Content uploaded by Edward Sinkule
Author content
All content in this area was uploaded by Edward Sinkule on Aug 08, 2016
Content may be subject to copyright.
Content uploaded by Edward Sinkule
Author content
All content in this area was uploaded by Edward Sinkule on Aug 08, 2016
Content may be subject to copyright.
Ann. Occup. Hyg., Vol. 57, No. 3, pp. 384–398, 2013
Published by Oxford University Press
on behalf of the British Occupational Hygiene Society 2012
doi:10.1093/annhyg/mes068
384
*Author to whom correspondence should be addressed.
Tel: +1-412-386-4054; fax: +1-412-386-6864;
email: ESinkule@cdc.gov
Evaluation of N95 Respirator Use with a Surgical
Mask Cover: Effects on Breathing Resistance and
Inhaled Carbon Dioxide
EDWARD JAMES SINKULE1*, JEFFREY BRYON POWELL1 and
FREDRIC LEE GOSS2
1National Personal Protective Technology Laboratory, National Institute for Occupational Safety and
Health, Centers for Disease Control and Prevention, Pittsburgh, PA, USA; 2Center for Exercise and
Health Fitness Research, University of Pittsburgh, Pittsburgh, PA, USA
Received 9 September 2011; in final form 28 July 2012; Advance Access publication 29 October 2012
Objective: For pandemic influenza outbreaks, the Institute of Medicine has recom-
mended using a surgical mask cover (SM) over N95 filtering facepiece respirators
(FFRs) among healthcare workers as one strategy to avoid surface contamination of the
FFR which would extend its efficacy and reduce the threat of exhausting FFR supplies.
The objective of this investigation was to measure breathing air quality and breathing
resistance when using FFRs with US Food and Drug Administration-cleared SM and
without SM.
Methods: Thirty National Institute for Occupational Safety and Health (NIOSH)-approved
FFR models with and without SM were evaluated using the NIOSH Automated Breathing and
Metabolic Simulator (ABMS) through six incremental work rates.
Results: Generally, concentrations of average inhaled CO2 decreased and average inhaled
O2 increased with increasing O2 consumption for FFR+SM and FFR-only. For most work
rates, peak inhalation and exhalation pressures were statistically higher in FFR+SM as
compared with FFR-only. The type of FFR and the presence of exhalation valves (EVs) had
significant effects on average inhaled CO2, average inhaled O2, and breathing pressures.
The evidence suggests that placement of an SM on one type of FFR improved inhaled
breathing gas concentrations over the FFR without SM; the placement of an SM over
an FFR+EV probably will prevent the EV from opening, regardless of activity intensity;
and, at lower levels of energy expenditure, EVs in FFR do not open either with or without
an SM.
Conclusions: The differences in inhaled gas concentrations in FFR+SM and FFR-only
were significant, especially at lower levels of energy expenditure. The orientation of the
SM on the FFR may have a significant effect on the inhaled breathing quality and
breathing resistance, although the measurable inhalation and exhalation pressures
caused by SM over FFR for healthcare users probably will be imperceptible at lower
activity levels.
Keywords: breathing resistance; effects from using N95 respirators; extending N95 respirators during influenza out-
break; inhaled carbon dioxide; inhaled oxygen; metabolic simulator; N95 respirator; respiratory protection; surgical
mask; using N95 respirators with surgical or procedure masks
by guest on August 8, 2016http://annhyg.oxfordjournals.org/Downloaded from
Evaluation of N95 respirator use with a surgical mask cover 385
INTRODUCTION
At the request of the Department of Health and
Human Services, the Institute of Medicine (IOM)
convened a Committee on the Development of
Reusable Facemasks for Use During an Influenza
Pandemic in order to report on the solutions, limita-
tions, threats, and possible opportunities of reusing
respirators for infection control during an influenza
pandemic (Institute of Medicine and Committee
on the Development of Reusable Facemasks for
Use During an Influenza Pandemic, 2006). The
Committee offered recommendations for extend-
ing the life of disposable N95 filtering facepiece
respirators (FFRs) for individual users. One recom-
mendation involved placing a surgical mask cover
(SM) over the respirator in order to prevent respi-
rator surface contamination. Previous studies have
reported elevated concentrations of inhaled carbon
dioxide (CO2) and decreased concentrations of
inhaled oxygen (O2) associated with wearing FFRs
(Sinkule et al., 2003). Others have proposed that the
adverse effects of wearing FFRs (e.g., headache and
increased sick days) are the result of elevated inhaled
CO2 concentrations (Lim et al., 2006). The increased
inhaled CO2 concentrations and decreased inhaled
O2 concentrations within the breathing zone of neg-
ative-pressure air-purifying respirators, including
FFRs, are directly related to dead space. Changes
resulting from the addition of an US Food and Drug
Administration (FDA)-cleared surgical mask could
include increased levels of CO2, breathing resist-
ance, and temperature of inspired air, and decreased
levels of O2. National Institute for Occupational
Safety and Health (NIOSH) certification tests that
measure minimum and average inhaled CO2 con-
centrations, and maximum and average inhaled O2
concentrations, apply only to respirators certified
for protection against chemical, biological, radio-
logical, or nuclear (CBRN) hazards. A European test
standard (EN 149:2001) characterizes inhaled CO2
concentration and breathing resistance, among other
assessments from FFR use.
The effects of wearing FFRs and other types of
respiratory protection have been widely studied
using a variety of measurement methods (Li et al.,
2005; Radonovich et al., 2009; Roberge et al., 2010).
Some of these investigations have been quantitative
(e.g., levels of inhaled CO2), qualitative (e.g., levels
of fatigue), or can reflect characteristics that range
from inconvenient (e.g., decreased levels of comfort)
to potentially hazardous (e.g., decreased inhaled
levels of O2). The physiological effects of breath-
ing elevated inhaled CO2 may include changes in
visual performance (Yang et al., 1997), modified
exercise endurance (Raven et al., 1979), headaches
and dyspnea (Raven et al., 1979). The psychological
effects include decreased reasoning and alertness,
and increased irritability (Sayers et al., 1987); with
CO2 at 7–7.5%, severe dyspnea, headache, dizzi-
ness, perspiration, and short-term memory loss have
been reported (Sayers et al., 1987; Compressed Gas
Association, 1999). Subjects performing physical
activity while breathing decreased O2 concentra-
tions (17%) produced higher levels of lactic acid
accumulation at lower levels of energy expenditure
as compared with normal O2 concentrations (21%),
in addition to achieving lower levels of peak exercise
performance (Hogan et al., 1983). Increased breath-
ing resistance with respirators has been identified as
the cause of respiratory fatigue and impaired physi-
cal work capacity, a shift to anaerobic metabolism
from an increased rate of O2 debt; and, early exhaus-
tion at lighter workloads.
