NIH Releases New Propaganda on School Mask Policy
The latest attempt by public health officials to defend the indefensible fails massively
Given the significant waning of Covid-19 cases and deaths, and the rising acknowledgement that the cloth & surgical masks in widespread use are ineffective in preventing the spread Covid, along with the increasing media coverage of the harms resulting from authoritarian school covid policies, I was hoping that my recent Pure Science Substack article, Stringent School COVID-19 Policies are Child Abuse would be my last article (for a while at least) involving Covid-19, and in particular the issue of mandatory masking of school children. However, on 3/10/2022 the National Institute of Health (NIH) published a News Release inferring that mandatory mask school policies resulted in a 72% reduction of secondary (schooled related) Covid-19 infections; the methods, scientific rigor, and inferences of the so-called “study” on which it was based are so egregiously poor that I could not let it stand without comment.
As I've noted in previous articles, the public’s confidence in science is justifiably low and continues to be damaged further when science is corrupted in the name of achieving a purported social benefit, or in this case as an ex post facto defense of the indefensible stringent covid school mask policies enacted in much of the United States since the pandemic began.
The new study compares rates of primary and secondary Covid-19 infection in school students and school staff in schools with mandatory mask, partial mask mandates, and optional mask policies. The authors attempt to estimate the effectiveness of mandatory mask mandates by comparing the rates of secondary infections to primary infections in schools with differing mask policies. Figure 2 below shows a plot of secondary versus primary cases for the 61 districts included in the study. The higher slope of the red line relative to the green and yellow lines graphically represents the author’s claim of much higher Covid-9 transmission rates in mask optional schools.
Let's go through some of the flaws in this study one by one.
Flaw #1: Not Randomized or Controlled
As we've seen in discussions of other potential Covid-19 mitigations such as Ivermectin (How Many Covid Deaths Would Ivermectin Have Saved?), the highest quality studies are Randomized Control Trials (RCTs). Retrospective studies like this one are considered more suspect, especially when the authors make no attempt to match populations in the test and control groups (no attempt was made to match groups in this new study).
Flaw #2: Inconsistent Methodology
Despite vague references by the authors to “contact tracing”, the study leaves the determination of whether or not a Covid-19 case was primary or secondary (i.e. from school) to each of the individual 61 districts; in other words there could be 61 standards for deciding if a case is primary or secondary. The authors freely admit that many districts don't have the resources at all times to adequately trace the source of each case.
The fact that it is impossible to accurately trace the source of a virus like Covid-19 with a high frequency of asymptomatic cases and a very long latency between exposure and infection is a fundamental flaw in the methodology of the paper.
Additionally, it is not clear why the authors chose to group students together with school staff.
It is likely there were disparities between these two subgroups and the additional effort to segregate data for these two groups is very minimal. Barring an explanation by the authors, one is left to question their motives. Possibly student infections were much lower than the combined values reported by the authors?
Flaw #3: Poor Survey Response and Self Selection
The study was based on an invitation sent to 13,800 school districts. The number of school districts responding to the survey and complying with its requirements was only 61, which corresponds to an abysmal 0.0044 response and inclusion rate! Low survey response rates are very susceptible to self-selection bias - those responding to the survey do not represent the larger population and may be biased. Acceptable Survey Response Rate.
It gets worse! Of the 61 school districts included in the study only 6 districts are in the mask optional or trial group. This group has only 4,576 students and staff, thus the NIH claim that over 1.1 million subjects were in this study is more than a little misleading!
Most troubling is the fact that almost half of the secondary cases reported for the mask optional group came from just one school district!
The authors do not report this directly which is disturbing in and of itself, but can be inferred by combining the information from the graph in Figure 2 and the tables in their paper. A single mask optional school district reported approximately 150 primary infections and approximately 37 secondary infections. The conclusions of the authors are highly dependent on this one school district that chose, for whatever reason (possibly unhappiness with that school’s mask optional policy), to be included in the study, and then reported an unusually high ratio of secondary/primary infections.
The omission of this fact by the authors, at a minimum, borders on scientific misconduct.
Flaw #4 - Abuse of Statistics
The authors of the study show a fundamental lack of understanding of statistics and confidence intervals. Look at Figure 2 again. The shaded regions correspond to the 95% confidence intervals of their predictive model.
Notice however that most of the points in green (mandatory masks) fall outside of the green confidence interval! In other words the predictive model of the authors DOES NOT EVEN CORRESPOND WITH THEIR CONTROL (MANDATORY MASK) DATA.
This paper is the epitome of Junk Science. Every author on the paper (see below) should be shamed often and repeatedly until it is withdrawn.
Paper Authors:
This study picked districts that favored the preconceived outcome. It would have been equally feasible to pick districts that supported an outcome of masks didn't work. Of course this report doesn't give method or procedures only conclusions, not helpful at all.