The National Institute for Occupational Safety and Health’s methods and requirements for air-purifying respirator breathing resistance in 42 CFR Part 84 do not include work of breathing. The International Organization for Standardization Technical Committee 94, Subcommittee 15 utilized work of breathing to evaluate airflow resistance for all classes of respiratory protective devices as part of their development of performance standards regarding respiratory protective devices. The objectives of this study were: (1) to evaluate the relationship between the International Organization for Standardization’s work of breathing measurements and the National Institute for Occupational Safety and Health’s breathing resistance test results; (2) to provide scientific bases for standard development organizations to decide if work of breathing should be adopted; and (3) to establish regression equations for manufacturers and test laboratories to estimate work of breathing measurements using breathing resistance data. A total of 43 respirators were tested for work of breathing at minute ventilation rates of 10, 35, 65, 105, and 135 liters per minute. Breathing resistance obtained at a constant flow rate of 85 liters per minute per National Institute of Occupational Safety and Health protocol was correlated to each of the parameters (total work of breathing, inhalation, and exhalation) obtained from the work of breathing tests. The ratio of work of breathing exhalation to work of breathing inhalation for all air-purifying respirators is similar to the ratio of exhalation to inhalation resistance when tested individually. The ratios were about 0.8 for filtering facepiece respirators, 0.5 for half-masks, and 0.25 for full-facepiece respirators. The National Institute for Occupational Safety and Health’s breathing resistance is close to work of breathing’s minute ventilation of 35 liters per minute, which represents the common walking/working pace in most workplaces. The work of breathing and the National Institute of Occupational Safety and Health’s breathing resistance were found to be strongly and positively correlated (r values of 0.7–0.9) at each work rate for inhalation and exhalation. In addition, linear and multiple regression models (R-squared values of 0.5–0.8) were also established to estimate work of breathing using breathing resistance. Work of breathing was correlated higher to breathing resistance for full-facepiece and half-mask elastomeric respirators than filtering facepiece respirators for inhalation. For exhalation, filtering facepiece respirators were correlated much better than full-facepiece and half-mask elastomeric respirators. Therefore, the National Institute for Occupational Safety and Health’s breathing resistance may reasonably be used to predict work of breathing for air-purifying respirators. The results could also be used by manufacturers for product development and evaluation.

National Institute for Occupational Safety and Health (NIOSH) approval testing of air-purifying respirators (APRs) is performed using constant airflow methods, which uses two separate tests to determine the inhalation and exhalation resistances independently. These tests use continuous airflow at a steady rate of 85 l/min to determine airflow resistance. The standard test procedures (STPs) currently used for testing respirators with this method ensure these respirators meet the minimum inhalation and exhalation resistance requirements set forth in 42 CFR Part 84.

In contrast to constant flow, the work of breathing (WOB) method uses a dynamic airflow and continuously measures the inhalation and exhalation pressures. By using dynamic measurements over the current constant airflow methods, the full breath cycle is measured instead of just a single flow rate, allowing for a more complete understanding of how the respirator functions during use. Dynamic and non-linear flow effects contribute to airflow resistance, which for the latter is significant at high work rates. Such contributions are readily seen in WOB (_{T}, the power per minute ventilation. The first WOB and related parameters have been used to evaluate underwater breathing apparatuses since the 1970s (

The comparison between International Organization for Standardization (ISO) WOB and NIOSH breathing resistance has not been studied due to its recent introduction. Therefore, the main objective of this study was to evaluate the relationship between ISO WOB measurements at each of the specified work rates and NIOSH breathing resistance test results, provide a scientific basis for standards development organizations (SDO) to decide if WOB should be adopted, and establish linear and multiple linear regression equations for manufacturers and test laboratories to estimate WOB measurements using breathing resistance data.

A total of 43 APR was tested.

The WOB test setup was configured based on ISO standard 16900-5 (

In these tests, a medium headform, based on ISO 16900-5 (

ISO 16900:12 outlines eight breathing rates based on tidal volume (L) and breathing frequency (breathing cycles/min) (10, 20, 35, 50, 65, 85, 105, and 135 l/min). These specified breathing rates were used to create sinusoidal waveform files (

WOB methods followed ISO Standards 16900-5 and 16900-12. The test procedure, data treatment, and calculations were based on the ISO test standard 16900-12. The quantities were calculated for the five ISO work rates: resting, dynamic sinusoidal at 10 l/min; W1 flow rate, 35 l/min; W2 flow rate, 65 l/min; W3 flow rate, 105 l/min; and W4 flow rate, 135 l/min. RPD pressure and volume data were recorded for all eight ISO waveforms run in order of increasing minute volume with a 1-min duration for each waveform. RPD were not pre-conditioned prior to testing, and all WOB tests were recorded at ambient conditions. Retesting was performed if the respirator seal was found to be leaking. For more information regarding the WOB methods development see