To date, no laboratory or f ield studies have been
published to provide data on the effect of protective
covers (e.g., surgical masks) on the breathing pres-
sures and concentrations of inhaled respiratory gases
among multiple models and types of FFRs, including
many matched FFR models paired with and without
exhalation valves (EVs). The major purpose of this
study was to evaluate the inhaled CO2 and O2 concen-
trations and breathing pressures of NIOSH-certified
FFRs with and without an SM using an Automated
Breathing and Metabolic Simulator (ABMS)-based
test. An ABMS test protocol was used to character-
ize performance in terms of minimum and average
inhaled CO2 concentrations, maximum and average
inhaled O2 concentrations, and inhalation and exha-
lation pressures. The evaluations were repeated with
identical FFRs worn with an SM.
METHODS
Respirator and surgical mask selection
The selection of NIOSH-certified FFR mod-
els used for this study was determined by market
analysis, with a focus on those FFR models in the
US Strategic National Stockpile (SNS) at the time
of this investigation (US Department of Health and
Human Services, 2009). None of the FFRs in the
SNS had EVs. In order to evaluate both conditions,
FFRs without and with EVs were included in this
study. No assumptions were made regarding the
materials or construction details of the entire sample
of FFRs, including EVs, FFR fabric, or head straps.
The SM model was a device also identified as being
by guest on August 8, 2016http://annhyg.oxfordjournals.org/Downloaded from
386 E. J. Sinkule, J. B. Powell and F. L. Goss
in the SNS. A sample of at least four respirators of
a consistent common size (medium, medium/large,
or universal) among each the following types was
tested in the present investigation: N95 cup, N95
horizontal flat-fold, and N95 other flat-fold. The
N-series FFR is restricted for use in workplaces
free of oil aerosols, and 95 means the filter device
is 95% efficient for filtering a mean particle size up
to 0.075 ± 0.020 μm. The type of FFR (for example,
cup or horizontal flat-fold) is a manufacturer desig-
nation and is not a classification or differentiation in
a NIOSH standard.
A consistent common size in FFR does not mean
the same dead space. Constructional differences
between the different types of FFRs complicated the
characterization of dead space even further. In the
sample of FFRs used in this investigation, includ-
ing those from the SNS, “medium”-sized FFRs were
selected when various sizes were available. Size
specifications for several FFRs, however, included
intermediate sizes, such as, “one-size-fits-all” (also
known as “universal size”) and “medium/large” sizes.
Intermediate sizes and manufacturer-specific FFRs
were unbalanced among respirator models between
FFRs with and without EVs. Furthermore, as there
are no federal regulations or industry standards for
sizing FFRs, dissimilar sizes between manufacturers
would occur and would affect respirator dead space.
Comparison of respirator dead space among the vari-
ous respirators, therefore, may not be meaningful for
individual FFR performance. In the paired-valve
FFR subset, manufacturer-specific respirators and
sizes between FFRs with and without EVs were the
same, i.e. balanced among the respirator models. It
was necessary to include a variety of manufacturers
for the sample in order to ensure that a representative
market sample was evaluated (see Table 1).
Thirty NIOSH-approved FFR models and one
flat-fold SM model were selected for this investi-
gation. The FFR sample represented approximately
10% of the 300+ NIOSH-approved FFR models
(National Institute for Occupational Safety and
Health, 2011). Table 1 provides the grouping among
the respirator models for each type and the presence
of an EV. The SM used was Medline NON27382,
which was also in the SNS. This SM was a random
selection between two SMs available from the SNS
that did not have ear loops for attachments thus com-
patible with a head form without ears. Among the
30 respirator models, 18 were of the cup type, six
of the horizontal flat-fold type, and six of the other
flat-fold type. The “other flat-fold” types included
three vertical flat-fold and three tri-fold respirators.
The results were grouped by manufacturer’s clas-
sification and the relative number of models (see
Table 1) in each group. Eighteen respirator models
(five cup pairs, two horizontal flat-fold pairs, and
two other flat-fold pairs) were paired-valve models,
that is, the same respirator except one with and one
without an EV.
Table 1. Grouping of the FFR type and valve.
FFR type FFR without exhalation valve FFR with exhalation valve
Cup 3M 1860 (M)a3M 8211 (O)
3M 8000 (O)a3M 8212 (O)
3M 8210 (O)a3M 8511 (O)
Inovel 3002 (M)a3M 8512 (O)
AO Safety N9504C (O)bAO Safety N9505C (O)b
Crews RPN951 (O)bCrews RPN952 (O)b
Gerson 1730 (O)a,b Gerson 1740 (O)b
Moldex 2200 (M/L)a,b Moldex 2300 (M/L)b
Moldex-Metrics 2600 (M/L)bMoldex-Metrics 2700 (M/L)b
Horizontal Kimberly-Clark 46727 (O)aWillson N9520FM (M)
Flat-fold Crews RPFN951 (O)bCrews RPFN952 (O)b
San Huei SH2950 (O)bSan Huei SH2950V (O)b
Other 3M 1870 (O)aDräger Piccola (O)
Flat-fold 3M 9210 (O)a,b 3M 9211 (O)b
San Huei SH3500 (O)bSan Huei SH3500V (O)b
Respirators are FFR (size).
M, medium size; M/L, medium/large size; O, one-size-fits-all/universal size.
aSelected FFR from the Strategic National Stockpile.
bPaired-valve respirators: the same FFR with or without an exhalation valve.
by guest on August 8, 2016http://annhyg.oxfordjournals.org/Downloaded from
Evaluation of N95 respirator use with a surgical mask cover 387
Automated breathing and metabolic simulator
The ABMS is ideal for quantitative and repeatable
testing and evaluation of FFRs. The ABMS (Ocenco,
Inc., Pleasant Prairie, WI) has the capability to simu-
late the following human metabolic parameters: O2
consumption, CO2 production, respiratory frequency,
tidal volume, breathing waveform shape, and heated
and humidified exhaled breathing gas. In addition,
any number of work rates may be serially combined
in any order to simulate various activities. The capac-
ity ranges for the parameters are as follows: minute
ventilation, 0–160 l·min−1; O2 consumption, 0–7
l·min−1; CO2 production, 0–7 l·min−1; respiratory fre-
quency, 0–100 breaths·min−1; tidal volume, 0–5 l; and
human-like breathing gas temperatures, 30–45°C. All
gas volume parameters were at standard conditions
of temperature (0°C), pressure (760 mmHg), and dry
(no water vapor) unless indicated otherwise. A sinu-
soid waveform was used for ventilation rates below
50 l·min−1. A trapezoid or human-like waveform was
used for ventilation rates above 50 l·min−1.
