The Anesthesia Gas Machine
Michael P. Dosch CRNA PhD, Darin Tharp CRNA MS
University of Detroit Mercy - Nurse Anesthesia
This site is https://healthprofessions.udmercy.edu/academics/na/agm/index.htm.
Rev Jan 2024
ANESTHESIA GAS MACHINE> TEST YOURSELF
There are two types: a galvanic type sensor (an older "plug in" type), and the paramagnetic. For the galvanic oxygen sensor, calibrate to room air (the time to 90% response is 15-20 seconds, so if it takes longer than 40-60 seconds to read 21%, change the sensor). Then expose to 100% oxygen and ensure it reads close. You may recalibrate at 100%, but it is not necessary with all monitors.
Newer paramagnetic sensors use internal calibration routines. So they only need periodic (every 3-6 months) exposure to calibration gas, and they last for years. However, I make sure they read 21% when exposed to room air when I do my morning check.
Don't attempt to fix it- you must trust monitors until you can prove they are wrong.
If desaturation is the problem, check midaxillary breath sounds- a common cause of decreased oxygen saturation is endobronchial intubation.
The hospital pipeline is the primary source of all gases, and the pressure within is 50 psi, which is the normal working pressure of most machines. Cylinder oxygen is supplied at around 2000 psi (regulated to approximately 45 psi after it enters the machine). Nitrous oxide cylinders hold a pressure of 745 psi when full. Air cylinder pressures are similar to oxygen.
The hanger yoke: orients cylinders, provides unidirectional flow, and ensures a gas-tight seal. The check valve in the cylinder yoke functions to: minimize trans-filling, allow change of cylinders during use, and minimize leaks to atmosphere if a yoke is empty.
There is a check valve in each pipeline inlet as well. So you can give an anesthetic even when there is no connection to the hospital pipeline, hoses are disconnected, or if a tank is missing.
It is important to recognize that the fail-safe guards against decreased oxygen pipeline pressure and not against crossovers or mislabeled contents. As long as there is any pressure in the oxygen line, nitrous oxide (and any other gases) will continue flowing. If oxygen pressure is lost, the fail-safe shuts off the flow of all other gases.
The hypoxic guard system works on oxygen pressure as well. It controls the ratio of oxygen and nitrous oxide so that there is a minimum 25% oxygen. It does not analyze what is in the oxygen pipeline for the presence of oxygen. Nor does it analyze the FIO2 in the breathing circuit.
The first device to inform one of a crossover will likely be the oxygen analyzer. The second monitor to respond to a crossover (especially if you ignore the first) might be the pulse oximeter, depending on circumstances.
If you do not disconnect the pipeline supply hose at the wall, the pipeline pressure exerted on the oxygen cylinder regulator diaphragm (downstream side) keeps the cylinder gas from flowing, since the pipeline is maintained at a slightly higher pressure (50 psi) than the cylinder regulator (45 psi). The situation is similar to dropping the level of the main intravenous fluid bag when you want a piggyback to run- whichever is higher will flow.
Just like a crossover,
Why? Something is wrong with the oxygen pipeline. What if the supply problem evolves into a non-oxygen gas in the oxygen pipeline? If so, it will flow (pipeline pressure 50 psi) rather than your oxygen cylinder source (down-regulated to 45 psi). If you are lucky, the oxygen alarm will sound to warn you of the change.
If for some reason the oxygen analyzer does not warn of the crossover, the pulse oximeter will- but only after the oxygen has been washed out (by ventilation) from the patient's functional residual capacity and vessel-rich group.
So disconnect the pipeline connection at the wall if oxygen pipeline pressure is lost. It's also easier to remember one strategy which works for any problem with the pipeline, than to remember that sometimes you must, and sometimes it is optional, to disconnect. And use that oxygen analyzer always!
Driving a vent with cylinders will cause their rapid depletion. So manually ventilate the patient, assist spontaneous ventilation if possible, use air or nitrous oxide with oxygen if possible, and use low flows.
We want to find how long the current E tank contents will last. The calculation works for oxygen only. For compressed gases which are stored as liquids (nitrous oxide, carbon dioxide), the relationship between pressure and contents is not proportional.
Calculate:
With an E tank pressure of 1000 psi, and at a oxygen flow of 2 L/min, the tank will last 150 minutes (150 = (0.3*1000)/2). Note that some use a conversion factor of 0.28 (e.g. https://www.respcalc.com/oxygen-tank-duration-study-guide/) but using the easier to remember 0.3 introduces no significant inaccuracy.
