Several devices marketed as CPAP devices are not in fact devices designed to deliver constant pressure. For example the Infant Flow device, widely used in NICUs, delivers a constant flow into its circuit, but there is no mechanism to respond to changes in resistance, and the pressure in the circuit is therefore very variable depending on down-stream resistance, and should therefore I think be called a positive airway flow device. It is often referred to as variable flow CPAP, as the flow into the baby varies depending on the phase of respiration of the infant, so this can get very confusing. However, as I said, the pressure is not constant, so CPAP is not the right terminology; and the flow into the circuit is actually constant, it is supposedly the flow into the baby that varies, based on the so-called ‘fluidic flip’.
Now the flow across the prongs changes according to this ‘fluidic flip’ principle, which is supposed to reduce expiratory work of breathing; but as others have pointed out, in normal use the baby does not breathe out through the prongs, most of the expiratory gas leaves the airway through the mouth, or around the prongs, wherever the resistance is lowest. So although under very specific experimental conditions there may be a reduction in respiratory work, I am unsure if this translates into any real advantage in clinical use. If the infants changes head or mouth position, or the prongs are slightly displaced, anything which changes the resistance, the delivered pressure may change dramatically, very often falling well below the desired pressure. I find this most problematic during non-invasive intermittent positive pressure ventilation, we might set up the device to give a peak pressure of 10 and an expiratory pressure of 5, but when I go round to review the baby a few minutes later the pressures being delivered may be 5 over 4, or even less.
In contrast using a ventilator to deliver CPAP uses a very expensive machine to deliver a constant pressure, the pressures can be maintained over a much wider range of infant conditions. A simple constant pressure circuit which is very much cheaper, and probably as good as ventilator derived CPAP, is the so-called bubble CPAP which delivers a constant pressure as a result of the height of the water column, and may deliver some pressure oscillations as a result of the bubbling, more of which below.
Does the difference between CPAP and PAF matter? A new multicenter RCT from Poland randomized 276 babies between 750 and 1500 g birth weight to either the Infant Flow, or CPAP using a Draeger Babylog within the first 6 hours of life. The primary outcome was never needing to be intubated, and there were no substantial differences in outcomes between the groups. (Bober K, Swietlinski J, Zejda J, Kornacka K, Pawlik D, Behrendt J, et al. A multicenter randomized controlled trial comparing effectiveness of two nasal continuous positive airway pressure devices in very-low-birth-weight infants. Pediatr Crit Care Med. 2012;13(2):191-6. Epub 2011/06/15.)
Another RCT from 3 years ago compared the Infant flow device to bubble CPAP. (Gupta S, Sinha SK, Tin W, Donn SM. A Randomized Controlled Trial of Post-extubation Bubble Continuous Positive Airway Pressure Versus Infant Flow Driver Continuous Positive Airway Pressure in Preterm Infants with Respiratory Distress Syndrome. The Journal of Pediatrics. 2009;154(5):645-50.e2.) Samir Gupta and colleagues from the NICU in Middlesborough randomized 140 babies of 24 to 29 weeks gestation when they were extubated. The primary outcome of extubation failure was not different between the groups, however, the failure rate was lower than expected, so the study was underpowered, the non-significant differences in outcomes favored the bubble CPAP group, and subgroup analyses also showed some benefits for bubble CPAP.
A recent review by Thomas Shaffer and his colleagues reviews many of the physiologic effects of non-invasive respiratory support. (Shaffer TH, Alapati D, Greenspan JS, Wolfson MR: Neonatal non-invasive respiratory support: Physiological implications. Pediatric Pulmonology 2012 http://onlinelibrary.wiley.com/doi/10.1002/ppul.22610/abstract) this very complete and clearly written review covers all forms of non-invasive support, including oxygen, CPAP, nIMV and high flow cannulae. They point out that the pressure oscillations which occur during the bubbling of bubble CPAP are far less than those which are produced by a high frequency ventilator, and that under usual circumstances it seems unlikely that there is measurable ventilation from the oscillations caused by the bubbles.
So overall, I think that a true constant pressure system is probably preferable, and that bubble CPAP is as good, and may be preferable, to a ventilator, and is very much cheaper. The main limitation being that bubble CPAP can’t do non-invasive ventilation.