The ABMS is capable of monitoring the follow-
ing variables: flow-weighted average inhaled con-
centrations of O2 and CO2, minimum inhaled CO2
concentration, maximum inhaled O2 concentra-
tion, breathing pressures, and inhaled dry-bulb and
wet-bulb gas temperatures. The capacity ranges for
these parameters are as follows: O2 concentration,
0–100%; CO2, 0–15%; breathing pressure, ±700 mm
of water (H2O); and inhaled dry-bulb and wet-bulb
gas temperatures, 0–100°C. The results of this inves-
tigation only will report the average inhaled CO2 and
O2 concentrations, and peak inhalation and exhala-
tion pressures. A more detailed explanation is given
in the appendix of the US Bureau of Mines publi-
cation by Kyriazi (1986). Deno (1984) provides a
description of the development of the ABMS. There
has been one research report where the responses
from the ABMS with air-purifying CBRN escape
hoods were shown to be very similar to responses
from human subject volunteers, although a statisti-
cal comparison between the ABMS and human sub-
jects was not conducted (Sinkule and Turner, 2004).
Experimental design and variables
Each week, FFR testing was preceded by instru-
ment calibration (pressure transducer, dry-bulb ther-
mocouple, and wet-bulb thermocouple) and routine
room air validation studies. Before each test, the fast-
response ABMS gas analyzers were calibrated using
primary standard calibration gases (15% O2 and 8%
CO2), when response time (<100 ms) and transport
time were calculated (<300 ms) and used to electron-
ically offset sample time. Prior to attaching the FFR,
a leak test was performed on the ABMS. According
to the manufacturers’ instructions, respirators were
then placed on a head form attached to the trachea of
the ABMS. A sealant (Poli-Grip®) was applied to the
contact area between the head form and FFR in order
to create a seal between the facial surface of the head
form with each FFR. The face seal with a NIOSH-
approved FFR among human users is assessed with a
fit-test. No sealant was used while donning an SM to
the FFR since a fit-test is not needed for using surgi-
cal masks. FFRs, with or without SM, were tested for
a minimum of 5 min at each work rate.
The breathing frequencies (f), tidal volumes (VT),
minute ventilation rates (V
.
E), O2 consumption rates
(V
.O2), CO2 production rates (V
.CO2), and respira-
tory quotients (R) programmed into the ABMS are
shown in Table 2. These metabolic rates represent a
progression from light to very intense energy expen-
ditures. For healthcare workers, the range of energy
expenditure can be from very light (e.g. desk work
used for writing patient notes, 1.8 METs; or per-
forming procedures in an operating room, 3 METs)
to moderate (e.g. moving patients 34 kilograms or
more, 7.5 METs) to very intense (e.g. responding to
emergency calls by paramedics, patient care by phys-
ical therapists, and emergency calls performed by
Table 2. Metabolic variables for the ABMS exercise protocol.
Test level f (breath/min) VT (l, BTPS)
V
.
E (l·min−1,
BTPS) V
.O2 (l·min−1, STPD) V
.CO2 (l·min−1, STPD) R
1 12.9 0.92 11.9 0.5 0.4 0.80
2 19.5 1.57 30.6 1.0 0.8 0.80
3 28.0 1.63 45.6 1.5 1.3 0.87
4 32.6 2.30 75.0 2.0 1.9 0.95
5 34.2 2.42 82.8 2.5 2.5 1.00
6 36.4 2.66 96.8 3.0 3.15 1.05
(f=frequency of breathing; VT=tidal volume; V
.
E=minute ventilation, expired; V
.O2=oxygen consumption; V
.CO2=carbon dioxide
production; R=respiratory quotient; BTPS=volumes expressed at body temperature (37ºC), ambient pressure, and saturated with
water vapor.)
by guest on August 8, 2016http://annhyg.oxfordjournals.org/Downloaded from
388 E. J. Sinkule, J. B. Powell and F. L. Goss
flight nurses, >10 METs) (Ainsworth et al., 2000).
One MET (metabolic equivalent) is equal to a rest-
ing metabolic rate equal to quiet sitting, or 3.5 ml of
oxygen consumed·kg−1·min−1.
In a randomized fashion, selected models of FFRs
and the same models of FFRs with a selected SM
were evaluated as the test conditions. Flow-weighted
average inhaled CO2 and O2 concentrations as well as
peak and average inhaled and exhaled pressures at the
mouth were measured by the ABMS, and arithmetic
means of these variables were calculated. The data
during the last minute of each variable at each level of
energy expenditure (Table 2) for each condition (FFRs
only and FFRs with SM) were used for analysis.
Data from a previous NIOSH study that inves-
tigated the inhaled CO2 concentrations in various
respirators using the ABMS were used to deter-
mine sample sizes in the current study (Sinkule
et al., 2003). The mean fractional minimal inhaled
CO2 concentration was 0.25%, the within-group, or
within-respirator standard deviation (pooled over the
different units) was 0.09%. The between-group vari-
ability represented 94.4% of the overall variability in
CO2 measurements. For the initial power estimate,
the analysis of variance contrast between those with
and without an SM will have similar power to the
two-sample t-test with a pooled standard deviation
of 0.09% and a sample size of n = 0.944 × 96, which
(conservatively) ≈90 per group. The sample size,
therefore, of at least four respirators of each model
was tested for 5 min at each oxygen consumption
rate, both with and without an SM. A total of 281
trials were completed, or an average of 4.7 trials for
each respirator model tested with and without an SM.
Dependent variables (average inhaled CO2 con-
centrations, average inhaled O2 concentrations, and
peak inhalation and exhalation pressures) were ana-
lyzed using 2 × 2 × 3 (SM × EV × FFR type) factorial
analysis of variance (ANOVA). For each dependent
variable, comparisons between the treatments (SM
versus no SM) were performed for respirators with
and without EVs and for each type of FFR (cup,
horizontal flat-fold, and other flat-fold). The Tukey
multiple comparison test was used for all post hoc
analysis of significant effects. Statistical significance
for ANOVA and Tukey analyses was set a priori at
P < 0.05. Data analyses were performed using SPSS,
version 17 (SPSS, Inc., Chicago, IL, USA).