Pressure-volume (PV) data used in WOB calculations was obtained from recorded results by averaging over ten consecutive breaths at each recorded point throughout the period (_{T}/V_{T}), inspiratory work of breathing (WOB_{in}), and expiratory work of breathing (WOB_{ex}) can be found in Annex B of ISO 16900-12:2016 (

The static flow resistance test consisted of two separate tests: an inhalation resistance test reflecting NIOSH STP TEB-APR-STP-0007 (

Breathing resistance for inhalation and exhalation obtained from NIOSH breathing resistance test method was correlated to each of the parameters obtained from the ISO WOB test (WOB (total, inhalation, and exhalation), peak pressures (inhalation, exhalation), and elastance) for corresponding respirator models.

The degree of correlation between the two methods can be determined by comparing current NIOSH STP results to WOB test results for all varieties of air purifying respirators. _{T} and NIOSH breathing resistance for inhalation and exhalation for all APRs. Each of the APR breathing resistance test results for each individual APR tested in this study met the NIOSH breathing resistance requirements (42 CFR PART 84). From ^{−3} m^{3}/s), respectively.

_{T} with breathing resistance by RPD classes at different work rates. The mean WOB/V_{T} of the full-facepiece respirators was generally much higher than half-mask elastomeric and FFR for inhalation and exhalation at all work rates. The same trend was observed with NIOSH breathing resistance. Each Pair Student’s t-test and All Pairs Tukey-Kramer test were used to compare mean WOB/V_{T} and breathing resistance for FFR, half-mask elastomeric, and full-facepiece respirators at different work rates. The test results showed a significant difference of WOB_{in}/V_{T} and inhalation resistance among the three RPD classes at different workflow rates at resting, W1, W2, W3, and W4. For exhalation, no significant differences of WOB_{ex}/V_{T} and exhalation resistance among these RPDs at low and moderate work rates at resting, W1, and W2 were found; and no significant difference was found for WOB_{ex}/V_{T} and exhalation resistance between half-mask elastomeric and full-facepiece respirators, whereas for FFRs a significant difference was found with half-mask elastomeric and full-facepiece respirator at work rates W3 and W4 (the high work rates).

From _{T} for full-facepiece exceeded the ISO WOB limits for high work rates W3 and W4 (see _{T} met the ISO WOB limit.

_{T} at different work rates _{T} can be predicted well by NIOSH breathing resistance for inhalation, while for exhalation, the best match to WOB/V_{T} from NIOSH breathing resistance is at work rate W1 and W2. Most fibrous filters for particulates exhibit laminar flow characteristics at realistic breathing rates, and therefore should be expected to have a linear dependence between the flow rate and the resistance, with the linear coefficient depending upon the particulars of the media. As seen in _{T} for the different breathing rates can be well matched with linear fits, with the slopes increasing with increasing breathing rate. Similarly, _{T} is proportional to the work rates tested.

The Pearson correlation coefficient r was used as a measure of the strength and the direction of the relationship between the two variables. The coefficient r approaches unity when the two variables are positively and linearly correlated. The P value expresses the significance of the relationship between two variables. _{T} and breathing resistance at different work rates for inhalation and exhalation. Most correlation coefficients between WOB/V_{T} and breathing resistance were above 0.7, indicating that they were strongly correlated except at resting (r = 0.24) for inhalation and (r = 0.29) for exhalation. P-values were low, below 0.0001 except for inhalation at rest with a P-value of 0.127, due to the variation between respirator types. _{T} and breathing resistance at different work rates for inhalation and exhalation by classes of respirators. The average correlation coefficient for full-facepiece, half-mask elastomeric, and FFR are 0.79, 0.79, and 0.61 for inhalation and 0.81, 0.53, and 0.92 for exhalation, respectively. WOB/V_{T} correlations to breathing resistance for full-facepiece and elastomeric half-mask were better than FFR for inhalation; for exhalation, FFR was much better than full-facepiece and half-mask elastomeric.

