Why a patient needs a ventilator and what the clinician sees
Translate a clinician's complaint about a ventilator - wrong volume, PEEP not held, oxygen alarm, 'it breathes by itself' - into a technical question about a specific part of the machine, its settings or its accessories.
The job a ventilator does
Breathing has two jobs. The first is to move gas in and out of the lungs, so that carbon dioxide leaves the body. The second is to get enough oxygen across the lung into the blood. A patient ends up on a ventilator when one or both jobs fail: the breathing muscles are too weak or too tired, the brain is not driving breathing (deep sedation, anaesthesia, injury), or the lungs are so stiff or so full of fluid that normal effort cannot inflate them.
An ICU ventilator takes over or assists the first job by pushing gas into the lungs under positive pressure, then letting the lungs empty passively. It helps with the second job by raising the oxygen concentration of the gas and by keeping a little pressure in the lungs at the end of each breath so that the small air sacs do not collapse. That is all the physiology this course needs. You do not set the therapy; the clinician does. Your job is to make sure the machine delivers exactly what is set, measures exactly what is delivered, and raises an alarm when either goes wrong.
Keep one fact in front of you at all times: the ventilator breathes for a person who usually cannot breathe alone. A ventilator that stops, or delivers the wrong breath, harms the patient in minutes. Everything in this course, from the gas supply to the final verification, follows from that.
The words on the screen
You will hear these terms in every complaint. Learn what each one means physically, because each one points to a different part of the machine.
Pressures on ventilators are shown in cmH2O (centimetres of water) or in hPa or mbar. These are almost the same: 1 cmH2O is about 0.98 hPa, and 1 hPa equals 1 mbar. Volumes are in millilitres, flows in litres per minute. Minute volume is simply tidal volume multiplied by respiratory rate. If a nurse says the minute volume is low but the rate is correct, the question is about volume.
- Tidal volume (Vt): the volume of one breath. Delivered volume is measured on the inspiratory side; exhaled volume is measured on the expiratory side. When the two disagree, think of leaks or a flow sensor.
- Respiratory rate: breaths per minute, set by the clinician or produced by the patient.
- Inspiratory time or I:E ratio: how long the breath in lasts compared with the breath out.
- FiO2: the fraction of inspired oxygen, from 21% (room air) to 100%. It depends on the gas supply, the mixing valves and the oxygen sensor that monitors it.
- PEEP: positive end-expiratory pressure, the pressure the machine holds in the circuit at the end of exhalation. It is produced mainly by the expiratory valve.
- Peak pressure: the highest airway pressure during the breath. Plateau pressure: the pressure during a short pause at the end of inspiration, when flow is zero.
- Trigger: the threshold at which the machine decides the patient is trying to breathe and starts or supports a breath. It is either a pressure drop or a flow change in the circuit.
Modes: who decides the breath
Manufacturers use dozens of mode names, many of them trademarks. Underneath, almost all modes are combinations of three ideas. If you know which idea a mode is built on, you know which reading should be stable and which one is allowed to change.
In volume-controlled ventilation the clinician sets the tidal volume and the flow pattern. The machine delivers that volume, and the airway pressure is whatever it takes to push it in. If the lungs get stiffer or the tube gets blocked, the pressure rises; the high-pressure alarm protects the patient. So in volume control, a complaint of 'pressures are high' is often a patient or circuit problem, while 'the volume is wrong' points at the machine's flow delivery or measurement.
In pressure-controlled ventilation the clinician sets a pressure and an inspiratory time. The machine holds that pressure, and the volume is whatever the lungs accept at that pressure. If the lungs get stiffer, the volume falls - correctly. So in pressure control, a complaint of 'the volume dropped' may be the patient, not the ventilator, while 'it does not reach the set pressure' points at the inspiratory valve, a leak, or the supply.
In pressure support, and in other spontaneous modes, the patient starts every breath. The machine detects the effort through the trigger, adds a set pressure to help, and ends the breath when the patient's flow falls. If the patient stops breathing, a backup (apnoea) ventilation takes over and an apnoea alarm sounds. Many modes mix these ideas: synchronised mandatory breaths with pressure-supported breaths between them, or a pressure-controlled breath that adjusts itself to reach a target volume. When you meet a mode name you do not know, open the operator's manual and find which of the three ideas it is built on.
PEEP, FiO2 and the trigger - the three settings that generate most calls
PEEP is held by the expiratory valve, which closes enough at the end of exhalation to keep the set pressure in the circuit. If PEEP drifts or cannot be reached, the usual suspects, in order, are a leak in the circuit or around the patient's tube, an expiratory valve that was reassembled wrongly after cleaning, a worn or torn valve diaphragm, and only then the valve drive or pressure measurement. A leak anywhere makes PEEP hard to hold, because the valve can only stop gas leaving through itself.