RESULTS
Respiratory gases
Results for the average inhaled CO2 concentra-
tion among FFRs with and without SM are shown
in Table 3. Among the six levels of energy expendi-
ture, the average inhaled CO2 concentrations were
higher (P < 0.05) among the cup FFRs with SM as
compared with cup FFRs alone at V
.O2 of 0.5 l·min−1,
2.5 l·min−1, and 3.0 l·min−1. The average inhaled
CO2 concentrations were lower (P < 0.05) among
horizontal flat-fold FFRs with SM as compared with
horizontal flat-fold FFRs alone at V
.O2 of 1.0 l·min−1
and 1.5 l·min−1. The average inhaled CO2 concen-
trations were not different between other flat-fold
FFRs with and without SM. Significant interactions
(P < 0.05) between FFR type and SM use on average
inhaled CO2 were observed for
VO2of 1.0 l·min−1,
Table 3. Average inhaled carbon dioxide concentrations (%) among FFRs with and without SM.
Oxygen consumption
(l·min−1) Treatment
Horizontal Other
Cup (n = 18) Flat-fold (n = 6) Flat-fold (n = 6)
0.5 FFR only 2.49 ± 0.51 3.52 ± 0.93 2.65 ± 0.57
FFR + SM 2.93 ± 0.38*3.14 ± 0.64 3.13 ± 0.40
1.0 FFR only 1.64 ± 0.53 2.87 ± 1.12 1.93 ± 0.66
FFR + SM 1.98 ± 0.39 2.00 ± 0.44*2.01 ± 0.12
1.5 FFR only 2.09 ± 0.82 3.23 ± 1.32 2.31 ± 0.94
FFR + SM 2.31 ± 0.41 2.30 ± 0.46*2.21 ± 0.09
2.0 FFR only 1.43 ± 0.60 1.81 ± 0.82 1.65 ± 0.73
FFR + SM 1.75 ± 0.33 1.67 ± 0.33 1.58 ± 0.15
2.5 FFR only 1.28 ± 0.57 1.66 ± 0.77 1.52 ± 0.73
FFR + SM 1.65 ± 0.38*1.52 ± 0.26 1.48 ± 0.16
3.0 FFR only 1.52 ± 0.65 1.90 ± 0.87 1.79 ± 0.89
FFR + SM 1.99 ± 0.33*1.75 ± 0.32 1.71 ± 0.22
Values are means ± SD.
*Significantly different from FFR only, P < 0.05.
by guest on August 8, 2016http://annhyg.oxfordjournals.org/Downloaded from
Evaluation of N95 respirator use with a surgical mask cover 389
2.5 l·min−1, and 3.0 l·min−1; between FFRs with EVs
and SM use at
VO2of 1.5–3.0 l·min−1.
Results for the average inhaled O2 concentration
among FFRs with and without SM are presented in
Table 4. Among the six levels of energy expenditure,
the average inhaled O2 concentrations were lower
(P < 0.05) among the cup FFRs with SM as com-
pared with cup FFRs alone at
VO2of 0.5 l·min−1,
2.5 l·min−1, and 3.0 l·min−1. The average inhaled O2
concentrations were higher (P < 0.05) among hori-
zontal flat-fold FFR with SM as compared with hori-
zontal flat-fold FFRs alone at
VO2of 1.0 l·min−1 and
1.5 l·min−1. The average inhaled O2 concentrations
were not different between other flat-fold FFRs with
and without SM. Significant interactions (P < 0.05)
between FFR type and EV on average inhaled O2
concentration were observed for
VO2of 0.5 l·min−1
only; between FFR type and SM use on average
inhaled O2 for
VO2of 1.0 l·min−1, 1.5 l·min−1, and
3.0 l·min−1; between FFRs with EVs and SM use at
VO2of 1.5–3.0 l·min−1.
Breathing pressures
Peak inhalation and exhalation pressures are used
as measures of breathing resistance among respirator
users. Results for peak exhalation pressure among
FFRs with and without SM are presented in Table 5.
Among the six levels of energy expenditure, the peak
exhalation pressures were higher (P < 0.05) among
the cup FFRs with SM as compared with cup FFRs
alone at
VO2of 1.5 l·min−1, 2.0 l·min−1, 2.5 l·min−1,
and 3.0 l·min−1. Peak exhalation pressures were not
different between horizontal flat-fold FFRs with or
without SM, or between other flat-fold FFRs with
or without SM. In addition, a significant (P < 0.05)
main effect of SM use and FFRs with an EV on peak
exhalation pressure was observed for
VO2 of 1.5–
3.0 l·min−1.
Results for peak inhalation pressure among FFRs
with and without SM are presented in Table 6.
Among the six levels of energy expenditure, the peak
inhalation pressures were higher (P < 0.05) among
the cup FFRs with SM as compared with cup FFRs
alone at every level of O2 consumption. Peak inhala-
tion pressures were different between other flat-fold
FFRs with or without SM at
VO2 of 2.0 l·min−1,
2.5 l·min−1, and 3.0 l·min−1. Peak inhalation pres-
sures were not different between horizontal flat-fold
FFRs with and without SM. In addition, a significant
(P < 0.05) main effect of SM use on peak inhalation
pressure was observed for
VO2of 1.0–3.0 l·min−1.
The presence of an EV is intended to affect
breathing resistance by reducing exhalation pres-
sure. Fifteen of the 30 FFR models contained an EV.
The EV is a flexible dam (usually made of rubber)
anchored to a circular frame that is mounted in the
wall of the FFR directly in the front of the breath-
ing zone. Upon negative mask pressure created by
inhalation, the flexible dam is pulled into its frame
to create a seal and prevent air leaking into the FFR
mask. Table 7 contains results of peak inhalation and
exhalation pressures in FFRs with and without EVs
as compared with FFRs with and without SM.
Paired-valve FFR models
Eighteen paired-valve FFR models (five pairs of
cup, two pairs of horizontal flat-fold, and two pairs of
other flat-fold) provided a homogeneous subset for
Table 4. Average inhaled oxygen concentrations (%) among FFRs with and without SM.
Oxygen consumption (l·min−1) Treatment
Horizontal Other
Cup (n = 18) Flat-fold (n = 6) Flat-fold (n = 6)
0.5 FFR only 17.40 ± 0.81 16.10 ± 1.14 17.31 ± 0.77
FFR + SM 16.81 ± 0.54* 16.52 ± 0.79 16.58 ± 0.67
1.0 FFR only 18.84 ± 0.77 17.30 ± 1.39 18.47 ± 0.89
FFR + SM 18.39 ± 0.50 18.39 ± 0.55* 18.29 ± 0.17
1.5 FFR only 18.49 ± 1.04 17.15 ± 1.52 18.22 ± 1.13
FFR + SM 18.22 ± 0.49 18.25 ± 0.51* 18.25 ± 0.09
2.0 FFR only 19.33 ± 0.70 18.92 ± 0.84 19.08 ± 0.84
FFR + SM 18.96 ± 0.37 19.05 ± 0.35 19.05 ± 0.15
2.5 FFR only 19.52 ± 0.65 19.12 ± 0.77 19.26 ± 0.82
FFR + SM 19.11 ± 0.41* 19.25 ± 0.28 19.19 ± 0.16
3.0 FFR only 19.32 ± 0.71 18.95 ± 0.83 19.03 ± 0.96
FFR + SM 18.82 ± 0.38* 19.06 ± 0.34 18.98 ± 0.23
Values are means ± SD.
