
benzodiazepines — less delirium, fewer ventilator days, better
mortality. The active metabolite also accumulates in renal impairment.
- namelorazepam
whyBenzodiazepine, same reasoning.
- namediazepam
whyBenzodiazepine, same reasoning.
benzodiazepine indications
"Alcohol or benzodiazepine withdrawal"
"Status epilepticus"
"Procedural sedation"
note
A benzodiazepine is not forbidden — it is the default that is wrong. Where
one of the indications applies, record it.
interruption
required dailytrue
pair with breathing trialtrue
evidence
ABC trialpairing the awakening trial with a breathing trial reduced
ventilator days by 3.1 and one-year mortality by 14%.
safety screen
"Active seizures"
"Alcohol withdrawal requiring sedation"
"Agitation requiring escalating sedative doses"
"Neuromuscular blockade in use"
"Raised intracranial pressure"
"Myocardial ischaemia in the last 24 hours"
failed screen note
A failed safety screen is a reason, not a failure. It must be recorded so
the difference between "unsafe today" and "nobody did it" is visible.
checks
- idno_target_set
severitywarn
textNo sedation target recorded
why
Without a target, depth cannot be too deep — which is how a patient
sits at -4 for a week with nobody accountable.
- idnot_scored
severitywarn
textRASS not scored within the assessment interval
whyOver-sedation is invisible without scoring.
- iddeeper_than_target
severitywarn
textDeeper than target
whyEach point below target adds delirium risk and ventilator days.
- iddeep_without_indication
severitycritical
textDeeply sedated with no recorded indication
why
The distinction that matters. Deep by decision and deep by drift look
identical on a chart.
- idbenzodiazepine_default
severitywarn
textBenzodiazepine without a recorded indication
whyPADIS suggests propofol or dexmedetomidine over benzodiazepines.
- idinterruption_not_done
severitywarn
textDaily sedation interruption not done
whyABC trial: 3.1 fewer ventilator days, 14% lower one-year mortality.
- idagitated
severitywarn
textAgitated above target
why
Under-sedation is also a failure, and usually means pain rather than a
need for more sedative. Assess analgesia first.