Cheap, portable, easy-to-use ventilators for respiratory support in pandemics

I think it is worth exploring inventions for cheap, portable, easy-to-use ventilators to address coronavirus, or to prepare for the next respiratory pandemic. OneBreath, a company out of Stanford, apparently [has already invented this](http://onebreathventilators.com/), although I’m not sure if this company is still active (I just emailed them to check). Although I have not fact-checked much of the below, my hunch is that this avenue is promising.

From https://www.popsci.com/diy/article/2010-05/invention-awards-breathing-easy/: “[In 2006], when Matthew Callaghan was a surgery intern at the University of California at San Francisco, the medical world was buzzing over the prospect of a global flu pandemic. One of the biggest potential problems was logistical: Because 95 percent of the ventilators in the U.S.—which keep critically ill patients breathing when their respiratory system is unable to function—are already in use, thousands of patients would die for lack of available life support. Ventilators cost hospitals from $3,000 up to $40,000 for state-of-the-art models, making it impractical for most hospitals and clinics to stockpile them for emergencies.

Callaghan, [in 2010] a postdoctoral fellow in Stanford University’s biodesign program, knew that in a pandemic situation, hospitals would have to come up with triage procedures that would leave some to die. If he could develop a reliable, no-frills ventilator, it would eliminate many such heart-wrenching decisions. “I thought, these ventilators cost 40 grand, and they just push air around. It isn’t complicated engineering. You don’t need all the bells and whistles.” That thought was the impetus behind the OneBreath, the ventilator Callaghan invented with a small team of fellow Stanford biodesign students. The device is just a fraction of the cost of a low-end conventional ventilator, runs on a 12-volt battery for six to 12 hours at a time, and is smaller than a toolbox so it can be easily deployed wherever needed.”

Ten years after that article, I’ve [read](https://medium.com/@joschabach/flattening-the-curve-is-a-deadly-delusion-eea324fe9727) and confirmed with friends in Boston hospitals that ventilator shortage remains a concern in today’s coronavirus pandemic. “If we take the number of ventilators [existing in hospitals and a CDC stockpile] as a proximate limit on the medical resources, it means we can take care of up to 170,000 critically ill patients at the same time.” Compare that to a loose estimate of the number of ventilators needed, whose numbers I have *not* checked (and note that this estimate is based on an overstatement of Marc Lipsitch’s true estimate of COVID prevalence, which I last read was 20-60% of American adults, in the absence of mitigation): “assume that 55% of Americans catch COVID-19 until the end of 2020, and 6% (10.8 million) of them will need ventilators at some point [with each intense case needing a ventilator for 4 weeks]” (https://medium.com/@joschabach/flattening-the-curve-is-a-deadly-delusion-eea324fe9727). Knowing that that estimate needs to be taken with a fistful of salt, it still seems quite plausible that there will be a ventilator shortage. So what happened to OneBreath?

“A round of successful tests on pigs wrapped up [in December 2009], and the FDA is expected to review the device for humans [in fall 2010]. The OneBreath should not need to undergo clinical trials, Callaghan says, since it performs the same air-moving function as existing ventilators. He anticipates that the U.S. government will want to stockpile the device for use during pandemics, but clinicians who have been privy to the OneBreath’s development are excited about its prospects elsewhere as well.”

What happened to it after that? I would be curious to find out.

For more on how OneBreath works (in case you were interested in designing your own), see the full article: https://www.popsci.com/diy/article/2010-05/invention-awards-breathing-easy/.

The US public health system’s preparedness in the early 1990s

I’ve been reading about how the US public health system’s ability to diagnose for coronavirus is slower than that in other countries. This reminded me of results of an audit of the US public health system in the early 1990s (from Laurie Garrett’s _The Coming Plague_), which found there was much room for improvement in the US public health system. Perhaps this is a guide for what could be done better in the future, although I’m not sure if similar room for improvement exists today, or if infectious disease writer Laurie Garrett’s diagnosis of the barriers to improvement at the time were correct (“two decades of government belt tightening, coupled with decreased local and state revenues due to both taxation reductions and severe recessions, had rendered most local and regional disease reporting systems horribly deficient, often completely unreliable”), or if that diagnosis is also true today. However, this is an interesting historical example:

“In response to [an early 1990s] Institute of Medicine’s report on emerging diseases, the CDC gave Dr. Ruth Berkelman the task of formulating plans for surveillance and rapid response to emerging diseases. For a year and a half Berkelman coordinated an exhaustive effort, identifying weaknesses in CDC systems and outlining a new, improved system of disease surveillance and response.
Berkelman and her collaborators discovered a long list of serious weaknesses and flaws in the CDC’s domestic surveillance system and determined that international monitoring was so haphazard as to be nonexistent. For example, the CDC for the first time in 1990 attempted to keep track of domestic disease outbreaks using a computerized reporting system linking the federal agency to four state health departments. Over a six-month period 233 communicable disease outbreaks were reported. The project revealed two disturbing findings: no federal or state agency routinely kept track of disease outbreaks of any kind, and once the pilot project was underway the ability of the target states to survey such events varied radically. Vermont, for example, reported outbreaks at a rate of 14.1 per one million residents versus Mississippi’s rate of 0.8 per million.27
Minnesota state epidemiologist Dr. Michael Osterholm assisted the CDC’s efforts by surveying the policies and scientific capabilities of all fifty state health departments. He discovered that the tremendous variations in outbreak and disease reports reflected not differences in the actual incidence of such occurrences in the respective states, but enormous discrepancies in the policies and capabilities of the health departments.28 In the United States all disease surveillance began at the local level, working its way upward through state capitals and, eventually, to CDC headquarters in Atlanta. If any link in the municipal-to-federal chain was weak, the entire system was compromised. At the least, local weaknesses could lead to a skewed misperception of where problems lay: states with strong reporting networks would appear to be more disease-ridden than those that simply didn’t monitor or report any outbreaks. At the extreme, however, the situation could be dangerous, as genuine outbreaks, even deaths, were overlooked.
What Osterholm and Berkelman discovered was that nearly two decades of government belt tightening, coupled with decreased local and state revenues due to both taxation reductions and severe recessions, had rendered most local and regional disease reporting systems horribly deficient, often completely unreliable. Deaths were going unnoticed. Contagious outbreaks were ignored. Few states really knew what was transpiring in their respective microbial worlds.
“A survey of public health agencies conducted in all states in 1993 documented that only skeletal staff exists in many state and local health departments to conduct surveillance for most infectious diseases,” the research team concluded. The situation was so bad that even diseases which physicians and hospitals were required by law to report to their state agencies, and the states were, in turn, legally obligated to report to CDC, were going unrecorded. AIDS surveillance, which by 1990 was the best-funded and most assiduously followed of all CDC disease programs, was at any given time underreported by a minimum of 20 percent. That being the case, officials could only guess about the real incidences in the fifty states of such ailments as penicillin-resistant gonorrhea, vancomycin-resistant enterococcus, E. coli 0157 food poisoning, multiply drug-resistant tuberculosis, or Lyme disease. As more disease crises cropped up, such as various antibiotic-resistant bacterial diseases, or new types of epidemic hepatitis, the beleaguered state and local health agencies loudly protested CDC proposals to expand the mandatory disease reporting list—they just couldn’t keep up.
Osterholm closely surveyed twenty-three state health department laboratories and found that all but one had had a hiring freeze in place since 1992 or earlier. Nearly half of the state labs had begun contracting their work out to private companies, and lacked government personnel to monitor the quality of the work.29 In a dozen states there was no qualified scientist on staff to monitor food safety, despite the enormous surge in E. coli and Salmonella outbreaks that occurred nationwide during the 1980s and early 1990s.
At the international level the situation was even worse. The CDC’s Jim LeDuc, working out of WHO headquarters in Geneva, in 1993 surveyed the thirty-four disease detection laboratories worldwide that were supposed to alert the global medical community to outbreaks of dangerous viral diseases. (There was no similar laboratory network set up to follow bacterial outbreaks or parasitic disease trends.) He discovered shocking insufficiencies in the laboratories’ skills, equipment, and general capabilities. Only half the labs could reliably diagnose yellow fever; the 1993 Kenya epidemic undoubtedly got out of control because of that regional laboratory’s failure to diagnose the cause of the outbreak. For other microbes the labs were even less prepared: 53 percent were unable to diagnose Japanese encephalitis; 56 percent couldn’t properly identify hantaviruses; 59 percent failed to diagnose Rift Valley fever virus; 82 percent missed California encephalitis. For the less common hemorrhagic disease-producing microbes, such as Ebola, Marburg, Lassa, and Machupo, virtually no labs had the necessary biological reagents to even try to conduct diagnostic tests.”

My questions about coronavirus (possible PhD research questions)

  1. Why do infectious pathogens like coronavirus not end up infecting 100% of people? Marc Lipsitch claims this is because enough of the population becomes immune at a certain point, creating “walls” through which the pathogen cannot transmit. Where is the study that shows this is the explanation? Because the other possibility is that the estimates that “20% of the world was infected with 2009 H1N1” were taken too early (the pathogen was still spreading) or had limitations in some other way. Edit 3/12: Per https://medium.com/@tomaspueyo/coronavirus-act-today-or-people-will-die-f4d3d9cd99ca, this has implications: if we lift social distancing measures in x time, will cases just reappear? This is suggested by Spanish flu, and China being able to limit cases now but lifting social distancing soon will test this. :Edit 3/12.
  2. My friend told me hospitals will lack enough ventilators to provide breathing support; these ventilators may also be bulky and expensive. What happened to the portable, cheap ventilators called for in the [Johns Hopkins Center for Health Security report on technologies for global catastrophic biorisks](https://www.dropbox.com/s/5h3r7e9m0zrujgl/181009-gcbr-tech-report.pdf?dl=0)? For example, what happened to the [OneBreath ventilator](https://www.popsci.com/diy/article/2010-05/invention-awards-breathing-easy/)? Maybe that could be deployed now.
  3. According to Marc Lipsitch, on average it takes “three weeks to die from infection” from coronavirus. I assume this number is calculated only from those who have died? (This makes it less relevant to someone who does not know, in advance, whether they will die.) Also, how was the infection date measured, given that coronavirus is often asymptomatic? This is relevant for knowing how urgently one should fly home to see one’s elderly relatives if they start showing symptoms. Edit 3/12: https://github.com/midas-network/COVID-19/tree/master/parameter_estimates/2019_novel_coronavirus#time-from-symptom-onset-to-death. :End edit.
  4. What is the status of the coronavirus diagnostic from the CDC? How about those from other developers, e.g. SHERLOCK out of the Broad? Is SHERLOCK field-ready and better for this coronavirus situation?
  5. Let’s say two friends are considering whether to hang out. If they check that neither of them have symptoms, is it safe to meet up? (This is not to say preemptive social distancing measures like closing down public gatherings are not useful; clearly they are, because even symptomatic people might come to those.)