*Significantly different from FFR only, P < 0.05.
by guest on August 8, 2016http://annhyg.oxfordjournals.org/Downloaded from
390 E. J. Sinkule, J. B. Powell and F. L. Goss
analyses. Where the main sample of FFR represented
a variety of respirator shapes across the different
FFR types (cup, horizontal flat-fold, and other flat-
fold) for those with and without EVs, the subset of
paired-valve FFR represented identical FFRs, with
and without EVs.
Figure 1 illustrates the effects of EVs on the aver-
age inhaled CO2 concentration among FFRs with
and without SM. The bars are the mean difference
(delta) of the average inhaled CO2 concentration in
FFRs with EVs and the average inhaled CO2 con-
centration in FFRs without EVs. The black filled
bars are the delta CO2 concentrations among FFRs
without SM. The white filled bars are the delta
CO2 concentrations among FFRs with SM. From
Figure 1 it can be seen that FFRs with a SM (light
bars) produced small and similar delta values in
average inhaled CO2 concentrations among a spec-
trum of energy expenditures in the matched models
of FFRs paired with and without EVs, and there
were no significant differences in the delta values
of average inhaled CO2 between FFRs and FFR+SM
among the matched models of FFRs paired with and
without EVs below
VO2of 1.5 l·min−1. A signifi-
cant (P < 0.05) main effect was observed with SM
use at
VO2of 1.5 l·min−1 for average inhaled CO2.
Significant interactions (P < 0.05) between FFR type
and SM use on average inhaled CO2 concentration
Table 5. Peak exhalation pressures (mmH2O) among FFRs with and without SM.
Oxygen consumption (l·min−1) Treatment
Horizontal Other
Cup (n = 18) Flat-fold (n = 6) Flat-fold (n = 6)
0.5 FFR only 8 ± 2 7 ± 2 7 ± 4
FFR + SM 8 ± 2 8 ± 2 9 ± 3
1.0 FFR only 10 ± 3 11 ± 3 9 ± 3
FFR + SM 12 ± 3 12 ± 3 11 ± 2
1.5 FFR only 14 ± 4 15 ± 3 14 ± 4
FFR + SM 17 ± 4* 18 ± 4 17 ± 2
2.0 FFR only 23 ± 7 24 ± 4 22 ± 6
FFR + SM 29 ± 7* 30 ± 8 28 ± 4
2.5 FFR only 20 ± 6 21 ± 4 19 ± 5
FFR + SM 25 ± 6* 26 ± 6 24 ± 4
3.0 FFR only 24 ± 8 25 ± 4 23 ± 6
FFR + SM 30 ± 7* 31 ± 8 29 ± 4
Values are means ± SD.
*Significantly different from FFR only, P < 0.05.
Table 6. Peak inhalation pressures (mmH2O) among FFRs with and without SM.
Oxygen consumption (l·min−1) Treatment
Horizontal Other
Cup (n = 18) Flat-fold (n = 6) Flat-fold (n = 6)
0.5 FFR only −6 ± 1 −5 ± 2 −6 ± 2
FFR + SM −7 ± 2* −6 ± 2 −7 ± 1
1.0 FFR only −12 ± 2 −12 ± 4 −12 ± 2
FFR + SM −15 ± 3* −14 ± 5 −14 ± 2
1.5 FFR only −19 ± 4 −19 ± 6 −18 ± 2
FFR + SM −23 ± 5* −23 ± 8 −23 ± 3
2.0 FFR only −35 ± 6 −34 ± 10 −33 ± 3
FFR + SM −41 ± 7* −43 ± 16 −44 ± 11*
2.5 FFR only −35 ± 6 −34 ± 10 −33 ± 3
FFR+ SM −42 ± 8* −44 ± 16 −45 ± 12*
3.0 FFR only −41 ± 7 −42 ± 13 −40 ± 3
FFR + SM −49 ± 9* −54 ± 22 −56 ± 17*
Values are means ± SD.
*Significantly different from FFR only, P < 0.05.
by guest on August 8, 2016http://annhyg.oxfordjournals.org/Downloaded from
Evaluation of N95 respirator use with a surgical mask cover 391
were observed for
VO2 of 0.5 l·min−1, 1.0 l·min−1,
1.5 l·min−1, and 3.0 l·min−1; between FFRs with EV
and SM use on average inhaled CO2 concentration
observed at
VO2 of 1.5–3.0 l·min−1.
Figure 2 illustrates the effects of EVs on the aver-
age inhaled O2 concentration among FFRs with and
without SM. The bars are the mean difference (delta)
of the average inhaled O2 concentration in FFRs with
EVs and the average inhaled O2 concentration in FFRs
without EVs. The black filled bars are the delta O2
concentrations among FFRs without SM. The white
filled bars are the delta O2 concentrations among
FFRs with SM. Like the conclusions presented for
Figure 1 (average inhaled CO2), FFRs with an SM
Table 7. Peak inhalation and exhalation pressures (mmH2O) among FFRs with and without EV between FFRs with and without
SM.
Oxygen consumption (l·min−1) Treatment
Peak inhalation pressure Peak Exhalation Pressure
FFR − EV FFR + EV FFR − EV FFR + EV
0.5 FFR only −6 ± 1 −5 ± 2 7 ± 2 8 ± 2
FFR + SM −7 ± 2 −7 ± 2* 8 ± 2 8 ± 3
1.0 FFR only −11 ± 2 −12 ± 3 11 ± 3 10 ± 3
FFR + SM −14 ± 3* −15 ± 4* 13 ± 2* 10 ± 3
1.5 FFR only −18 ± 3 −20 ± 4 16 ± 4 13 ± 4
FFR + SM −22 ± 4* −24 ± 6* 19 ± 3* 15 ± 3
2.0 FFR only −33 ± 5 −36 ± 8 27 ± 6 20 ± 5
FFR + SM −40 ± 8* −45 ± 11* 33 ± 6* 25 ± 5*
2.5 FFR only −33 ± 5 −36 ± 8 23 ± 5 17 ± 5
FFR + SM −40 ± 8* −45 ± 12* 28 ± 5* 22 ± 4*
3.0 FFR only −39 ± 5 −43 ± 9 27 ± 6 20 ± 6
FFR + SM −49 ± 12* −54 ± 15* 34 ± 6* 26 ± 5*
Values are means ± SD.