FiO2 depends on two gas supplies and on the mixing. A low or unstable oxygen reading points at the oxygen sensor first - especially a galvanic cell, which is a consumable that wears out - then at the oxygen supply pressure, then at the mixing valves. Never 'fix' an FiO2 complaint by calibrating the sensor until the reading matches what you hope is delivered. Measure the delivered oxygen with an independent, calibrated analyser first.
The trigger is where clinical setting and machine behaviour meet most often. If the trigger is set very sensitive, small pressure or flow changes - water moving in the tubing, a small leak, even the heartbeat transmitted through the lungs - can start breaths the patient did not ask for. Clinicians call this 'the ventilator breathes by itself' or 'the rate is too high'; the term is auto-triggering. If the trigger is too insensitive, or the circuit is leaking, the patient struggles and the machine responds late. Both look like faults, and both can be caused by settings, condensate or leaks rather than by a broken part.
From complaint to technical question
A complaint arrives in the language of the bedside. Your first task is to turn it into a question that you can test on a bench. Ask what mode was running, what was set, what was displayed, which alarm appeared, whether it happened with one patient or with several, and whether the circuit, filter or humidifier had just been changed. Ask to see the alarm log if the machine keeps one. Write the answers down before you open anything.
Then sort the complaint into one of four places: the settings and the patient, the circuit and accessories, the gas and power supply, or the machine itself. Experience across ICUs is consistent: most 'ventilator faults' turn out to be the circuit, water traps, a wet expiratory flow sensor, or a wall gas supply that holds pressure at rest and sags when gas flows. The engineer who opens the machine first loses time and parts.
Remember also what you are not. You do not change clinical settings on a patient, and you do not decide whether a patient can stay on a machine you suspect. If a clinician calls you to a ventilator in use, the patient is first moved to a backup - a manual resuscitation bag or another ventilator - by the clinical staff, with your help if they ask for it. Diagnosis starts after that, on a test lung.
- 'Volume too low' in volume control: inspiratory flow delivery, flow sensors, leak.
- 'Volume too low' in pressure control: possibly the patient; check the circuit and leak before suspecting the machine.
- 'Exhaled volume does not match': leak, wet or dirty expiratory flow sensor, sensor calibration.
- 'PEEP not held': circuit leak, expiratory valve assembly and diaphragm.
- 'Oxygen alarm': oxygen sensor, oxygen supply, mixing - in that order, verified with an independent analyser.
- 'It breathes by itself': trigger setting, condensate in the circuit, small leaks.
Why the self-test is not the end of the story
Every ICU ventilator has a self-test or pre-use check. It is valuable and must be run exactly as the manufacturer describes, but it checks the machine against its own sensors. If a flow sensor reads 10% high and the controller trusts it, the machine will deliver a smaller breath than set and still report everything as normal. Only an independent, calibrated gas-flow analyser connected to a test lung tells you what a patient would actually receive. Lessons 5 and 7 cover this in detail; for now, carry the idea into every conversation with the ward: 'passed the self-test' is not the same as 'delivers what is set'.
The ventilator that 'breathed by itself'
An ICU called about a ventilator that, right after ventilation started, ran an erratic rate and PEEP for about half a minute before settling. The checkout passed and the leak was low. On the bench, with a small test lung and a large set tidal volume, the peak pressure also looked high, and the ward suspected the expiratory valve. The engineer ran the same settings and the same test lung on a second ventilator of the same model: the peak pressure was the same. Reducing the trigger sensitivity removed the erratic start completely. The air inlet water trap was also checked, because a wet air supply can put water into the flow valves, but it was dry.
Before you suspect a valve, check whether the 'fault' is the trigger setting, the test set-up or the circuit. Compare with a second machine under identical conditions, and only then go inside.
The leak that was a setting
A service engineer was called to an ICU ventilator because the clinicians saw an exhaled tidal volume about a fifth lower than the delivered volume and suspected a leaking circuit or a failing flow sensor. Before opening anything, the engineer looked at the configuration and at how the patient was connected. Leak compensation was switched off while the patient had a noticeable leak around the airway. On a second machine of the same type the oxygen source had been configured as low-pressure, so the FiO2 control did not appear on the screen. On a test lung with the specified adaptor both machines delivered and measured the set volume within the manufacturer's limits. The engineer went through the settings with the staff, recorded the findings and left the hardware untouched.
A complaint about volumes or oxygen is a question, not yet a fault. Check settings, compensation options and how the number is measured, then prove the machine on a test lung before replacing parts.
On-site checklist
- Record mode, settings, displayed values and alarm text exactly as the clinician reports them.
- Ask whether the problem follows the machine, the patient or the circuit (did it happen with another circuit or on another patient?).
- Ask what was changed recently: circuit, filter, humidifier, expiratory valve after cleaning, gas outlet.
- Confirm the patient has been moved to a backup before any diagnosis.
- Reproduce the complaint on a test lung with the same mode and settings, and compare with a second machine if one is available.
- Do not adjust or calibrate a sensor to match an expected value; measure with an independent analyser.