Short answer:
These devices can save oxygen because most of the oxygen we deliver by nasal canula is never even inhaled, let alone makes it to the alveoli. Only about 1/3 of the oxygen is inhaled and maybe 1/6 actually gets to the alveoli. By strategically timing how much oxygen is delivered when, you can eliminate much of the waste. And by sensing pressure at the other end of the oxygen delivery tube or a separate sampling tube (both methods are used), we can synchronize the system to the patients respiration using pneumatic control systems (diaphragm operated valves) or electronic control systems (sensors, electronics, microcontrollers, and solenoid valves) inside the oxygen conservation device.
Long answer:
I don't have experience with these devices but my educated guesses have been mostly confirmed by a little research. It appears there is more than one method in use, so there is no single answer for how these device work.
Some sources of waste: Oxygen delivered while you are exhaling blows out into the atmosphere and is wasted. Oxygen delivered to accommodate peak activity levels will in part be wasted when you are at rest. Oxygen delivered late in the inspiration cycle may be caught in dead air space and wasted. Oxygen which is delivered when you are inhaling slowly (at the beginning and end of the inspiration cycle) may be partially wasted because you are breathing less than it is putting out or the flow is slow enough that the oxygen can still escape the nostril. And some oxygen may just escape out the nostril when you are inhaling. So most of the oxygen delivered by a continuous flow nasal canula system is wasted. Oxygen delivered when the patient removes the canula without turning the system off is also wasted.
The pulsed devices appear to deliver a brief high flow rate pulse of air during the peak flow time of your inspiration or from the time they sense inspiration to the time they sense inspiration ends. So most of the oxygen delivered will actually be carried down into the alveoli.
Getting technical about sensing methods: There are a variety of methods that could be used to sense inspiration and/or expiration including pressure, flow, sound, capnogrophy, one way valve position (like a rebreather mask), Doppler shift, or chest diameter, or a combination of methods.. My first guess would be small changes in pressure. It could be sensed by a pressure sensor near the canula, a separate tube to the canula (similar to capnography), or through the same tube that delivers oxygen. Some actually use a dual lumen canula where each nostril tube is divided in half while others use the delivery tube. If only the oxygen delivery tube is used, it is difficult to measure inspiration while oxygen is actually being delivered which is one reason why the oxygen may be delivered as a short pulse on those units. Sensing the movement of air through a sampling tube is somewhat similar to the Pitot tube on an airplane that measures airspeed by comparing pressure from a forward facing tube to ambient pressure. It appears that some devices may use pressure, some may use mass flow, and some may use a combination of the two though I didn't dig too deeply into the patent descriptions. Mass flow can be sensed because a portion of the inspired and expired air will flow through a tube inserted in the nostril if the other end is partially open, though there can be significant resistance in a long narrow tube. I suspect this method isn't used that much in commercially available devices.
The control system could be electronic or pnuematic. It could be powered by battery or by air flow/pressure. Both electronic and pneumatic devices are commercially manufactured. Typically the ones on an oxygen bottle may be pneumatic and the ones used on an oxygen concentrator are probably electronic.
The oxygen supply could be an oxygen cylinder or an oxygen concentrator. The conserving device could be built into the oxygen concentrator or separate for use with a bottle. In the cases of portable battery powered concentrators, the device would be built in so the device can be engineered with a smaller battery and concentrator and a common power source and controller can be used.
The liters per minute setting is going to be confusing on a pulsed device. Is it the peak flow rate, average flow rate, or equivalent flow rate? If I am on 3 liters per minute on a normal canula, maybe I am using 1 lpm and 2 lpm is never even inhaled. Maybe on the pulsed device, I get 1lpm delivered at a peak flow rate of 6lpm for 16% of the time. So is my setting 1lpm, 6lpm, or 3lpm? On another pulsed device, I might get 3 lpm peak flow for 1/3 of the time for an average of 1lpm. Another device might use a 5:1 or 6:1 ratio instead of a 3:1 ratio. It appears, however, that the user setting is probably an equivalent dose to a plain continuous flow system to minimize confusion for patients and providers. Remember to take into account the conservation ratio, though, if you need to calculate oxygen cylinder life.
To add another complication, some devices deliver a set amount of oxygen per minute while others deliver a set amount per breath. In the later, the oxygen consumption and delivery varies with respiratory rate (demand). One could even make a device that increased FiO2 in response to hyperventilation.
In the case of the original patient, the device could have broken, the patient could have been breathing too weakly to trigger the device, the tube could have been pinched, or the patient could have been mouth breathing. Or the problem could have been at least partly psychogenic, possibly triggered by a temporary oxygen shortfall. Your reassuring presence, bootstrapping the patient with higher flow oxygen, and instructions to the patient may have rectified a temporary problem. Your presence and actions may have broken a viscous circle.
These devices could potentially save time and money refilling oxygen cylinders in EMS as well. It would be great if we could use a much smaller , lighter cylinder in a jump bag but we probably need the larger cylinder for patients who need to use 15 lpm non-rebreather and oxygen powered CPAP devices.