*Significantly different from FFR only, P < 0.05.
Fig. 1. Average inhaled CO2 delta (FFR with exhalation valve minus FFR without exhalation valve) concentrations (mean ± SD)
in FFR with and without surgical mask covers for all levels of oxygen consumption. FFR without surgical mask covers (FFR
only) = black bars, FFR with surgical mask covers (FFR+SM) = white bars. *Significantly different from FFR+SM, P < 0.05.
by guest on August 8, 2016http://annhyg.oxfordjournals.org/Downloaded from
392 E. J. Sinkule, J. B. Powell and F. L. Goss
(light bars) produced small and similar delta values
in average inhaled O2 concentrations among a spec-
trum of energy expenditures in the matched models
of FFRs paired with and without EVs, and there were
no significant differences in the delta values of aver-
age inhaled O2 between FFRs and FFR+SM among
the matched models of FFRs paired with and without
EVs below
VO2 of 1.5 l·min−1. Signif icant interac-
tions (P < 0.05) between FFR type and SM use on
average inhaled O2 concentration were observed for
VO2 of 0.5 l·min−1, 1.0 l·min−1, and 1.5 l·min−1, and
3.0 l·min−1; and, between FFR with EV and SM use on
average inhaled O2 concentration observed at
VO2of
1.5 l·min−1, 2.0 l·min−1, 2.5 l·min−1, and 3.0 l·min−1
.
For both peak inhalation and peak exhalation pres-
sures, the delta values between the FFRs with and
without EVs were not different for the FFRs and
FFR+SM. A significant (P < 0.05) main effect of
SM use was observed for peak exhalation pressure
at
VO2 of 1.5–3.0 l·min−1. A significant (P < 0.05)
main effect of SM use on peak inhalation pressure
was observed for
VO2 of 1.0–3.0 l·min−1.
DISCUSSION
Respiratory gases
Respirator scientists have known that respiratory
protection may have adverse effects on breathing
pressures from restricted flow characteristics and
inhaled CO2 and O2 concentrations from increased
dead space. Increased dead space causes an increase
in tidal volume and respiratory rate (Harber et al.,
1982). However, little data exist on the effects of
an additional cover (i.e. SM) on the respiratory gas
concentrations and pressures resulting from the res-
pirator, and the effects of the shape of the additional
cover on the characteristics of the FFR (Roberge,
2008). Other NIOSH research has indicated signifi-
cant elevated inhaled CO2 associated with various
respirators. Sinkule et al. (2003) investigated five
types of respiratory protection using the ABMS:
air-purifying respirators (n = 27), air-supplied res-
pirators (n = 20), gas masks (n = 6), powered air-
purifying respirators (n = 11), and FFRs (n = 26).
Using the same six levels of energy expenditure as
the present investigation, FFRs (type was not strati-
fied) produced the highest levels of average inhaled
CO2 concentrations and lowest average inhaled O2
concentrations for all levels of energy expenditure
as compared with all other respiratory protective
devices examined. Table 3 contains average inhaled
CO2 concentrations among the FFRs used in the pre-
sent investigation. The practical significance of these
findings includes the influence of dead space upon
the inhaled CO2 concentrations among horizontal
flat-fold FFRs, which were larger in comparison than
the other types of FFRs without an SM.
Fig. 2. Average inhaled O2 delta (FFR with exhalation valve minus FFR without exhalation valve) concentrations (mean ± SD)
in FFR with and without surgical mask covers for all levels of oxygen consumption. FFR without surgical mask covers (FFR
only) = black bars, FFR with surgical mask covers (FFR+SM) = white bars. Significantly different from FFR+SM, P < 0.05.
by guest on August 8, 2016http://annhyg.oxfordjournals.org/Downloaded from
Evaluation of N95 respirator use with a surgical mask cover 393
There currently is no inhaled CO2 concentration
threshold value for NIOSH certification testing of
FFRs. The average inhaled CO2 concentrations were
above the NIOSH Short-Term Exposure Limit of
0.5% among all levels of energy expenditure, without
and with SM. The threshold for the NIOSH Ceiling
of CO2 is 3% by volume. The NIOSH Ceiling is used
to describe occupational exposures that shall not be
exceeded through any part of the workday (American
Conference of Governmental Industrial Hygienists,
2008). Individually, three of the 30 FFR models tested
without SM produced average inhaled CO2 concen-
trations above 4%, which is Immediately Dangerous
to Life and Health (IDLH). IDLH is the designation
of maximal exposure above which only highly reli-
able respiratory protection provides maximal worker
protection. One FFR model was withdrawn from the
analysis based upon the exceptionally large average
inhaled CO2 concentration—5.8% at
VO2 of 0.5
l·min−1. The performance of this FFR model was a
true outlier that skewed analyses. Manufacturing of
this FFR was discontinued during the course of this
investigation and it is no longer commercially avail-
able. In addition to the discontinued FFRs, another
FFR (without SM) produced average inhaled CO2
concentrations above 5% at the energy expenditure
of 1.5 l·min−1. Of the FFRs without SM, three hori-
zontal flat-fold FFR models (one with an EV) pro-
duced average inhaled CO2 concentrations between
4 and 5% at
VO2 of 0.5 l·min−1, as well as one hori-
zontal flat-fold FFR at
VO2 of 1.0 l·min−1 and one
cup FFR at
VO2 of 1.5 l·min−1.
The respirator provides a micro-environment for
the exposure pathway of inhaled CO2 (Checkoway
et al., 2004; Nieuwenhuijsen, 2006). According to
CFR 42 Part 84, the highest inhaled CO2 concentra-
tion specified for respiratory protection is 2.5% for
≤30 min among the self-contained breathing appa-
ratus (Approval of Respiratory Protection Devices,
2006). A standard test procedure used by NIOSH
for the evaluation of negative-pressure air-purify-
ing hooded respirators for escape only contains an
inhaled CO2 concentration threshold of 2.5% for
apparatus of 15–30 min duration, and 2.0% for an
apparatus of 45–60 min duration (National Institute
for Occupational Safety and Health, 2006). Of the
FFRs in this investigation, most average inhaled
CO2 concentrations were lower than 2.0% for levels
of energy expenditure at 2.0 l·min−1 or greater, both
without and with an SM. At levels of energy expendi-
ture of 1.5 l·min−1 or lower, most average inhaled CO2
concentrations appeared above 2.0% for all FFRs
and more so at the lowest level of energy expendi-
ture (rest). The recognizable effect of inhaled CO2 is
the stimulating action upon respiration, i.e. respira-
tory removal of CO2 occurs through the increase in
ventilation rate. Respiratory rate, tidal volume, and
alveolar CO2 become elevated with inhaled CO2 con-
centrations above ambient (Schneider and Truesdale,
1922; Consolazio et al., 1947; Patterson et al., 1955).
