A new qualitative fit test was developed using Bitrex (Macfarland Smith Limited) as the test agent. It was validated by running a series of paired qualitative and quantitative fit tests. Quantitative tests were conducted with a small corn oil aerosol, using a condensation nucleus counter as a detector. Qualitative fit tests were run with Bitrex and saccharin, following the established protocol for the saccharin fit test. Four models of National Institute for Occupational Safety and Health--approved replaceable filter respirators were used in the study. All were half mask models equipped with high efficiency filters. In some cases, respirators expected to be the correct size for test subjects were tested. In other cases, respirators expected to be too small or too large for the subjects were tested. Test results were analyzed using fit test method validation criteria recommended in the American National Standards Institute draft standard on fit testing (ANSIZ88.10). The Bitrex and saccharin tests were found to have virtually identical performance. Both met proposed American National Standards Institute requirements for a valid qualitative fit test.

American National Standards Institute (ANSI) standard Z88.10, Respirator Fit Testing Methods, includes criteria to evaluate new fit-tests. The standard allows generated aerosol, particle counting, or controlled negative pressure quantitative fit-tests to be used as the reference method to determine acceptability of a new test. This study examined (1) comparability of three Occupational Safety and Health Administration-accepted fit-test methods, all of which were validated using generated aerosol as the reference method; and (2) the effect of the reference method on the apparent performance of a fit-test method under evaluation. Sequential fit-tests were performed using the controlled negative pressure and particle counting quantitative fit-tests and the bitter aerosol qualitative fit-test. Of 75 fit-tests conducted with each method, the controlled negative pressure method identified 24 failures; bitter aerosol identified 22 failures; and the particle counting method identified 15 failures. The sensitivity of each method, that is, agreement with the reference method in identifying unacceptable fits, was calculated using each of the other two methods as the reference. None of the test methods met the ANSI sensitivity criterion of 0.95 or greater when compared with either of the other two methods. These results demonstrate that (1) the apparent performance of any fit-test depends on the reference method used, and (2) the fit-tests evaluated use different criteria to identify inadequately fitting respirators. Although "acceptable fit" cannot be defined in absolute terms at this time, the ability of existing fit-test methods to reject poor fits can be inferred from workplace protection factor studies.

No abstract

A workplace protection factor (WPF) study was conducted with a half-mask air-purifying respirator during fiber glass boat production. Styrene was the measured analyte, and the geometric mean WPF found was 39.7. Analytical detection limits, sample contamination, and pulmonary elimination from previous exposures or from skin absorption were identified as important considerations that can bias the WPFs measured. There were significant differences in the mean concentrations found inside the respirator when analyzed by time period. An increase in the concentration found inside the facepiece cavity and a decrease in the WPF over time was found for people with three or four measurements. This indicates either a change in performance of the respirator over time or a bias from low-level exposures during the day or skin absorption.

A method for performing quantitative fit tests (QNFT) with N95 filtering facepiece respirators was developed by earlier investigators. The method employs a simple clamping device to allow the penetration of submicron aerosols through N95 filter media to be measured. The measured value is subtracted from total penetration, with the assumption that the remaining penetration represents faceseal leakage. The developers have used the clamp to assess respirator performance. This study evaluated the clamp's ability to measure filter penetration and determine fit factors. In Phase 1, subjects were quantitatively fit-tested with elastomeric half-facepiece respirators using both generated and ambient aerosols. QNFT were done with each aerosol with both P100 and N95 filters without disturbing the facepiece. In Phase 2 of the study elastomeric half facepieces were sealed to subjects' faces to eliminate faceseal leakage. Ambient aerosol QNFT were performed with P100 and N95 filters without disturbing the facepiece. In both phases the clamp was used to measure N95 filter penetration, which was then subtracted from total penetration for the N95 QNFT. It was hypothesized that N95 fit factors corrected for filter penetration would equal the P100 fit factors. Mean corrected N95 fit factors were significantly different from the P100 fit factors in each phase of the study. In addition, there was essentially no correlation between corrected N95 fit factors and P100 fit factors. It was concluded that the clamp method should not be used to fit-test N95 filtering facepieces or otherwise assess respirator performance.