The National Institute for Occupational Safety and Health (NIOSH) and the National Personal Protective Technology Laboratory (NPPTL) will offer the webinar, "Variability of Respirator Fit Test Panels," on Tuesday, July 23, 2013 from 1 p.m. to 3 p.m. ET. The purpose of this interactive webinar is to provide a description and results of a recent NIOSH study to assess the fluctuation between bivariate panels used in respirator fit testing.
Employers rely on NIOSH-approved respirators to protect employees from airborne toxic contaminants. As the last line of protection for workers, respirators must be designed and manufactured to perform reliably. They must also be tested for compliance to a baseline performance level. The performance of the facepiece-to-face seal and other potential sources of leakage are critical for half-facepiece air-purifying respirators since the potential leak areas determine the amount of contaminated air the worker might inhale. In the early 1970s, the Los Alamos Scientific Laboratory developed respirator fit test panels using detailed anthropometric data to evaluate this leakage. Since that time, these panels have been and continue to be used in respirator research, design, and certification. However, based on a 2003 survey of approximately 4,000 respirator users, the NIOSH developed new respirator fit test panels (i.e., bivariate and principle component analysis in 2007. It is important that panels comprised of different subjects provide comparable results when the same half-facepiece respirator is tested.
The study that will be discussed consisted of testing five N95 filtering facepiece and five elastomeric respirators with N95 filters on three different 40-member even subject distribution panels and three different 35-member U.S. respirator user population-based panels. Two ambient aerosol inward leakage tests on each respirator were performed on the subjects while performing the following exercises: Normal Breathing, Deep Breathing, Turning Head Side to Side, Moving Head Up and Down, Reciting the Rainbow Passage, Reaching for the Floor and Ceiling, Grimacing, and Normal Breathing.
Because there were three independent panels of each design (i.e., 35-member or 40-member), the variability between panels was determined by computing an estimate of the distribution of the inward leakage percentage. The data from the subjects were combined and bootstrap samples were created to generate multiple new panels to estimate the panel variability.
The bootstrap estimation procedure for panel variability was determined using random numbers to select 500,000 panels of a specified type and size:
1.Equal size cells of n=4
2.Equal size cells of n=3
3.Equal size cells of n=2
4.Percentage of population-sized cells of n total = 35 (population-weighted bivariate panel)
5.Percentage of population-sized cells of n total = 30 (population-weighted bivariate panel)
6.Percentage of population-sized cells of n total = 20 (population-weighted bivariate panel)
The results of these procedures will be discussed. The webinar will be interactive, with participants having the opportunity to ask questions and provide comments on the test panels.
Webinar presenters are:
- Christopher Coffey, PhD, associate director for science at NIOSH's National Personal Protective Technology Laboratory. From 1993 to 2001, Coffey was involved in the area of respiratory protection, both certification and research. His respirator research was on fit-test methods. He was the first to demonstrate that the results of all fit-test methods did not correlate with a wearer’s actual exposure. His work was also the first to provide a scientifically based explanation of the discrepancies between fit factors and workplace protection factors. He was also the first to characterize the errors (i.e., alpha and beta) associated with the most commonly used fit-test methods. His work also involved determining the fitting characteristics of various types of air-purifying respirators.
- James Wassell, PhD, a statistician in the NIOSH Division of Safety Research. Wassell has 23 years of experience as a research mathematical statistician at the CDC, including writing award-winning research publications in occupational epidemiology and statistical methods. For this project, he developed the experimental design and will perform the data analysis.
This webinar will be available via Live Meeting. Registration by July 19 is required and is limited to 100 seats.