These physiological responses occur to compensate
for abnormal diffusion of CO2 from the blood, due
to a decrease in the ratio of alveolar to capillary CO2
(Schulte, 1964). In addition to the increased rate and
depth of breathing, cardiac output will increase to
compensate for the additional CO2 (Schulte, 1964).
While inhaling 1–2% CO2 for 17–32 min, slight
increases have been reported in systolic and diastolic
blood pressures (Schneider and Truesdale, 1922).
Exposures of increased inhaled CO2 between 2 and
3% have been known to produce sweating, headache,
and dyspnea for some subjects at rest after several
hours (Schneider and Truesdale, 1922). If inhaled
CO2 concentrations are between 4 and 5%, dysp-
nea can occur within several minutes and increased
blood pressure, dizziness, and headache can occur
within 15–32 min (Schneider and Truesdale, 1922;
Patterson et al., 1955; Schulte, 1964). If inhaled
CO2 exposures are at 5%, mental depression may
occur within several hours (Consolazio et al., 1947;
Schulte, 1964). As noted in several of these studies,
headaches have been reported at inhaled CO2 con-
centrations similar to those found in this investiga-
tion. This is consistent with one study which found
that 37% of healthcare workers surveyed reported
headaches following FFR use (Lim et al., 2006).
A striking unanticipated finding among the hori-
zontal flat-fold FFRs was a reduction in the average
inhaled CO2 concentration when an SM was applied
as an additional layer of protection at
VO2 of 1.0
and 1.5 l·min−1 (Table 3). The high average inhaled
CO2 concentrations among horizontal flat-fold FFRs
without an SM were caused by the larger respirator
dead space as compared with the cup type FFR or
other flat-fold FFR models. The effective respirator
dead space was increased by inflating the horizon-
tal flat-fold FFRs during exhalation. The application
of the horizontal flat-fold type of SM—a glove-to-
hand sleeve over the horizontal flat-fold FFRs—
restricted the inflation effect during exhalation and
reduced the effect of an increased dead space. The
average inhaled CO2 concentration among cup FFRs
increased among those with an SM as compared
with those without the SM at
VO2 of 0.5 l·min−1,
2.5 l·min−1 and 3.0 l·min−1, due to the additional dead
space caused by the horizontal flat-fold type of SM.
For the other flat-fold FFRs, three FFR models were
of the tri-fold type and three FFR models were of the
by guest on August 8, 2016http://annhyg.oxfordjournals.org/Downloaded from
394 E. J. Sinkule, J. B. Powell and F. L. Goss
vertical flat-fold type. The orientation of these other
flat-fold types on the user’s face and other dead space
features would affect the average inhaled CO2 con-
centrations. With an SM, the orientation of the other
flat-fold FFRs (vertical flat-fold and tri-fold) would
change on the user’s face. Placing a horizontal flat-
fold SM on a vertical flat-fold FFR would require
bending the corners/ends of the vertical flat-fold
FFR, which probably would decrease the dead space
within the FFR. Bending of the folds in tri-folded
FFR would be needed with the application of a hori-
zontally flat-folded SM, which also would reduce the
dead space in the FFR. Any bending or folding of
the FFR filter material also compromises the total
surface area of the filter media and f iltration effi-
ciency. The variability of the average inhaled CO2
concentration results among the various types of flat-
fold FFR combined in the “other flat-fold” category
contributed to insignificant differences between this
type of FFR with and without SM because this cat-
egory contained only six models between two types
of FFR (vertical flat-fold and tri-fold).
The changes in average inhaled O2 concentration
closely followed the expected reciprocal displacement
by average inhaled CO2, whereas average inhaled O2
concentration increased in conditions where average
inhaled CO2 concentration decreased and vice versa.
One reason for the changes in average inhaled O2
concentration relative to average inhaled CO2 concen-
tration is because of the relative displacement of the
gases in air; the changes in one gas directly allows for
a greater or lesser proportion of the other gases. Like
the unanticipated change that occurred among the hor-
izontal flat-fold FFRs, where a reduction in the aver-
age inhaled CO2 concentration was observed when an
SM was applied as an additional layer of protection
at
VO2 of 1.0 l·min−1 and 1.5 l·min−1, an increase in
the average inhaled O2 concentration also occurred
for this select subset of FFRs. According to CFR 42
Part 84, a hazardous atmosphere includes any oxygen-
deficient atmosphere of less than a partial pressure
of 148 mmHg or 19.5% O2 (Approval of Respiratory
Protection Devices 2011). From Table 4, the average
inhaled O2 concentrations were below 19.5% for all
conditions and all levels of energy expenditure, except
for the condition of “N95 only” at the level of O2
consumption of 2.5 l·min−1. The average inhaled O2
concentration of ≤15% occurred in one FFR without
SM during the 0.5 l·min−1 and 1.5 l·min−1 levels of
energy expenditure. In a clinical trial, inhaled oxygen
concentration of 15% caused more time needed to
travel a standard distance with the lowest power out-
put measured and coincided with the highest measured
capillary blood lactate concentrations when compared
with normoxia and hyperoxia (FIO2, 100%); caused
the recruitment of specific muscle fiber types (reduc-
ing fatigue-resistant type I fibers and increasing type
II fibers), muscle fatigue, reduced release of calcium
ion from the sarcoplasmic reticulum, increased min-
ute ventilation by 26%, and decrease O2 consumption
by 10% (Amann et al., 2006). At the threshold par-
tial pressure of O2 at 132 mmHg (17.4% O2), symp-
toms include headache, lightheadedness, drowsiness,
muscular weakness, dyspnea on exertion, nausea, and
vomiting (Schulte, 1964). Neurological symptoms,
such as reduced memory, mental work capacity, audi-
tory and visual disturbances, vertigo, tinnitus, and irri-
tability, may be manifested if O2 deficiency continues
(Schulte, 1964).
Future research may consider human subject test-
ing of various FFR models, adjusted for age and gen-
der, while measuring volume-weighted mean inhaled
CO2 and O2 gas concentrations and comparing the
responses to average inhaled CO2 and O2 gas con-
centrations from the ABMS. This proposed research
may provide a connection of ABMS results with
human subject responses for use in the development
of an ABMS-based standard test procedure for eval-
uating negative-pressure air-purifying respiratory
protective devices. Certain special groups also may
benefit from exploratory research using the ABMS
to evaluate respiratory protection, e.g. children that
use respiratory protection while performing activi-
ties in the agricultural industry.
