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.

weaning.yaml

Configuration · weaning.yaml · 194 lines · editable in place

WEANING READINESS — spontaneous breathing trial screen ============================================================================ Idea 14. Deterministic: a screen against published criteria, run on every ventilated patient every day, which is exactly the checklist nobody runs daily under pressure.

WHY THIS IS WORTH COMPUTING --------------------------- The criteria are not difficult and not disputed. The failure is that a ventilated patient on a busy morning does not get screened, and each day of unnecessary ventilation carries pneumonia risk, delirium, and cost.

Bundle element B pairs the awakening trial with the breathing trial and reaches only 10-17% delivery in published point-prevalence data. This engine answers the question that element asks: is this patient ready today.

A NOTE ON RSBI -------------- The rapid shallow breathing index (respiratory rate / tidal volume in L) is the classic teaching, threshold 105. The AARC 2024 clinical practice guideline states it is NOT needed to determine readiness — it adds little to the other criteria and delays trials in patients who would have passed.

It is therefore computed and displayed but NOT used as a gate. That is a deliberate departure from what many clinicians were taught, so it is stated rather than buried.

Sources: AARC Clinical Practice Guideline: Spontaneous Breathing Trials, 2024. Standard weaning protocols (Penn, Colorado SICU, AHRQ SAT/SBT).

meta
version0.1.0
ownerUNASSIGNED — set to the responsible clinician
last clinical reviewnot set
verified bynot set
note
Thresholds vary between units. These are common values and must be
reconciled with the unit's own weaning protocol before use.
criteria
- idcause_improving
nameCause of respiratory failure improving or resolved
kindclinical
why
The single most important criterion and the only one requiring
judgement. Cannot be computed — it is asked, not measured.
requires cliniciantrue
- idoxygenation_pf
nameAdequate oxygenation
kindnumeric
variablepao2_fio2_ratio
operator>=
value150
alternative
variablespo2
operator>=
value90
why
PaO2/FiO2 at least 150, or SpO2 at least 90%. Below this the patient is
still dependent on the support being withdrawn.
- idfio2_low
nameFiO2 at or below 0.40
kindnumeric
variablefio2
operator<=
value0.40
whyA trial on high FiO2 tests the oxygen, not the patient.
- idpeep_low
namePEEP at or below 8 cmH2O
kindnumeric
variablepeep
operator<=
value8
why
Commonly stated as 5, but 8 is widely used and avoids delaying patients
who would pass. Set to 5 if the unit protocol says so.
- idhaemodynamic_stable
nameHaemodynamically stable
kindnumeric
variablemap
operator>=
value65
why
Adequate perfusion pressure without escalating vasopressor support. A
low or rising pressor requirement is a separate check below.
- idvasopressor_low
nameNo high or escalating vasopressor requirement
kindvasopressor
max dose mcg kg min0.15
why
A low stable dose does not preclude a trial. An escalating requirement
does — the patient is not stable, whatever the current number says.
- idno_neuromuscular_blockade
nameNeuromuscular blockade stopped
kindflag
flagneuromuscular_blockade
expectfalse
whyA paralysed patient cannot breathe spontaneously.
- idadequate_mentation
nameAble to protect the airway
kindnumeric
variablegcs
operator>=
value8
why
Consciousness sufficient to cough and protect the airway. GCS 8 is a
common floor; some protocols require a cough and gag assessment instead.
- idacid_base
nameAcceptable acid-base status
kindnumeric
variableph
operator>=
value7.25
whyProfound acidosis increases respiratory drive and workload.
- idsecretions_manageable
nameSecretions manageable
kindclinical
why
Frequent suctioning predicts extubation failure even when everything
else passes. Asked, not measured.
requires cliniciantrue
rsbi
computetrue
gatefalse
threshold105
formularespiratory rate / tidal volume in litres
guidance
AARC 2024 states RSBI is not needed to determine readiness for a trial.
Values above 105-120 predict a patient who may not tolerate one, so it is
shown as context. It does not block a trial here, which is a deliberate
departure from the classic teaching.
trial
duration minutes30
acceptable range minutes[30, 120]
modes["T-piece", "CPAP 5", "PSV 5-8 over PEEP 5"]
note
30 minutes is as informative as 120 in most patients. Longer trials tire
the patient without adding information.
failure signs
"Respiratory rate above 35 for more than 5 minutes"
"SpO2 below 90%"
"Heart rate above 140 or a 20% rise"
"Systolic pressure above 180 or below 90"
"Agitation, diaphoresis, or distress"
"New arrhythmia"
on failure
Return to the previous settings and rest for 24 hours. Repeated failed
trials in one day fatigue the patient and delay liberation.
cadence
screen every hours24
earliest after intubation hours12
pair with awakening trialtrue
pair note
Bundle element B counts only when BOTH the awakening and breathing trials
were done. Screening for one without the other is how that element reaches
10-17% delivery.