blowers
- CPAP+ blower
- 80L/min against 2 PSI
- ~$100
- HEPA filter
- online instrumentation
Activity
- Power supply: 120-240v AC, low voltage (>24v) DC would be great
- Speed control: Flow rate range is ~5 to 80 l/min
- Air source could be blower or pump (e.g. membrane/aquarium/mattress pump style )
- Noise; my memory of the stndard is shay, but CPAP systems are expected to operate ~30-50 dBA (req. added 1-Jul-20)
- Flow and pressure monitoring/alarms on the blower/pump could be (re)used to support patient monitoring
- Mfg. options looked at a few years back considered an off the shelf membrane pump retrofitted with a bio-compatible membrane. This approach could be used for the blower too.
Edited by Christopher Howard
What limits use of these kinds of BLDC blowers?
https://www.digikey.com/products/en/fans-thermal-management/dc-brushless-fans-bldc/217?k=blower
Interesting. In a previous spec we were given an upper bound of 40cmH20 - from an Italian doctor using commercial CPAP 'helmets', which is (nicely) about 40kPa - that's in-bound for a collection (5-6) of COTS parts: these that do 800Pa each. I was tracking that here: https://gitlab.cba.mit.edu/pub/coronavirus/pph-ppe
Bigger versions of the same part exist: this is a 97mm diameter impeller, here's one for 1.5kPa and about the biggest will do 2kPa:
That's kind of irrelevant though if we want 2psi, which is 14kPa. We could of course get 7 of these together, but it's hardly the right approach.
I did spend some time designing a diaphragm (membrane) compressor at Otherlab. Pumps are hard, and this seems like it's in a strange regime of high flow / high(ish) pressure.
Would a regenerative blower be more appropriate, shifting into higher pressures and lower flow? It doesn't look like any of these are available at the right scale, but I can imagine printing a casing / impeller to test, if specs of existing parts look like they're in the right part of parameter space.
Membrane (aquarium) pumps are used in this bubble CPAP device (Pumani), created originally by Rice Uni students in ~2011. (Nice repair video about the system is here: https://www.picturinghealth.org/cpaptraining13/) The membrane pumps we tested had a max pressure of ~20 cmH2O and flow rate of ~20 l/min. It also created a ~ +/- 1cmH2O pressure oscillation ~60Hz, but that may not be a problem for our application. The oscillation pressure our device deliberately creates is about +/1 to 2cmH2O, and the frequency is stochastic. The membrane is a biocomaptibility risk, but could be replaced with a different-but comparable material a la the Pumani, I believe.
I don't recall bumping not the regenerative blower. Looks interesting. Our max flow rate is relatively low (80 l/min to get good breath volumes/minute ventilation (2x1000 ml drager ventilator test lungs to fill at ~60 breaths per min) so that could be an great option.
Back pressure/pressure drop in the bubblevent is minimal so we may be able to get by with higher flow/lower pressure option (e.g. 1 psi) in a pinch. What we have is, essentially, a bubble cpap device with a valve we can control the timing on by adjusting gas flow rate. Our max PIP at this time is 30 cmH20, but we can likely go (much) higher. We're working on a publication so I'm cautious about sharing too publicly yet, but I can send some data snippets and a link to a video of the device running to interested folks privately.
I'm running my entire system off a 1.5 gal., ~US$100 Home Depot Porter Cable compressor. It is WAY to noisy (we need to be ~30-50 dBA and the noise spec listed for this compressor 82 dBA), and I'm not certain it has a suitable duty cycle for our needs (i.e. we need weeks of continuous use). It does show one option out there though. (https://www.portercable.com/products/power-tools/air-compressors/6gallon-oilfree-pancake-compressor/c2002). I need mine currently, so haven't pulled it apart (yet) to see if it could easily be made quieter. My guess is it can be. If the duty cycle could also get there, the next question then becomes bio compatibility of the pump and tank...
We looked at scroll pumps; good pressure and flow performance but very expensive (used to drive pneumatic road drills and support other industrial applications...). Basically a non starter from what we found, but we didn't go very deep on this research. Maybe there is a small version for ~$500 somewhere in the world...
Edited by Christopher Howard
Agree with @bransord re: 'sick' COVID patients eventually end up as really challenging to ventilate and require "real critical care" ventilators... Possibly the simplest 'low cost' ventilator (mode) to manufacture is a vent that's capable of APRV mode, promoted by Dr. Nader Habashi. Read/listen to the APRV/Habashi proponents and you might be lulled into thinking this is the answer, but the topic does have considerable controversy (utility of APRV and in what circumstances).
