Thread by @LazarusLong13

What is the right thing to do?

When you are wrong, you apologize - as I did after seeing @bencowling88 and @jburnmurdoch's analysis of the 53% masking study.

You retract if the honest errors are particularly large or impactful.

I propose that Loeb (2009) and Radonovich (2019) need to be retracted as they were measuring the efficacy of N95's versus surgical masks in protection from an airborne disease - but using droplet protocols.

Not airborne protocols.

First - we can see that

influenza is airborne via aerosols:

Cowling et al (2013)
ncbi.nlm.nih.gov/pmc/articles/P…
Tellier (2006)
ncbi.nlm.nih.gov/pmc/articles/P…
High-level review of Milton et al (2018)
cidrap.umn.edu/news-perspecti…
Study:
pnas.org/content/115/5/…

While hospitals tend to have excellent ventilation, it is not enough to prevent infections. To wit:
khn.org/news/article/h…

In the below study, a CoVid ward with 12 ACH (double the standard excellent ventilation rate of 6 ACH), "(w)hen the door to the patient room was open,

high levels of aerosols had crossed the corridor and entered the nurses’ station at baseline measurement."
ncbi.nlm.nih.gov/pmc/articles/P…

So, flu is airborne via aerosols, and even excellent ventilation in hospitals is not enough to prevent aerosols from flowing into corridors.

H/t @CPita3 for catching these gaps in the these studies.

Loeb et al (2009)
jamanetwork.com/journals/jama/…

"Objective: To compare the surgical mask with the N95 respirator in protecting health care workers against influenza."

They acknowledge that influenza is spread via aerosols.

The key parts are in the Interventions section.

Cntrl + F "1 m of a patient" to jump to it.

The nurses wore gloves and gowns when entering the room of a patient. Per the instructions of the Ministry of Heath and Long-Term Care, they would put on a surgical mask when within 1 meter of a patient.

Those are droplet protocols. Not airborne protocols.

Someone asked - "What if the nurses put on the mask prior to entering the room?"

It doesn't matter. As shown above, aerosols flow into corridors.

There were Audits - the auditor never entered the patient room.

We might assume that the nurse put on the PPE in the corridor

in front of the auditor.

But that is an assumption - and we have no idea what happened in room itself as the auditors never went into the room, as shown above.

A clue of what went on - can be found from the timing of the study.

Study period was between January 12, 2009 and April 23, 2009.

The audits were from March 11 to April 3, 2009.

Study was ended on April 23, 2009 - due to the H1N1 Pandemic, and requirement of the usage of N95's. (pic).

So, what were the default protocols for influenza, then?

The June 19, 2009 / 58(23);641-645, MMWR indicates that standard droplet protocols were in place.

Therefore, except for the 18 events from the 23 days when an audit was conducted, we have 58 days where the default protocol was in place.

3 feet away.

cdc.gov/mmwr/preview/m…

And keep in mind - all events were handled at the very best with the donning of PPE done right outside the door.

If you enter an asbestos abatement site with asbestos swirling around, does it matter if you put on your N95 when you've already inhaled it? H/t @sameo416 for this.

Back to the study, it's no wonder that surgical masks were found to be noninferior to N95's.

And the results were found to suggest that small aerosols do not dominate transmission.

This cannot be understated - the entire IPC field's understanding of the usage of N95's

is based in part on this study.

Let me go back to where this thread started.

Will an editor's note suffice to undo the damage of these honest errors?

Is that the right thing to do?

While you think upon that, let's move onto Radonovich et al (2019).

jamanetwork.com/journals/jama/…

If you don't want to read any more on this study, you can skip down to the tweet with this picture, again.

cdn.jamanetwork.com/ama/content_pu…

In essence, the picture of this diagram contained all that we really need to know - 6 feet is not enough for aerosols.

But let's dive in.

The question it is seeking to answer

"Is the use of N95 respirators or medical masks more effective in preventing influenza infection among outpatient health care personnel in close contact with patients with suspected respiratory illness?

They found that, counter-intuitively,

N95's performed worse. 8.2% of HCW's ended up with influenza, as opposed to those with surgical masks at 7.2%

Not a significant difference, but certainly raises questions given that N95's with a good seal have 95% Inward Protective Efficiency per NIOSH.

Surgical masks?

Not even rated by NIOSH for airborne protection.

Of interest to all, should be the fact that this study cited Loeb (2009) as inconclusive evidence that N95's have been clinically effective.

In the "Procedures, Interventions, and Group Allocation" section we find the key part:

"Participants were instructed to put on a new device whenever they were positioned within 6 feet (1.83 m) of patients with suspected or confirmed respiratory illness."

Any audits? Yes, randomly - and did not enter the patients' room.

Reviewing the supplementary material, we find:
cdn.jamanetwork.com/ama/content_pu…

that the protocol was at 6 feet - or entering a small room.

And in the supplemental eAppendix, we find:

"Study monitor observed device wearing before entering and after exiting patient rooms was 40.6 % in
the N95 arm and 33.5 %in the MM arms (p = 0.02), suggesting adherence was larger in the N95 arm.
"
cdn.jamanetwork.com/ama/content_pu…

But - those enforcing this adherence were told within 6 feet - of a suspected or confirmed respiratory infection.

cdn.jamanetwork.com/ama/content_pu…

Influenza has a 2 day presymptomatic period, so this is another issue. But let's put that aside and continue on our 6 foot long journey.

The pre Study Activation and Study Activation Letters - say nothing about doors.

Does say "within 6 feet"

Daily Exposure Form - does not ask about doors.

Does ask about 6 feet. (pics below are zoomed in, and then the relevant part of the Exposure form).

I am still adding to this thread.

I recommend not replying quite yet - I will say "/fin" upon completion.

The Patient-Based Observation Form - merely asks if the HCW wore a mask during the visit.

Not if they wore it through the door.

So, to conclude with Radonovich - it was a study designed to test PPE in a clinical setting - versus an airborne disease.

That found those in N95's got sick more than those in surgical masks. Not significantly scientifically, but significantly enough to be a red flag.

What is the right thing to do?

The WHO IPC Guidelines for CoVid recommend continuous medical (surgical) mask usage.

who.int/publications/i…

Which is based on:
apps.who.int/iris/bitstream…

In that document, a Cochrane review,

located here:
ncbi.nlm.nih.gov/pmc/articles/P…

In which the key message was "We are uncertain whether wearing masks or N95/P2 respirators helps to slow the spread of respiratory viruses."

As based on 5 studies. Two of which were Loeb and Radonovich.

Without those 2 studies?

N95's are statistically superior to surgical masks in a clinical setting.

Nothing short of retraction of Loeb and Radonovich will do to right the mistakes that have been made to the world's Health Care Workers - and to the IPC field, itself.

Those studies have been, and will

continue to be cited by those seeking to make their mistaken cases that surgical masks are equivalent in the clinical setting.

As we have seen during this pandemic, an editor's note will not suffice. Those are generally missed in the speed of today's internet-connected world.

Loeb and Radonovich must be retracted.

/Fin