Search This Blog

Follow by Email

Thursday, February 27, 2020

I'm back

Its  been 11 years since I last posted.   I am no longer a dialysis tech working in a hospital.  I now engineer, design, build and test new technology for dialysis full time at a tech startup.

The test equipment I had and the prototypes of home built ventilators that my son and I made when the H1N1 virus was of concern sat for a long time in my basement collecting dust.  Ironically, I cleaned out that room in December 2019 and threw out most of it.

Years came and went and no major pandemic ever appeared.  H1N1 was much less lethal than many expected, and MERS was largely contained.  I was happy for that.  Although I warned 11 years ago of the risks of not having enough ventilator surplus capacity to deal with a serious problem, I was happy that no serious emergency ever occurred.  I thought we were safe.  I hope we still are safe, and I pray that all the reports we hear are drastically exaggerated and nothing bad will happen.

So now we have a new coronavirus now called SARS-CoV-2, causing the disease Covid-19.
A lot of people are concerned.  The stock market certainly seems to be concerned at the end of February.  And suddenly there is a large spike in traffic on this blog again.  So how do we know if this will be severe or if it will turn out to be nothing?  First we look at what the experts have to say:

Here is a good article from the New England Journal of Medicine.

  • Basically they are concerned but it is still too early to say how bad it will be.  
  • They say that the estimated case fatality ratio among medically attended patients so far is approximately 2%.

Here is an article on estimating the demand for mechanical ventilators in a pandemic.

  • The highest  estimated case fatality ratio they considered was 0.5%
  • They used an estimate of a clinical attack rate of between 20% to 30%.
  • Two of the experts in the list below predict a clinical attack rate of 60% or even 70%.  I hope it is not that high or we will have a lot of excess deaths due to lack of basic ventilator support.

Among other factors, epidemiologists are interested in the severity of the outbreak and what is known as the virus’ “attack rate.” The attack rate means the percentage of people who will develop the disease if it spreads.  With seasonal influenza, the Centers for Disease Control and Prevention estimates between 3% and 11% of people become sick with flu every year. But immune systems have experience with influenza; the attack rate might be higher with a virus such this that is wholly new to humans.

So, have hospitals stockpiled enough ventilators to deal with this risk?
Here is an article from th Baltimore Sun in 2016 that deals with this issue:

  • "Maryland's attorney general issued an opinion in December that found that the Catastrophic Health Emergencies Act of 2002 gives the governor the authority to ration the machines that support breathing in flu patients suffering respiratory complications."   "The flu typically sickens and kills thousands in the United States every year, but pandemics remain uncommon. During the last one, in 2009, a vaccine hurried into production was credited with stemming a virus that led to about 12,000 deaths."
  • That 2009 one was the one that was worrisome when I started the Pandemic Ventilator Project.
  • "A severe pandemic like the Spanish flu of 1918, which led to more than 500,000 U.S. deaths, likely would necessitate rationing.  A severe pandemic today could sicken 90 million people and lead to 1.9 million deaths, the planners estimated. Some 740,000 ventilators could be needed."
  • That death rate of 2% is very similar to current projections for COVID-19.  That number is for a clinical attack rate of 30%.  Some experts predict a clinical attack rate of 60% for COVID-19.  so the death rate and ventilator need could then be twice that number.
  • "Studies show U.S. hospitals have about 70,000."
  • Not good news.
  • "Fortunately, we have not experienced a pandemic event that has resulted in a ventilator shortage in the U.S," said Alan C Regenberg, director of outreach and research support at Hopkins' Berman Institute of Bioethics.  But "even a more moderate event would push our current capacity to its limits," he said, "and a severe event would leave us with major mechanical ventilation shortages."

Here are a few more collected expert opinions on the coronavirus outbreak

 The director general of the WHO has recently spoken of a narrowing of the window of opportunity to control the current epidemic. The tipping point - after which our ability to prevent a global pandemic ends - seems a lot closer after the past 24 hours
Prof. Paul Hunter
Professor of Health Protection
University of East Anglia
Feb. 23, 2020

 The number of reports from multiple different countries in the past 36 hours showing what is most likely community human to human spread of SARS-CoV-2 confirms fears that the virus is on its way to causing a pandemic
Prof. Dr. Benhur Lee, MD
Professor of Microbiology
Icahn School of Medicine at Mount Sinai (ISMMS)
Feb. 21, 2020