To learn more about the relationship between the five waveforms and breathing resistance for inhalation and exhalation, multiple regression analysis was used to create a prediction model. The data were fairly well matched using a multiple linear fit (R^{2} value = 0.88 for inhalation, from ^{2} value = 0.79 for exhalation, from

Since airflow through fibrous particulate filtration media is typically low Reynold’s number, in the ideal case the instantaneous resistance is directly proportional to the instantaneous flow rate. The maximum flow rate for Work Rate W4 (ISO 16900-12:2016) is
_{T} is the tidal volume. A maximum inhalation resistance _{85l/minl/min} of 343 Pa (35 mm H_{2}O) at 1.42 10^{−3} m^{3} s^{−1} (85 l/min), the NIOSH limit for non-powered air-purifying particulate respirators, would ideally correspond to a maximum resistance of 1760 Pa at this _{Max}, some-what below the ISO limit of 2000 Pa. If one assumes a linear resistance coefficient,

The work of breathing for one inhalation of one cycle of the ISO breath, _{Max} sin_{T} = 1:375kPa, some-what less than the ISO limit of 1.6 kPa. These calculations depend upon the idealized linearity of the filter resistance, and deviations are likely to lead to higher resistances at higher flow rates. Exhalation resistance for respirators with exhalation valves can be assumed to be strongly nonlinear, and these calculations would not apply. For exhalation with respirators with no valve the analysis is similar.

The above analysis and discussion allow the following conclusions to be drawn.

NIOSH breathing resistance was strongly and positively correlated with WOB. Linear and multiple linear regression models (R-squared values of 0.5–0.9) were established to estimate WOB using breathing resistance.

Manufacturers and test laboratories can estimate WOB by using breathing resistance data for selected respirators.

WOB was better correlated to breathing resistance for full-facepiece and half-mask elastomeric respirators than FFRs for inhalation; for exhalation, FFRs had better correlation to WOB than full-facepiece and half-mask elastomeric respirators; the average WOB for full-facepiece was significantly higher than for half-mask elastomeric respirators and FFRs for both inhalation and exhalation.

The ratio of exhalation WOB to inhalation for all APRs is similar to the ratio of exhalation to inhalation resistance. The ratios were about 0.8 for FFRs, 0.5 for half-masks, and 0.25 for full-facepiece respirators.

Statistical analysis showed that there was a significant difference of WOB_{in}/V_{T} and inhalation resistance among three RPDs at different work rates at resting, W1, W2, W3, and W4. For exhalation, there was no significant difference found for WOB_{ex}/V_{T} and exhalation resistance among three RPDs at low and moderate work rates at resting, W1, and W2; and there was no significant difference found for WOB_{ex}/V_{T} and exhalation resistance between half-mask elastomeric and full-facepiece respirators, whereas FFRs were significantly different with half-mask elastomeric and full-facepiece respirators at work rates W3 and W4 (which are high work rates).

Currently, respirators undergo a specific set of approval testing procedures based on the type of respirator and the desired outcome (e.g., determination of breathing resistance, determination of exhalation valve leakage, and maintenance of positive pressure). All respirators must meet a minimum inhalation and exhalation resistance requirement set forth in 42 CFR Part 84.

This study examines the correlation between the test results from current set of NIOSH test methods and the test results for newly proposed ISO WOB method. How the methods compare may provide a basis on which better decisions can be made by NIOSH or other standard development organizations and suggests NIOSH requirements remain relevant even as international standards are updated. Additionally, the test equipment needed to execute the NIOSH test procedures for inhalation and exhalation resistance are relatively simple and can be constructed using commercially available parts. The ISO WOB test method, by contrast, is more complex because a sinusoidal breathing machine is needed as well as the headforms described in ISO 16900-5:2016.

Manufacturers and test houses without the capability to run WOB tests can measure breathing resistance and use our regression equations to estimate WOB values. In comparison to the complex equipment required to measure WOB over full breathing cycles, equipment to determine breathing resistance at a single flow rate is very generic. Indeed, common filter efficiency testing equipment reports the breathing resistance during penetration testing (see, for example,

Manufacturers may use the variability for product development and evaluation. Large variability in WOB results was observed in this study, due to usual experimental variability and notably as well to the wide variety of respirator designs. The analysis in this work consciously neglects the effects of design differences, to allow predictions for the widest array of respirators. If the variability is reduced, the likelihood of false negatives in conformance testing could also be reduced. The respirators used in this study were limited to FFRs, half-mask elastomeric, full-facepiece respirators, and loose- and tight-fitting PAPRs. Other types of respirators such as supplied air respirators and self-contained breathing apparatus are not yet to be evaluated.

The authors thank the following persons for their contributions: Andrew Viner, Dan Warkander, David Cowgill, Wolfgang Drews, Andrew Capon, Colleen Miller, Ian Maxwell, Stan Ellis, Troy Baker, Ewa Messaoudi, Graham Bostock, Margaret Sietsema, George Niezgoda, Joe DuCarme, Lee Portnoff, Terry Thornton, Aaron Reeder, Jeffrey Palcic, and Rosalyn King.

