Homemade protective mask project

Due to the COVIDpocalypse, there's suddenly a worldwide (and nationwide, and statewide) need for surgical masks and N95 masks. Current supply is not enough for the wave of demand that is rushing through US hospitals as of 2020-03.

Meanwhile, the rest of us need them, too. We should learn from the countries that experienced SARS earlier this century, and are generally faring better than most other countries experiencing the SARS-CoV-2 pandemic now. We should also learn from studies that have been done on various aspects of this problem, rather than just taking something off the shelf and hoping it works.

Commercial PPE

Personal protective equipment (PPE) for hospitals is regulated, and the stuff health care workers (HCWs) are supposed to use is regulated and standardized/specified.

With regard to protective masks, there are three basic kinds:

• Surgical mask: Loose-fitting, tied round the back of the head.
  • These masks protect the wearer and the patient from each other's aerosolized droplets. For example, one protocol for treating COVID-19 patients is a nasal cannula covered by a surgical mask.
• N95 respirator: Tight-fitting, roughly dome-shaped singular piece of filter paper/fabric specified to filter (block) 95% of particles 0.3 microns in size. These masks have a metal nose-piece that the wearer should form to their nose, and two elastic bands that are meant to go over the crown of the head and the base of the neck, in order to form a seal around the edge of the mask. Accordingly, there are also standards for keeping facial hair shaved where the edge of the mask makes contact, so the diameter of one's facial hairs does not become the width of a gap.
  • These masks protect the wearer from small particles, such as smoke particles (hence the demand during wildfires) and virions (virus particles). Note that respiratory droplets are considerably bigger than 0.3 microns; an N95 mask would be necessary for airborne transmission, and the WHO recommends “airborne precautions” for certain procedures.
• Cartridge respirator: Often labeled as “P100” (a designation similar to N95 but different in ways that aren't relevant here). Here, the respirator is a hard plastic reusable enclosure, and the filters are cartridges that one changes out from time to time (e.g., when visibly dirty or when breathing is difficult). These are used more in construction than health care.

Surgical masks and N95 masks are normally treated as consumables: wear for some limited time, then dispose of as a biohazard. But given shortages, we're quickly going to see HCWs reusing the supplies they've got, all while getting priority on all supply being produced—so don't expect to be able to buy these at retail for awhile.

Requirements

Safety

“First, do no harm” is the golden rule whenever doing anything medical, including making (essentially) medical equipment.

The biggest hazard with masks is hypoxia. If the material does not permit gas exchange, it will trap dead air (CO2) and not admit fresh air, and the wearer will gradually become more confused as they breathe in their own exhaust and blood oxygen saturation (SpO2) drops.

Effectiveness

The material must permit gas exchange, but not be too permissive; it must also function as a filter. Nor should it be so occlusive that exhalation pushes the mask out and creates an opening for unfiltered air to enter.

These two properties are called filtration efficiency (percentage of contaminants blocked) and pressure drop (difference in fluid/air pressure between the outside and the respirating wearer's side).

It must be said that perfect is the enemy of the good, here. We're necessarily not working to the standards of commercial PPE; that would require specialized material (non-waven polypropylene fabric) and equipment (heat-sealing rather than sewing) that we don't have. It's OK for a homemade mask to be not as good as real PPE; it just has to be better than nothing, and not a false sense of security or actively harmful.

The fact that we're mainly concerned with droplets means N95 is not a goal. We need to catch and block droplets; virus particles should be blocked in the process. We're also not trying to pass an N95 fit test; the design will mimic that of a surgical mask, not a respirator.

(There is also the option of wearing a reusable mask over a commercial N95 mask, to prolong the life of the latter.)

Ability to obtain materials

Even on a good day, non-woven polypropylene is not something you find at your average fabric store. The sort of materials we're most likely to have on hand include:

• quilting cotton
• denim
• canvas (e.g., duck canvas)
• floral wire
• string or twine
• ribbon
• elastics
• cotton or polyester thread
• sewing machines

We're hampered by the shelter-in-place orders: most fabric stores are probably not open. Delivery may be an option, and Cliff's Variety in the Castro (which is also a hardware store, and thereby “essential” under at least the City's order) is still open. So we're limited to whatever supplies we've got on hand.

This requirement also means using materials rather than off-the-shelf finished parts such as commercial filter cartridges. You can buy cotton fabric from lots of places (even if we can't), but the WidgetDyne 9003A (not to be confused with the 9003B, which is incompatible) could become scarce quickly.

Not subtracting supply from commercial masks

Supposing we did buy a roll of non-woven polypropylene, that would be a roll that wouldn't go to making certified masks. Masks we make for the benefit of ourselves and others shouldn't come at the expense of someone else's.

This also applies to things that are spare parts for medical equipment, such as air filters for CPAP machines. The CPAP machines already need those.

What we know about making masks

TODO: Gather links to studies and summaries of their findings

Possible filter media

• Scraps from N95 mask or surgical mask: Violates the rule of not subtracting supply from commercial masks. If we have N95s, we should simply donate those to mutual aid efforts to collect N95s for HCWs.
• Coffee filter: May not have enough filtration efficiency; coffee filters are pretty porous.
• HEPA filter: Tends to be more of a cartridge form factor, but may be fine with a sufficiently thin filter (or unbothered wearer). Often some variety of these is used in hard-enclosure designs.
• Scraps from a vacuum cleaner bag: Interesting prospect. Vacuum bags likewise are made to pass air but block particles (normally dust).
• Paper towel: The HK Mask site says this is pretty occlusive (note the pressure drop for “kitchen paper”).
• Facial tissue: The HK Mask site says one layer is pretty good at pressure, but bad at filtering 0.3µm particles. OTOH, a facial tissue may work way better at catching droplets…
• Plain old fabric: Most designs take this option, leaving it to one or more layers of fabric to do the work of filtration.
  • Denim or canvas: Tighter weaves mean more occlusion and filtration. These would need to be tested to ensure they're suitable and better than nothing.

Known patterns

• Izumi's non-pleated pattern
• Izumi's pleated pattern for adults, and pleated pattern for kids
• “A.B. Mask — for a nurse by a nurse” (includes a suggestion to use this mask as a cover for a commercial N95)
• FreeSewing.org “Fu” pattern
• “HK mask” with a filter pocket (complete with a table of filtration efficiency and pressure drop numbers for various filter media)
• Video tutorial by Deaconess Health (hospital in Evansville, IN)
• Video tutorial for a PM2.5 filter mask (constructed to enclose a disposable PM2.5 filter paper)