 All predictions are important. Most predictions are wrong. And I think we must be careful with that.
[asked about prediction that Covid-19 could affect 2/3 of world's population]
Dr. Michael Ryan
Executive Director
WHO Health Emergencies Programme
Feb. 17, 2020

 I think it is likely we will see a global pandemic. If a pandemic happens, 40% to 70% of people world-wide are likely to be infected in the coming year. What proportion is asymptomatic, I can't give a good number
Prof. Marc Lipsitch
Prof. of Epidemiology, Harvard School of Public Health
Head, Harvard Ctr. Communicable Disease Dynamics
Feb. 14, 2020

 I think this virus is probably with us beyond this season, beyond this year, and I think eventually the virus will find a foothold and we'll get community based transmission and you can start to think about it like seasonal flu. The only difference is we don't understand this virus
Dr. Robert Redfield
Director, CDC
US Centers for Disease Control and Prevention
Feb. 13, 2020

This is really a global problem that’s not going to go away in a week or two.
What makes this one perhaps harder to control than SARS is that it may be possible to transmit before you are sick.
I think we should be prepared for the equivalent of a very, very bad flu season, or maybe the worst-ever flu season in modern times.
Prof. Marc Lipsitch
Prof. of Epidemiology, Harvard School of Public Health
Head, Harvard Ctr. Communicable Disease Dynamics
Feb. 11, 2020

 I hope this outbreak may be over in something like April
Prof. Nanshan Zhong
Leading epidemiologist, first to describe SARS coronavirus
Feb. 11, 2020

 It could infect 60% of global population if unchecked
Prof. Gabriel Leung
Expert on coronavirus epidemics
Chair of Public Health Medicine
Hong Kong University
Feb. 11, 2020

 It’s a new virus. We don’t know much about it, and therefore we’re all concerned to make certain it doesn’t evolve into something even worse
Prof. W. Ian Lipkin
Epidemiology Director
Columbia University
Feb. 10, 2020

 We are estimating that about 50,000 new infections per day are occurring in China. [...] It will probably peak in its epicenter, Wuhan, in about one month time; maybe a month or two later in the whole of China. The rest of the world will see epidemics at various times after that.
Prof. Neil Ferguson
Director, Institute for Disease and Emergency Analytics
Imperial College, London Feb. 6, 2020
 This looks far more like H1N1’s spread than SARS, and I am increasingly alarmed
Dr. Peter Piot
(Director, The London School of Hygiene and Tropical Medicine)
Feb. 2, 2020

 It sounds and looks as if it’s going to be a very highly transmissible virus [...] This virus may still be learning what it can do, we don’t know its full potential yet.
Robert Webster
(Infectious disease and avian flu expert at St. Jude Children’s Research Hospital)
Feb. 2, 2020

Click on the word Comments below 
(if you do not see the comments)


  1. When you did your initial work it looks like you were using a Windows based control system. Is that program you created available?

  2. Could a raspberry pie device be used as a controller?

  3. Hi Brendan,
    A Raspberry pie would definitely work. When you look at what I did over 10 years ago, I preferred to use PLC devices because they were simple and readily available. That windows program was Labview by National Instruments. It is proprietary software though and could be expensive to obtain and so put it out of reach for third world country use. We got the software for free back then when we explained what we were using it for. The software and the programming is now gone. I still have the laptop around somewhere but it was wiped and converted to Linux when WinXP support ended.

    There are many good free alternatives for graphical programming for such control systems now. I do not think PLCs are the optimal control solution today. There are many raspberry pies around that could be repurposed to function as controllers. The nice thing about the raspberry pie is that it directly supports memory, WiFi, Bluetooth, and a monitor through HDMI. This could make for an easier to use system that looks very professional.

    Another really good alternative is Arduino as there is a large choice in control type shields and may it be easier to program for control logic.

  4. We discussed this a bit 11 years ago. I am glad the PVP was not needed then and I hope it wont be needed now. But I am also glad you are back in the game in case PVP is needed

  5. Very simple diy ventilator

  6. I'm looking at starting an open source project for a portable ventilator more aligned with specs from US BARDA for a Mass Casualty Ventilator, the more general term than "Pandemic Ventilator".

    Open source to the World is much different than DIY.

    Anyone have suggestions on how to make the world aware and launch it?