  6. Elderly people are dying in higher numbers of coronavirus. This also seems to be true of flu and other potentially unrelated problems like air pollution. I never see this in my life because I’m not in the hospital and I don’t know many elderly people. What is the pathology in these diseases, i.e. how does death happen?
  7. Less important: I heard there are studies showing a bad health outcome if exposed to a high dose of virus, but a good outcome if exposed to a low dose. I don’t remember which virus, which organism (animal?), etc. Where are these studies? What is the implication, if any, for coronavirus? (For example, should we think of increased risk from being in crowded places as being explained by lots of virus particle from lots of people, i.e. high dose? This is different from how I usually picture it: a crowded place means more sick people.)
  8. How many tests did South Korea actually do when people say “they did a lot of tests”? Check https://ourworldindata.org/coronavirus.

My actions in response to coronavirus

My friend asked me, “Friends!! Are you worried about coronavirus at all?” Here’s my current response:

**~Although I will be adopting “common-sense” protective habits~, I will practice social distancing and other “common-sense” habits when I don’t lose too much by adopting the habit; this includes remote work. I’m mainly not worried about my own safety because fatality rates seem to be low for my age range. I’ll especially avoid being in contact with the elderly and immunocompromised to reduce risk of transmitting to them.**

The Chinese CDC-reported fatality rate for my age range is [0.2%, i.e. 1 in 500 chance of dying if I get it (maybe one could consider that too high), with this being a potential overestimate if it’s true that early case fatality rate is overstated due to better recording of deaths from coronavirus than coronavirus cases themselves](Chinese CDC: https://www.businessinsider.com/coronavirus-death-age-older-people-higher-risk-2020-2).[1] Edit 3/7: I agree with [Dr. Jeremy Faust](https://www.cnn.com/videos/tv/2020/03/07/how-vulnerable-is-the-average-person-to-the-coronavirus.cnn) that more of the focus should be to help those most vulnerable, i.e. the elderly and immunosuppressed. For a non-elderly/non-immunosuppressed person, this may mean not hoarding supplies, and not becoming a source of transmission (e.g. by following self-protective habits that don’t hoard and by self-quarantining if infected), which are both cases of worrying about the much higher risk to someone else than the much lower risk to oneself. :End edit.

By “common-sense” protective habits, I mean things like the following, where I or things I care about (e.g. environmental values) don’t seem to lose much by adopting the habit, and the main thing keeping me from adoption, especially _consistent_ adoption, is inertia:

  1. Biking to places instead of taking crowded subways (which I mostly already do) (Edit 3/9: however, if my only mode of transport were a subway and I wanted to make that trip, I’d do it without worrying, because biking is mainly for exercise and speed of travel in Boston :End edit),
  2. Avoiding non-essential flight or travel (which I mostly already do, because I dislike being on planes anyway and for environmental reasons),
  3. Doing my PhD work around 1-2 people in a relatively uncrowded office space instead of around 10-20 people in more crowded office space (which for me gets the best of both worlds of keeping each other productive and non-loneliness, plus reduce transmission risk; I mostly already do this), and
  4. Continuing to eat healthy, sleep well and stay low-stress to keep my immune system healthy (which I mostly already do).
  5. [“And self-protective activities include all the hygiene measures that were mentioned earlier [washing hands, not touching one’s face, avoiding public places], includes getting up-to-date on vaccinations to prevent needs for contact with the health care system. It includes, if you smoke, quitting smoking, because you need your lung function… It includes staying home from work to protect others if you’re sick. It includes making that possible if you’re a business owner.”](https://theforum.sph.harvard.edu/events/the-coronavirus-outbreak/).

[“These are all individually self-improving acts that will reduce one’s own risk. And this is a happy case where every one of those things also has benefits to the community. Slowing the epidemic is what we have to do if we can’t stop it. And all of those measures, small though some of them may be, help to slow the epidemic.”](https://theforum.sph.harvard.edu/events/the-coronavirus-outbreak/)

Footnotes:

  1. I understand that fatality rates are highly uncertain given that they’re calculated as “num deaths attributed to coronavirus” / “num cases of coronavirus,” [both of which are underestimated numbers right now due to lack of testing](https://theforum.sph.harvard.edu/events/the-coronavirus-outbreak/). However, I feel the low fatality rate for my age groups, even with error bars, makes me not worry about my own safety.