Disclaimer

The findings and conclusions in this study are those of the author(s) and do not necessarily represent the official position of the National Institute for Occupational Safety and Health, Centers for Disease Control and Prevention. Mention of a company or product name does not constitute endorsement by NIOSH.

Breathing machine and 3D medium headform.

Test fixture for evaluation of inhalation resistance at a constant flow rate of 85 l/min using the NIOSH Standard Test Procedure (

Mean WOB/V_{T} (kPa) and NIOSH breathing resistance for inhalation and exhalation for all APRs.

Comparison of Mean WOB/V_{T} (kPa) with breathing resistance by RPD classes.

WOB_{in}/V_{T} (kPa) vs. inhalation resistance at 85 lpm for all APR at different work rates.

WOB_{ex}/V_{T} (kPa) vs. exhalation resistance at 85 l/min for all APR at different work rates.

Linear dependence of the fitted coefficients from

Actual by predicted plot for inhalation.

Actual by predicted plot for exhalation.

RPD classes, abbreviations, and test numbers for each class.

RPDs and abbreviations | Number of tests |
---|---|

Filtering facepiece [FFR] (N95) half-mask | 17 |

Filtering facepiece [FFR] (P100) half-mask | 1 |

Half-mask elastomeric respirator [EHM] (N95, P100 cartridge) | 3,7 |

Full-facepiece elastomeric respirator [EFF] (CBRN Cap-1, P100 cartridge) | 5,2 |

Powered air-purifying respirators [PAPR], loose-fitting (loose-fitting hood, helmet) | 6 |

Powered air-purifying respirators [PAPR], tight-fitting respirator | 2 |

Ratio of exhalation to inhalation for breathing resistance and WOB/V_{T} (kPa) by class.

RPD classes | Resistance | Resting | W1 | W2 | W3 | W4 |
---|---|---|---|---|---|---|

Filtering facepiece | 0.81 | 0.98 | 0.90 | 0.84 | 0.76 | 0.67 |

Half-mask elastomeric | 0.41 | 0.59 | 0.42 | 0.44 | 0.49 | 0.58 |

Full-facepiece w/ CBRN | 0.23 | 0.53 | 0.25 | 0.23 | 0.25 | 0.29 |

All RPD (mean) | 0.48 | 0.70 | 0.52 | 0.51 | 0.50 | 0.51 |

Correlation coefficient r and associated p-value between WOB/V_{T} and breathing resistance at different work rates for inhalation and exhalation by classes of respirators.

Variable | Variable | Correlation coefficient r | p-value for all classes | |||
---|---|---|---|---|---|---|

Full-facepiece | Half-mask elastomeric | Filtering facepiece | All classes | |||

WOB_{in}/V_{T} Resting | Inhalation resistance | 0.84 | 0.72 | 0.77 | 0.26 | 0.127 |

WOB_{in}/V_{T} W1 | 0.81 | 0.80 | 0.83 | 0.71 | <0.0001 | |

WOB_{in}/V_{T} W2 | 0.79 | 0.83 | 0.74 | 0.86 | <0.0001 | |

WOB_{in}/V_{T} W3 | 0.72 | 0.82 | 0.46 | 0.89 | <0.0001 | |

WOB_{in}/V_{T} W4 | 0.80 | 0.79 | 0.25 | 0.90 | <0.0001 | |

WOB_{ex}/V_{T} Resting | Exhalation resistance | 0.66 | 0.17 | 0.91 | 0.79 | <0.0001 |

WOB_{ex}/V_{T} W1 | 0.91 | 0.80 | 0.93 | 0.84 | <0.0001 | |

WOB_{ex}/V_{T} W2 | 0.90 | 0.67 | 0.92 | 0.88 | <0.0001 | |

WOB_{ex}/V_{T} W3 | 0.81 | 0.47 | 0.92 | 0.73 | <0.0001 | |

WOB_{ex}/V_{T} W4 | 0.74 | 0.54 | 0.92 | 0.59 | <0.0001 |

Lack of fit sum of squares analysis for inhalation and exhalation.

Source | DF | Sum of squares | Mean square | F ratio | p-level | Max R^{2} | |
---|---|---|---|---|---|---|---|

Inhalation | Lack of Fit | 165 | 6.60 | 0.04 | 0.75 | 0.78 | 0.99 |

Pure Error | 10 | 0.53 | 0.05 | ||||

Total Error | 175 | 7.14 | |||||

Exhalation | Lack of Fit | 154 | 1.33 | 0.01 | 0.32 | 0.99 | 0.98 |

Pure Error | 5 | 0.13 | 0.03 | ||||

Total Error | 159 | 1.46 |