    John Strupat
    London ON CANADA

    1. Hello John,
      I'm gathering maker/diy solutions here: I hope it can be somehow useful

  7. Wow, that is a very interesting news article. Thanks. That was from 2013. At that time he was already kept alive for 5 years by his family. It is quite amazing to to have such a family that will do anything to keep you alive. Just look at what it did to their hands. Their ventilator was basically just an emergency resuscitator connected to a tracheotomy tube. It looks like they automated it by connecting a shaft to an an electric motor driven rotating arm. There is no monitoring, and very minimal controls, yet it works to keep someone alive for a long time.

    Here is an example of a resuscitator you can buy. ( for now anyway)

  8. Hey John,

    Good to hear from you again. That is a wonderful idea. Today, the very best ways of raising awareness and participation is through the various social media. Facebook, Twitter, Instagram, Reddit, Youtube etc. In particular you need to have a method of creating a viral (no pun intended) phenomenon. The mainstream media tends to get many of their clues as to what is newsworthy today from what is "trending" on social media. In particular the best way is to get an "influencer" to notice your story. If you can tell a compelling story that will get every reader on Facebook to forward it to 2 or more others it will become widely shared.

    To get started here are some links:

    Top Social Media sites

    Top Media Influencers

  9. Good advice!

    I've just posted to Facebook and am exploring whether CBC would like to do a story.

    Caution to everyone here that a ventilator for short term life support support during a pandemic needs much more capability than that needed for chronic ventilation support during paralysis.

    The clock is ticking as the virus spreads so fingers crossed.

  10. To Dreamer,

    Could you give us an updated list of parts/specs and instructions with the raspberry pi computer system in mind for an updated pandemic ventilator? I think your instructions for the original pandemic ventilator are great but with the new technology on the market and the pandemic looming around the corner the public could benefit from this. I understand if you don't have the time. Many of us are capable of assembling such a device but lack the programming skills to make it work correctly with the pi.

    I am of the opinion that this virus is going to reach pandemic status and that many of our loved ones will not have access to ventilators due to local health authorities being unprepared. I live in Canada and I have friends who work in the upper ranks of the public healthcare system here and they have told me that our institutions are not prepared for something like what China is dealing with.

    We must take the health of ourselves and our families in our own hands. If this virus has been downplayed or it mutates things could get much worse.

    1. I have not worked on this project for about 10 years now. Basically when the H1N1 flu pandemic came, a lot of people got the flu, but very few required ventilators, so they figured we were OK with what we had. I never got any more interest after that. I agree this could be quite different. Other people have continued on this work though. I suggest you look at this paper. It is a good design.

      Here is a description of their design

    2. So while I agree with you that Arduino or Raspberry Pi would be a better choice for control today. I do not have any thing developed on those lines.

    3. Hi @Dreamer. The MIT group had a follow-up thesis where they improved on their most published version: Powelson, Stephen K. (Stephen Kirby)

      Thank you so much for the Baltimore Sun article.

      Someone said above that it is impossible to use PEEP valves with bellows. However, PEEP valves are commonly incorporated into BVMs. The issue is that an additional valve / outlet is needed to block airflow back into the bag.

    4. Here is a good repository of information on the MIT Ventilator

  11. A pragmatic look at ventilator designs using a mechanism with bags or bellows gives no reason to use them. Newer technologies developed after 1980 offer reliability, small size, light weight and high performance.

    Patients with stiff lungs caused by the virus, and bacteria that soon follow, need high pressure and flow to survive. This is not practical with a bag or bellows design.

    The paralyzed guy in China can be ventilated with a modified manual bag because his lungs are relatively normal, just as most patients in the OR.
    The OR is the only site in a modern hospital where bags or bellows are used.

    A higher tech solution must be used here!

  12. Not practical to use a mechanism with bag or bellows to ventilate a patient with viral lung infection.

    Newer technologies developed after 1980 are better in every way.

    High pressures and flows are necessary for these patient to survive.

    The video of the paralyzed man in China is authentic, but does not portray what you would see with a viral/bacterial pneumonia. Don't be mislead by that video that a bag/bellow mechanism has any value for pandemic life support.

    1. Fair point on the Chinese example. He had a totally different issue.

      >High pressures and flows are necessary for these patient to survive.
      How might people make these machines fast?

    2. Recent designs use extremely high RPM turbine blowers to generate high flow at high pressure. Driven by electric motors that can stop, start and adjust flow in milliseconds. Small package, high reliability and tough construction. A flow sensor is mandatory for measuring the flow into the patient and to calculate the Tidal Volume (Vt). Lot's of engineering needed to make a useful and practical ventilator.