The cylinder should be turned off except when checking, or when the pipeline is unavailable- otherwise, silent depletion may occur. Pipeline pressure may decrease below 45 psi with flushing or ventilator use. If it does, oxygen will flow from an opened cylinder. Enough may be lost over a period of days or weeks to empty the tank. Then no reserve will be available if the pipeline supply fails. And the "low oxygen pressure" alarm that sounds will indicate no pipeline or cylinder oxygen available, rather than lack of pipeline oxygen (prompting you to open your full cylinder).
The hypoxic guard system only connects oxygen and nitrous oxide (some gas mixers also take desflurane into account). It is possible to create a hypoxic mixture when you give desflurane in air. Neither traditional machines nor newer gas machines will prevent this. But both will give visible and audible alarms.
Failure of the ventilator relief valve is the most common cause. If sustained high pressure in the chest occurs during mechanical ventilation, ventilate manually (with the breathing circuit). If the ventilator relief valve is at fault, this should be successful. If manual ventilation fails, disconnect the patient from the breathing circuit and ventilate by an Ambu bag. Don't forget to start total IV anesthesia, or assure adequate depth through other means.
The most common site is the Y-piece. Monitors for disconnection (apnea alarms) can be based on gas flow (tidal volume), circuit pressure (if peak inspiratory pressure is below threshold an alarm rings), chemistry (carbon dioxide) or acoustic (sound of the precordial, or normal sounds of the ventilator cycle). The most important is the precordial (or esophageal) stethoscope. Capnography is thought to be more important by some. The precordial is stated as most important in many references because it is inexpensive, reliable (cannot break or fail), and its "alarms" cannot be silenced. In older machines, alarms are not automatically enabled when the csse starts.
Disconnection is the most common preventable equipment-related cause of mishaps. Keep your vigilance high by:
Cleaning the bellows is necessary after anesthetizing a patient with diseases transmitted by oral secretions - so with respiratory disease, one or more of the following approaches should be used. Don't use mechanical ventilators, use bacterial filters at the Y or on each limb, use disposable soda lime assembly, or change soda lime after each case.
Oxygen flush during the ventilator inspiratory phase may cause barotrauma, since excess volume cannot be vented (the ventilator relief valve is closed). Just as the APL valve must be closed during manual ventilation to prevent gas loss to the scavenger, the ventilator relief valve is closed during the inspiratory phase of mechanical ventilation. Instead, if the bellows is empty, increase FGF to 8 L/m. The bellows will refill promptly without oxygen flush (which also dilutes the agent in the breathing circuit).
The disadvantages of the descending (hanging) bellows are unrecognized disconnection (due to their design, they may fill even when disconnected from the patient), and also collection of exhaled humidity in bellows (risking infection and lessening delivered tidal volume). To tell if a bellows is ascending ("standing") or descending ("hanging"), look at them during expiration (remember- ascend and descend have "e"s in them). The modern type is ascending. Only one (somewhat current) machine, the Anestar, uses a hanging bellows, but incorporates capnography and sensors to detect failure of the bellows to fill, both of which may lessen unrecognized disconnects.
Since you may work with a variety of ventilators, all of whom have different controls, safely initiate mechanical ventilation by:
With this sequence you can never go wrong. Don't take for granted that turning a few knobs will cause ventilation- check for chest movement.
A minimum safety test can be done even when time is critically short:
Tight mask fit is the most significant factor, since lack of a tight fit cannot be compensated for by increasing time (because the patient will not breathing 100% oxygen with a loose fit- see Anesthesiology 1999;91:603 at doi 10.1097/00000542-199909000-00006). Every time you place a mask on a patient's face, look back at the breathing bag (to ensure it is fluctuating with respirations) and the oxygen flowmeter (to ensure it is on). Pay attention to complaints that it "smells funny"- you may have left a vaporizer on.
In a traditional machine (Modulus or Excel), no. Increase the fresh gas flow (FGF) to 5 to 8 L/min for an adult (1 to 1.5 times minute ventilation). Petty (and Ehrenwerth & Eisenkraft) claims that this practically does away with the need for soda lime since this semi-open configuration is essentially non-rebreathing. Soda lime can be more easily changed in modern machines, without interrupting ventilation. But do be cautious- there are a number of case reports of inability to ventilate (mechanically or manually) after a CLIC-type (canister) is changed mid-case.
Failure of inspiratory or expiratory unidirectional valves, and problems with carbon dioxide absorbent granules (indicator fails, channeling, exhaustion) are the principal causes of rebreathing. While most instances should be detected by noting the increase in inspired carbon dioxide on the capnograph, it is still worthwhile to periodically review the clinical signs of respiratory acidosis:
Dark blood is not a sign of acidosis.
Keep the indicator float between the lines, and remember that the audible suction sound is an indication that it is functioning properly. This is unlike the closed interface, where if you can hear a hiss, waste gas is escaping into the room. The open interface is safer for the patient (open to atmosphere, so there is no chance of excess positive or negative pressure being transmitted to the breathing circuit), but less safe for the caregiver if you don't know how to use it (potential waste gas exposure).