Pairing a low cost vent with (as @jakeread remarks) a helmet-based (non-invasive) ventilator circuit could be of utility. The ventilator requirements are modest flow and little pressure. (Need just enough flow to overcome the dead space issues associated with the helmet and generate modest positive pressure, i.e 20 cm). Helmet-based ventilation shows promise in reducing mortality in ARDS (though 'sick' COVID may progress differently than and not be entirely the same as 'traditional ARDS', some significant overlap). The U of C folk have published possibly the most authoritative study on the topic. We (including @Zaid.Altawil and the 1st and last author of this study have met with the FDA seeking an EAU for a disposable helmet manufactured by Subslave as the few companies that make FDA approved (or already have their EUA) are backordered.
The other equally important draw helmet based ventilation is that it's associated with less leak than mask-based non-invasive interfaces. We would still need a local negative pressure environment such as @zfredin and @alfonso 's negative pressure COVID box.... or we have another idea specific to the helmet which we're working on to scavenge leak around the neck seal.
The U of C paper is crazy-promising re. interface. Does any one have a sense of the risk of pressure on the eyes? And what is the max. pressure these hoods can handle?
Guidance docs seem to indicate all CPAP machines are now expected to have a way of preventing aerosol from the patient escaping to atmosphere. There were some papers out of China following SARS (or perhaps N1H1) highlighting this risk of contamination that seemed to ended up influencing the initial guidance for CPAP use in Covid. Continuous flow into viral filters has been the anticipated go-to solution, but the cost and supply chain seems to be key non-clinical risks here, so other options could be very useful.
I wonder if there is a pressure differential or positive flow that could be used near or at/between the neck interface to prevent or attract particulate resulting from pressure relief pop-off around neck.
There are some promising low-cost CPAP options emerging, but they seem to need an air source. Bundled together (Air/O2, interface and pressure generator) there could be a very well simple, effective and low cost solution :)
One of the observations with the helmet is that compared to the mask - there is less leakage near the mouth and nose - the origin of the aerosol particles. But I have worn the helmet and while there is no leak at the oropharynx - there is a large leak around your clavicles. Perhaps this is better as the flow is downward toward the patient and the bed - but I think that is speculative.
-
@Jonathan.JILESEN have you tried modeling helmets?
-
@bransord if the helmets are fit well, they leak little.... we have them on campus (several different brands with different, but similar neck seals). We are in the process of putting together our in-service materials, much of which is based off the U. of C stuff (who we've worked with on the FDA EUA).
@Zaid.Altawil trialed a helmet with a Respironics V-60 and pushed it to high CPAP and virtually zero leak as measured by the V-60. @Zaid.Altawil, can you send the video, I seem to have misplaced it.
@neilg and @Jonathan.JILESEN, modeling of helmet leaks has been done. Here's an example of one way to model, (also see attached image) /Screen_Shot_2020-07-05_at_7.49.12_AM.png)
Keep in mind that these authors are trying to prove one brand of helmet leaks less than the others (StarMed with inflatable neck cuff). From what I know so far, the StarMed cuff may or may not achieve a better seal, but is adjustable whereas with the silicone or latex seals you have to 'get it right' the first time... or at least have it too tight on your first try.... method of adjustment being cutting the seal, so you can always make it looser, but never tighter.@zfredin and @bransord, I have a Zotero Group called "COVID-19 Related" that I'll add you to on 'stuff' related to COVID, most of the "Covid_vent_strategies" folder is dedicated to helmet-based ventilator materials and includes links to peer reviewed literature, educational materials (like the U of C stuff) and some proprietary materials (like manufacturer instructions). Click on the See all 93 items for this group in the Group Library on the bottom of the page to expand the selection. The public can see the library but need permission from me to join the group (hence no editing privileges).
@chris.howard pressure around eyes not an issue as it's not for divers, don't even notice it 20cm H20~ 10m underwater. It would be unusual to use more than 20 cm pressure for NIPPV, more likely working with 5-10 cmH20 peak.
Keep in mind, for COVID disease, the issue for most patients (early-mid course in their respiratory failure) is Type 1 respiratory failure (hypoxemic, can't oxygenate), not Type 2 (hypercapnea, can't ventilate). Ventilation strategies are being argued, including how much hypoxemia can be tolerated, how much FiO2 + PEEP is beneficial vs harmful, but the bottom line is that care of these patients consumes a lot of oxygen and a number of the strategies (including helmets, masks and high flow nasal) involve high flow... the combination of which consumes tremendous amounts of oxygen in a hospital with a high COVID patient burden. Hospitals in Lombardy have reported their oxygen lines freezing related to this. It is possible to provide effective CPAP using wall oxygen and a flow generator (simple, 'dumb' technology, old school, like original Downs or Whisperflow generators) which consumes more O2 than a ventilator which is 'smart'. For the helmets require even a bit more flow than face masks to mitigate deadspace concerns.
tagging @jbernard as he's my smart pulmonary critical care colleague here at BU/BMC (and I'm just an 'ER doc' )
Edited by William Baker