      Huge challenge here is that experienced ventilator users may also be in short supply so the operator interface for a Pandemic Vent must be greatly simplified over a conventional ventilator.

    3. Thanks for the reply.

      >extremely high RPM turbine blowers to generate high flow at high pressure. Driven by electric motors that can stop, start and adjust flow in milliseconds.

      Are suitable high RPM turbine blowers only used in this medical area or are they used elsewhere?

      > A flow sensor is mandatory for measuring the flow into the patient and to calculate the Tidal Volume (Vt).

      Flow sensors seem to be relatively easy to get. I am not sure if they are in the tolerances needed here.

    4. So what would we need for a fan? These are some (very rough) back of the envelope calculations:
      Assume we wanted to have the capability of providing 20 liters per minute to the patient. Now, these ventilators do not run the fan continuously. It is idle for most of the time and then runs at a very high flow rate and then shuts off right again. I think a brushless DC motor with a suitable drive would be a good choice. Let us assume a 20% on-time duty cycle. We would need an output of 100 liters per minute from our fan at whatever back pressure is being supplied by the lung. Most fans are flow rated at zero back pressure, so just to be sure we have enough capacity, let us put in a 10 times overspec. That would be 1000 liters per minute.

      Here on Amazon, for less than $20 is a brushless 12V DC centrifugal fan rated for 38.5 cubit feet per minute. That is 1,089 liters per minute.

      Wathai Brushless Cooling Blower Fan 120mm x 32mm 12V DC Centrifugal Fan High Airflow
      by Wathai

    5. Here is a Swiss-made blower from Micronel:
      They build for the medical industry. Unfortunately, it is highly unlikely they will sell you any units fro a DIY project.

      450 liter per minute at 5000 Pascal pressure (50 cm H2O)
      3 Phase brushless DC motor 24V, 1.4A
      Spec Sheet:

    6. Sorry those specs should be maximum 5000 Pascals, maximum 450 liters per minute.
      If you analyse the de-rating chart we get our 100 liters per minute at 4300 pascal (43 inches H2O)
      Probably a Respiratory Therapist would be a better judge of whether this meets the bill than me.

    7. Here is a nice summary of how pressure control ventilation works.

      They suggest a maximum peak airway pressure of 25 cm H2O.
      If you bought the Amazon blower, you could easily create a performance graph similar to what is on the Micronel spec sheet for comparison. Another really important aspect is how quickly the blower can speed up or slow down. This depends somewhat on the design and capability of the electronic driver system and also is dependant on the mass of the blower impeller and motor rotating element.

    8. The Amazon blower can barely produce flow against a few cm H20 but the Micronel part certainly has capability

      Pressure support/pressure control or BiPAP (2 level) ventilation is still not sufficient for patients with stiff lungs who may need 50 cm H20 or more peak pressure.

      I don't see any info on the response time of the Micronel products so you might let it blow continuously with large diameter valves in the patient circuit to block and then dump pressure. Typical of early ventialtors from the 1970's and all electric CPAP products.

    9. For Anyone looking for blowers, here are some more choices:

      Here is a blower from Digikey
      Digi-Key Part Number 603-2239-ND
      Manufacturer Delta Electronics
      Manufacturer Part Number BFN0724SS-01
      It has a Spec sheet with a performance curve
      This is a 5KPa blower that will produce about 50 cm H2O static pressure
      Looks very similar to the Micronel
      US $153
      50 in stock

      Here is an 8 KPa blower from China, US $35
      Unfortunately, it would probably take a long time to get here.
      They have a spec sheet as well.
      They also have a driver for it.

    10. Lot's of technical information here with pictures and parts lists for a successful blower based ventilator from 2006:

    11. Nice find and good information. That is very detailed useful to get an idea of what is required for parts and design for systems using blowers. This is, however, for a BiPAP design. BiPAP shares some of the same design elements as a pressure control ventilator with PEEP capability but is on its own, not a suitable replacement for patients with serious lung injury.

  13. Thanks very much for this comment. It is important to develop an effective device not just any device. The MIT paper said they could add PEEP (positive end-expiratory pressure) to the design. I am not exactly sure how they would do that. I understand that this is required to open keep open the stiffened alveoli in the lungs. Also you would want to increase the oxygen concentration. That could be done with a bellows though. In the published paper the MIT guys said they were adding that feature to their bag type device. But you say this is not possible with a bag or bellows style device. Important to know. So is PEEP used in every case of ARDS? A pressure style type device is much more complex to construct and control.

