Positive pressure ventilation pushes this whole curve up by increasing the functional residual capacity (through a combination of alveolar recruitment and alveolar distension)
Anyway: tidal volume is 7 ml/kg in a normally breathing subject, according to Kerry Brandis' "The Physiology Viva". If anaesthetic trainees are expected to regurgitate it, it must be true.
In ICU, a patient on CMV traditionally got around 12-15ml/kg. That works out to be roughly 840-1050ml for a 70kg person.
It turns out that this was harmful. It was way too much.
The famous ARDS network study which turned this on its head found that in people with acute lung injury ventilation with lower tidal volumes (something like 4-5ml/kg of true weight, or 6ml/kg of predicted weight) improves survival when compared to 12ml/kg.
What sucks is that the study just had these two volume settings. Nothing in between was tested (i.e. there was no comparison with 7ml/kg, 8ml/kg, 9 ml/kg). This is suboptimal, because most of the people you are going to ventilate will be having volumes in this range.
The consensus of learned elders seems to be that you should use your own judgement, and tolerate smaller and smaller volumes (and thus higher and higher CO2) as the ARDS becomes more severe. Thus, the tidal volumes you ought to aim for in ARDS and ALI are in the territory of 350-420ml for a 70kg person.
The influence of pressure support on tidal volume
Outside of a volume-controlled mode of ventilation, one uses increased pressure to drive deeper breaths. A pressure-support mode is flow cycled; when flow reaches a certain percentage of maximum the expiratory valve is open and the patient exhales.
By changing the targeted level of pressure, one changes the inspiratory flow.
Remember that pressure = flow x resistance, and resistance is not changing, so pressure translates directly into flow.
A higher pressure thus results in a higher peak flow rate.
This flow takes longer to fall to the flow cycling limit.
Another words, the inspiratory time is longer
Because flow x time = volume, this results in a higher volume.
Thus, with an ideally compliant lung and ideally patent airway
(i.e. in absence of resistance), pressure increases will yield predictable increases in tidal volume.
Obviously, no lung in ICU is ideal, and you will find that beyond a point increases in pressure support no longer yield satisfactory increases in volume, but only more pressure-related complications.
This is not an exceptionally intelligent setting on the ventilator.
You are telling the time-cycled mode how often to initiate a breath.
Alternatively, in a patient-controlled mode, you are setting a "backup rate"- i.e. once the patient stops breathing for whatever reason, the ventilator will take over, puffing away at your mandated rate.
Along with volume, frequency can be adjusted to change the rate of CO2 clearance (which is dependent purely upon ventilation).
In the abovementioned ARDS situation, you are limited to increasing your frequency as your tidal volumes drop.