Effect of leak location on measured respirator fit.

A significant difference in leak detection as a function of leak location was observed during a study assessing how well current models of quantitative fit-test systems detect leakage. Known sources of leakage (matched hypodermic needles) were introduced at three fixed locations into factory-probed half-mask and full-face respirators mounted on a headform-breathing machine system. The leak locations were the bridge of the nose, the cheek, and the chin. Baseline leakage into each respirator was determined by conducting a fit-test with all fixed leak sources capped. Fit tests were repeated with each individual source uncapped. Study objectives included determining (1) how well each system measured the leakage, and (2) whether leak location had any effect on leak measurement. An ambient aerosol fit-test system (Portacount Plus) and a controlled negative pressure (CNP) fit-test system (FitTester 3000) were used. The ambient aerosol system detected an overall average of 37.2% of the known leakage, with a coefficient of variation of 44.7%. An analysis of variance showed significant differences in aerosol system measurements of leakage as a function of leak location and mask type (p < 0.001). A different pattern of aerosol leak detection as a function of leak location was observed between half-mask and full-face respirators, which appears to be related to differences in in-mask airflow dynamics. The CNP system detected an overall average of 97.9% of the known leakage through the same hypodermic needles, with a coefficient of variation of 4.3%. CNP system results were not affected by leak location (p > 0.43) or mask type (p > 0.32).