  14. First year Respiratory Therapists would likely take at least 2 courses on this topic, starting with an Introduction to Mechanical Ventilators. Here is an older reference text that was popular in Canada, in part because the author lives here. Can't beat the price!

    In this text. the mechanisms that are mentioned in this blog are fully described and their relatively poor performance stated. No commercial ICU ventilators were manufactured with bags, bellows or pistons after 1980. That is 40 years ago by my estimation.

    Be very cautious with the article from Mechanical Engineering Group at MIT. Note that this is a student project in the MIT Precision Machine Design course. Nothing more. NOT anything to base a life support product on.

  15. You are correct, Clarence, but a high pressure blower could be substituted for the one used used in a BiPAP device like this, right?

    I suggest everyone read the technical details of this commercial design (BiPAP Synchrony) to see their solutions to the challenges with a blower-based product.


  16. Thats right. BiPap ventilators are generally intended for use with a mask and so have lower pressures than the high pressure units used in critical care directly connected to an endotracheal tube to seal the airway. Other than the higher pressures and more extensive controls and monitoring they are very similar. The positive pressure ventilator will have a more powerful blower as well. That design is well documented in the link and offers a lot of insight into how to make one.

  17. ICU's will generally avoid using BIPAP in these patients since the outlet valve will shoot aerosols across the unit. They will go straight to paralyzing and intubating.

  18. We're creating an engineering group to help in the event of a severe COVID-19 crisis.

    The idea is that a COVID-19 outbreak might produce demands that our medical system hasn't seen before. There may be shortages or special needs of medical equipment and other resources.

    Right now we have engineers, PhDs, and the owner of a devices manufacturing company, with a supply chain and offices in New York, Florida, and Taiwan. We probably can't build ICUs, but we could help a burdened medical front-line in a crisis.

    I'm posting because we are right now looking for doctors and nurses to get advice, particularly those with emergency room, triage experience.

    We're also looking for other help—people with experience in project management, engineering as well as non-technical leadership skills.

    If you're interested, please get in touch. My email is charles (at)

  19. There's an open source project for ventilators here: Please join us. We need your help.

  20. There's an open source project for ventilators here: Please join us! We need your help.

  21. How much for someone to make one for me. Older partner with all sorts of health issues. Terrified they'll get it and need a ventilator.

    1. I appreciate your concern for your partner. There is right now a large group of technologists and medical people working on solutions to the ventilator shortage. One of them may involve groups building them and providing them fro free to people that know how to use them. Liability concerns are a huge problem. Selling them is really not an option.

  22. Anyone know of medical doctors willing to appear on mainstream media to highlight the issue of worldwide ventilator shortage?

    I'm thinking Emerg Docs are ideal because their focus is on treatment of patients who are may not be admitted into the hospital or ICU.

    Someone open to stating that a pandemic ventilator must be proven to be safe and reliable safe to use but has value even with the most basic operating features.

    Yes, I'm talking about a "MEDICAL CHAMPION"

    John Strupat

    1. It is all in how you ask the question. There once was a nun who asked mother superior if it was Ok to smoke while praying. Mother Superior said no, it was not, you must devote all your efforts to praying. Another nun asked mother superior if it was okay to pray while she smoked. Of course said mother superior, you should pray every moment you can. Do not ask it if is OK to drop standards of care in a pandemic. Ask if it is OK to use every effort humanly possible to save the people that we do not have existing ventilator capacity for.

    2. Hello Mr. Strupat,

      We have a great shortage of ventilators in my Eastern European country - we have maybe 80 in total across the whole country, but almost certainly less. But we have some great engineers and some manufacturing capacity. I came across the CBC article on your version of the pandemic ventilator. Something like this, if we were to produce it here, could save hundreds or thousands of lives. Is there any way you could share the specs? Can I contact you directly somehow?

      Thank you and good health to you and yours

  23. A pandemoc machine has many patientcapacity &nd is simple.

  24. Out grandchildren's bouncy castle blower might have useful specs. I don't have much tech detail but it can blow up (deliver) perhaps 4m3 per minute to a fairly high static pressure probably way beyond what needed for a dozen ventilators.

  25. Thanks for sharing such an informative post. To Buy Bipap machine online , Medical Equipment Online at Best Prices from SSB Medicare Pvt Ltd in India. Top Quality with best price Health Equipments at your Doorstep. Order Now!
    Phone: +91 8400-600-800


Commenting has been limited for this post

Note: Only a member of this blog may post a comment.