The smell of gas during a case is abnormal and the cause should be sought. The threshold for smelling volatile agents is quoted as between 5 to 300 ppm, so if you can smell any at all, the concentration is above the NIOSH standard (not more than 2 ppm). Look for:
Reasons related to the scavenger include: open interface with no suction on, closed interface without enough suction, obstructed gas disposal tubing.
Nitrous oxide exposure may be more insidious. It cannot be smelled and it has proven ill effects on the reproductive system (both men & women). If you are concerned, beyond simply not using it, consider disconnecting the gas machine hose from the wall pipeline outlet at the beginning of the day (this junction is a prominent cause of leaks). Make sure your gas analysis system is scavenged. Participate or at least get informed about your department's pollution control program. Fill vaporizers at the end of the day rather than the beginning.
Barotrauma must result unless the same amount leaves the circuit each minute as enters; 4 L/min are exiting.
It's common to decrease fresh gas flow to 1-2 L/min right after intubation. This approach is good and bad. It creates a gradual onset of agent in the brain and thus helps to avoid hypotension. If you have a prolonged period to induce while waiting for surgery to commence, and the risk of awareness isn't high, it's a nice approach. Keep in mind that the redistribution of propofol can be fast, making a return to consciousness possible unless sufficient volatile anesthetic tension is created in the brain soon after induction. If flows are low, you can use overpressure to speed entry of agent into the brain.
Why is induction slower at low fresh gas flow? Imagine a 1 L sink with 1 L/min inflow (of which 1% or 10 mL is methylene blue), and the same outflow. You want to turn the initially colorless water in the sink as blue as the inflow. Think it would go any faster using 5 L/min inflow (of which 1% or 50 mL is methylene blue) and the same outflow? Of course. Not because the concentration is different (both inflows are 1% methylene blue) but because the rate of inflow is a greater proportion of the capacity in the second example.
One time constant (= capacity divided by flow) brings a system 63% of the way to equilibrium; two to 86%; three to 95%. Thus the first of the two systems will take 1 minute to reach 63% of equilibrium (1000 mL capacity/ 1000 mL inflow). The second, higher flow system achieves the same result in 0.2 min (1000 mL capacity/ 5000 mL inflow).
Inflow to the anesthesia breathing system (and thus speed of change of gas composition in the breathing circuit) is controlled by the flowmeters. The capacity of the functional residual capacity (FRC), hoses, and breathing circuit (less than 3 L in an AIsys breathing circuit) can be brought to equilibrium with the inflow more quickly as the rate of inflow increases. A rational approach to assure anesthesia, while conserving volatile agent, would seem to be a "non-rebreathing" induction (fresh gas flow 4-8 L/min for a few minutes) followed by 1-2 L/min during maintenance ("low flow") to conserve tracheal heat and humidity, gases and agent. End-tidal agent approaching close to inspired indicates that flows may be decreased. For a reasonable speed of emergence, choose the higher, non-rebreathing flows.
Ehrenwerth & Eisenkraft 1993 give the formula 3 x FGF (L/min) x volume% = mL used per hour.
No. One can overfill with this method, if the keyed filler is faulty, or the vaporizer dial is "on". It is better to fill vaporizers only to the top etched line within the sight glass (this is the method recommended by GE and Dräger).
There are two filling mechanisms; the funnel "screw-cap filler", and the agent specific keyed filler (notches on the neck of the bottle of agent fit a special pouring device which is keyed to prevent misfilling). The filler port is low to prevent overfilling, but this can be defeated with the method described in the question. Overfilling is dangerous because discharge of liquid anesthetic distal to the vaporizers causes overdose.
True- if you recognize they are empty. Not all vaporizers have low liquid anesthetic alarms. And a paralyzed patient who cannot mount much sympathetic response to lack of agent (elderly, trauma, beta blocked) could be awake with stable vitals.
If tipped more than 45 degrees from vertical, liquid agent can obstruct the control mechanisms and risk overdose on subsequent use. A typical treatment is to flush for 20-30 minutes at high flow rates with a low concentration set on the dial. Check the operating manual for the particular vaporizer, to be sure of the method before attempting it, since the correct procedure differs for each. Only two modern vaporizers can be tipped: the Aladin cassettes in the Aisys, and the Dräger Vapor 2000/3000 (if the dial is set to "T").
To prepare the gas machine:
If the patient develops an acute episode of malignant hyperthermia during operation, the treatment may include
You can contact the Malignant Hyperthermia Association of the United States for further information.
When the patient is more asleep than you are. "Vigilance" and "Watchful Care" are words chosen by anesthesia professional societies for good reason!