FOR INTENSIVE CARE CLINICIANS: Over decades, many methods for ventilator weaning have been devised.   At one time, the national standard for weaning from a ventilator was to decrease the number of effort free, large tidal volume, cleansing, SIMV breaths, thus allowing the patient to flex their diaphragm in between these synchronized, free breaths. They could pull as many breaths as they were capable, at a depth they were able. As the number of free breaths decreased they were required to make up the difference increasing their effort both in number and strength.  These natural inspiratory efforts placed a demand on the patient to perform the work, between the mechanical breaths, slowly toning their muscles. Once they were down to only 2 or 4/min mechanical breaths they were then considered to be good candidates for extubation and breathing on their own.

With the advent of PSV, Pressure Support Ventilation, the therapist was able to give a preset inspiratory nudge at the start of each spontaneous breath.  This would take the inertia out of the start of the breath and the source of greatest work and inspiratory demand. Our goal was to allow them to do the mid inspiratory work and as they improved, allow them to slowly resume the more difficult leading edge early inspiratory phase work- work that, if loaded on early in the wean caused them to fail more often than not. Bigger tidal volumes resulted. A refinement of SIMV, PSV soon became a stand alone weaning technique with no rate behind it at all, just alarms in case the patient fell off the edge.

But what is PAV?

PAV is a software program. Imagine that we had multiple math coprocessors that looked at numerous factors at the same time.  Suppose that every few hundred milliseconds the program could update the inspiratory flow through a tube, the amount of volume that moves in sections of that time frame, the inspiratory time, the starting pressure and the finishing pressure for the breath.   Through all these numbers, Warren Sanborn, was able to calculate the amount of “draw” on the endotracheal tube throughout the breathing cycle, since we know it’s length and diameter.  He factored in Poiseulle and voila we can now calculate resistance to flow on per breath basis, hence our term RPAV.  The determination of work of breathing is an extension of this number.

PAV uses a split second calculation to determine compliance also.  We call this CPAV. This is the ease of distensability of the lung. It is the reciprocal of elastance.  Compliance is “the change in pressure  given a change in volume.” To determine compliance, a known volume must be given and the airway closed so that the distribution of ventilation equilibrates inside the lung.  Pressure will also equilibrate. In a clever manuever, on random unpredicted breaths, so the brain cannot anticipate the test, the PAV program makes the patient hold their breath 300 milliseconds. a third of a second. The normal brain cannot prepare or predict the timing of this maneuver.

For time immemorial, we, and all who came before us have been taught that we need a time constant of inspiratoiry hold of 1 to 1.5 seconds for the flow and pressure wave to equilibrate in the most distal portions of the lung. This has been proved over and over in anesthetized patients. However, none of our patients are likewise medicated. In fact we have turned down their sedation to encourage good breathing effort. The breath hold time constant of 1-1.5 seconds is especially required for accuracy in measuring the diseased lung, with its blebs, bronchospasms, mucus blockages, atelectatic areas, with endless top to bottom V/Q mismatches etc. Some think a true measurement, with accuracy might require 2 seconds.

But, do we need this accuracy? No.  We are following a trend of CPAV and for our purposes the numbers obtained are accurate enough for usage. If we were to use 2 seconds for our inspiratory hold, most people would push against the closed valve and do a ValSalva manever, raising their pressure, giving faulty readings. Doing this twice a minute would cause anxiety and restlessness  In my questions to Mr Sanborn in an email, he stated that a 300 millisecond hold, is only “80%” accurate.  He said that it isn’t as true a number, but the accuracy doesn’t matter in that we are charting a trend- the real value is the relative index-not the actual number.

Though this sounds much like Washington budget accounting, simply said, we are sacrificing accuracy for measureability and reproducability of a useful trend.  If the trend is off by 20%, we correct for this in our software. The same ventilator alarms are in play as with PSV so there is no danger to the patient.  However, these alarms should not be set loose, since the entire idea behind it is to see the lactic surge coming before we would with PSV, so we can increase support or decrease it, if they are strengthening.  Our goal?  To keep the load level on the muscles consistent with their ability to do work. This is done through a graphic on the screen shows the patient’s amount of work in joules. Work is reduced or increased based upon patient effort to maintain a stable weaning ramp towards extubation.

Who is PAV not for?  Those patients under 25 kg and those without an inspiratory drive.  Neither is it proper for the neurologically hyperpnic.

According to Sanborn, “Patients with neurologically-related asynchronous breathing will not be good candidates for PAV. A patient who actively exhales during the pause will cause the maneuver to fail due to corruption of the raw data. Or if the patient forcibly exhales during exhalation, the RPAV value will be corrupted. If you observe a patient breathing as you described, he/she should be changed to another mode/breath type.

“The PA algorithm has built-in safety pressure and volume limits should the patient become over-supported; but the clinician would be expected to recognize the incompatibility between the selection of the PA algorithm and the wrong patient.”


The defect in PSV is that it is crude in the sense that one size fits all.  It does not quantify the patient’s individual work of breathing and unique parameters that could monitor this work. This is the difference between automobile carbueration and full computerized fuel injection that accounts for temperature and altitude. 

Until now, the patient information as to the sustainability of the wean was gathered by physical assessment. 

We monitored heart rate, blood pressure, flowloops, the nurse’s level of anxiety as well as the patient’s. Again this is the difference between an experienced mechanic listening to an engine run with a diagnostic ear, as opposed to putting it on a computerized display that lists problem codes.

In this, PAV definitely will help lesser experienced therapists make good clinical judgements. It will help all therapists to right size spontaneous support levels. After all, if we are looking at their WOB and seeing that their work is very low, this could give a early weaning signal as to readiness for extubation.  Readiness is less obvious to see than unreadiness. We may observe that they look worse than they are and have more reserves available for breathing on their own than we thought.

Ventilator weaning is much cussed and discussed in our realm, with armies aligned with each position, even those who still believe in T-Piece trials, all with good effect. Can PAV definitively prove itself superior to any and all comers? There is a limitation to all technology, because weaning lengths have not drastically changed.  There is art and science to it.  What is agreed is this. Weaning is a dependant function. We cannot bend the laws of the lung or trick pathophysiology, nor can we bait and switch the aging process.  None of the curves for homeostasis are in the right place. Determining patient readiness is the most intuitive aspect of starting a wean, since breathing costs are so high, as are their metabolic demands given their infections.  

PAV does provide, in LUNGLORD’S opinion, a capability to foretell where the patient will be soon, whether they will crump or not. By dialing up PAV support early, to a point they can meet their glycogen support, we can edge them away from a lactic build into anerobic metabolism and a wean fail.  If we are lucky, we may well keep these 1929 model lungers chugging along, breathing ugly, but ultimately prolonging their weans and getting them to the front door.


About these ads


  1. Great explanation of PAV support!!! I need to add another contraindication for the use of PAV. This would be a volume leak within the system, this could be a chest tube or cuff leak. The complex algorithm needs a closed system to determine compliance and resistance so flow and volume can be accurately maintain. If not maintained the term volume runaway develops and excess volume delivered may occur. As with any technology, it is as good as the clinicians assessment in the indicated use and the goals set to determine the best outcome. I believe PAV has it place along with the other modes of ventilation support with proper assessment. In my practice, the true weaning test is the T-piece. Even if for a minimum of 30 minute trial. Putting the lungs at a normal state and letting the diaphragm maintain negative pressure as nature intended it to be will provide a good assessment of extubation and will minimize reintubations.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s