From the equation of motion to the bedside — one idea, built up until you can name any mode and read any waveform.
Every manufacturer invents its own names — A/CSIMVPRVCVC+APRVASVPAVNAVA — for a small number of underlying behaviours.
It looks like a pile of facts. It is actually one cumulative idea.
The pressure the ventilator delivers, plus whatever the patient's muscles add, must at every instant overcome the elastic load (stiffness) and the resistive load (airways).
Pressure, flow and volume are not three separate pictures — they are one equation plotted over time.
Memorize the normal shapes; every finding is a deviation from them.
Whatever you set becomes a clean square wave and tells you nothing about the patient.
You set volume & flow →
read the patient off PRESSURE
You set pressure →
read the patient off FLOW & VOLUME
Constant flow makes pressure a resistive step (R·V̇) plus an elastic ramp (V/C).
Drag them — watch the step and the slope change while the tidal volume stays the same.
Pressure is the square wave now, so the patient shows up in flow & volume. Flow starts at ΔP/R and decays on the time constant.
Raise R → τ lengthens → flow decays slower, volume fills slower and smaller.
Continuous mandatory. Every breath is a machine breath. Spontaneous efforts just trigger another mandatory breath.
Intermittent mandatory. Mandatory breaths plus spontaneous breaths in between (SIMV and its cousins).
Continuous spontaneous. Every breath is patient-triggered and patient-cycled — pressure support, PAV, NAVA.
Control variable (2) × breath sequence (3) → five basic patterns. There is no VC-CSV.
Targeting scheme = how the machine hits its target: set-pointadaptiveservodual…
PRVC / VC+ / AutoFlow / “volume guarantee” set a tidal-volume target but control pressure breath-to-breath.
By behaviour they are PC-CMVa — pressure control, adaptive — not volume control. The name says volume; the waveform says pressure.
Driven by mean airway pressure and FiO₂.
Driven by alveolar minute ventilation = (VT − dead space) × rate.
Hypoxemia → think PEEP/FiO₂. Hypercapnia → think minute ventilation.
Name the mode
as a tag (control variable + breath sequence + targeting)
Read the load
off the waveform opposite the control variable — elastic vs resistive
Diagnose interaction
phase by phase: trigger → inspiration → cycle → expiration
The reference signal is the patient's effort, Pmus. Interaction has two axes:
In VC, effort scoops the pressure curve concave — “flow starvation”. If it pulls below PEEP the patient is doing work on the ventilator.
If expiratory time is shorter than the lung needs (Te < ~3τ), each breath starts before the last one finished emptying. Gas stacks.
PIP high & Pplat high
→ compliance problem
Pneumothorax, edema/ARDS, atelectasis, mainstem, abdominal pressure, breath stacking.
PIP high & Pplat normal
→ resistance problem
Kinked/bitten tube, secretions/mucus plug, bronchospasm.
The peak-to-plateau gap is the resistive pressure. One hold sorts the whole differential.
Every day, ask: is the reason for intubation improving?
Pair a spontaneous awakening trial (sedation off) with the breathing trial — daily.
Four virtual patients. In your groups, read the waveforms and vitals, propose a change, and watch the patient respond. ARDS · severe obstruction · dyssynchrony · the post-intubation crash.
Scenario console → vent-megacode.pages.dev
Everything in this hour, written to read top-to-bottom — plus the interactive simulator you saw the figures come from.
vent-waveforms.pages.dev
Educational reference built on the published mechanical-ventilation work of Mireles-Cabodevila & Chatburn. Not a medical device; not for clinical use.