Breathing pressures
The peak inhalation and exhalation pressures
could impact respirator comfort, in addition to inha-
lation and exhalation temperatures, respirator weight,
respirator valves, etc. The increased pressure may
cause a decrease in respiratory rate (Harber et al.,
1982; Louhevaara, 1984) and tidal volume (Harber
et al., 1982). Among older individuals, respiratory
rate may not change and tidal volume decreases with
increased inspiratory resistance (Louhevaara, 1984).
Tables 5 (Peak exhalation pressures) and 6 (Peak
inhalation pressures) show how the breathing pres-
sures increased with energy expenditure, where res-
piratory rate and tidal volume caused more air flow
during inhalation and exhalation. During exhalation,
the differences between FFRs with and without SM
occur only in cup type of FFRs. The difference in
the group with the largest representation (cup type
with 60% of the sample) would explain the variation.
In a previous NIOSH investigation of FFR breath-
ing pressures with and without SM using a breath-
ing machine, mean FFR alone and without EVs
(three models) at minute ventilations of 25 l·min−1
by guest on August 8, 2016http://annhyg.oxfordjournals.org/Downloaded from
Evaluation of N95 respirator use with a surgical mask cover 395
and 40 l·min−1 with a sinusoidal breathing wave-
form reported exhalation pressures of 7 mmH2O
and 11 mmH2O, respectively (Vojtko et al., 2008).
From the same report, the mean FFR with an SM
at minute ventilations of 25 l·min−1 and 40 l·min−1
and sinusoidal breathing waveform reported mean
exhalation pressures of 8 mmH2O and 12 mmH2O,
respectively. In FFR with EV (one model), Vojtko
reported 4 mmH2O and 5 mmH2O, respectively, at 25
l·min−1 and 40 l·min−1 for FFR alone; and, 4 mmH2O
and 6 mmH2O, respectively, for FFRs with SM. The
most significant factors contributing to the differ-
ences between the data reported from this study and
the Vojtko study could be due to a larger sample size
(30 FFR models versus 4 FFR models) and the dif-
ference in the minute ventilation values expressed by
the Vojtko study (atmospheric temperature and pres-
sure (ATP), ambient) and the present study (body
temperature (37°C), ambient pressure, and saturated
with water vapor (BTPS)). The conversion from ATP
to BTPS (used from the ABMS) would change min-
ute ventilation from 25 l·min−1 to 27 l·min−1 and from
40 l·min−1 to 44 l·min−1 (Cotes et al., 2006). Thus, the
BTPS-adjusted minute ventilations from the manikin
data used in the Vojtko et al. (2008) study are lower
than the similar minute ventilations of the ABMS,
30.6 l·min−1 and 45.6 l·min−1, respectively (Table 2).
The comparisons of pressures between FFRs that
were paired models with and without EVs were of
significant importance. First, the differences (deltas)
in average inhaled CO2 (Figure 1) and O2 (Figure 2)
between FFRs with EVs (FFR+EV) and FFRs without
EVs (FFR−EV) were largest at the middle and high
levels of energy expenditure, indicating the EVs were
remaining closed at the lowest levels of energy expend-
iture. With the SM, the delta values in average inhaled
CO2 (Figure 1) and O2 (Figure 2) between FFR+EV
and FFR−EV were consistently low among all levels
of energy expenditure, suggesting that the insufficient
pressure difference between the inside of the FFR and
the intra-mask space of the FFR and SM prevented the
opening of the EVs. Others have reported no differ-
ences in heart rate, breathing rate, tidal volume, minute
volume, transcutaneous carbon dioxide, and oxygen
saturation from human participants wearing FFR+EV
and FFR−EV, with and without SM, at
VO2 between
0.6 l·min−1 and 0.8 l·min−1 which also suggest the pres-
sure differences between the inside and outside of the
FFRs were insufficient to open the EVs both with and
without SM (Roberge et al., 2010).
Second, the exhaled and inhaled pressure results
for the paired FFRs were similar to the overall group
results, except at
VO2= 1.5 l·min−1 where the sig-
nificant effect (P < 0.05) of SM use with an EV
was not present among the paired FFRs and a sig-
nificant effect of SM use occurred only among the
paired FFRs. The homogeneous characteristics of the
paired FFRs would limit variability of various cup
depths on the user’s face. When homogeneous sizes
were procured for this investigation, the majority
of the sizes (77%) were universal, or one-size-fits-
all. Size specifications for FFRs are determined by
manufacturers. The medium size, or universal size,
for one manufacturer may not be equivalent to that of
another manufacturer. Furthermore, FFR sizes may
not be equivalent among various models from the
same manufacturer.
The various universal/medium-sized FFRs were
positioned on the ABMS head form by the same
person for all tests. Landmarks on the head form
(for example, eye, nose, chin locations) were used
to position the FFR and SM, per the manufacturer’s
instructions. Although uniformity of FFR position-
ing was maintained, the various depths and the vari-
ous types of FFRs positioned on the same head form
produced variations in dead space from the FFRs.
Clinically, it would be important to know when
humans find the added pressure from FFRs wear
intolerable or the point where users detect the added
pressure from an SM. Two reports investigated the
minimal pressures that can be detected in humans
from elastic and non-elastic loads (Campbell et al.,
1961; Bennett et al., 1962). Bennett et al. (1962)
conducted a study using added restrictive loads to
measure the ability to determine the lowest restric-
tion noticeable by humans. Participants were asked
to breathe (assuming inhalation and exhalation
were weighted equally) through progressively nar-
rowed calibrated tubes (between 2 mmH2O·l−1·s−1
and 12 mmH2O·l−1·s−1). The mean 50% level of
detection was 6 mmH2O·l−1·s−1 (BTPS). Bennett
reported a non-linear relationship between the pres-
sure and flow characteristics for each load. The
relationship between the results from Bennett et al.
(1962) and the ranges of pressures in Table 8, the
mean exhalation pressures (Table 5), and the mean
inhalation pressures (Table 6), can be used to esti-
mate the level of energy expenditure where an SM
addition to using an FFR is detected by humans.
During exhalation (Table 5), the difference in pres-
sures at the energy expenditure commensurate with
the flow rate in the Bennett study (at
VO2 between
0.5 l·min−1 and 1.0 l·min−1 by the ABMS) between
FFRs and FFR+SM were smaller than the 50%
level of detection for each flow. The same compari-
son analysis among the FFRs and FFR+SM dur-
ing inhalation (Table 6) also demonstrate that the
difference in pressures were smaller than the 50%
by guest on August 8, 2016http://annhyg.oxfordjournals.org/Downloaded from
396 E. J. Sinkule, J. B. Powell and F. L. Goss
level of detection for the flow at
VO2 between 0.5
l·min−1 and 1.0 l·min−1. These results suggest that
the increased pressures resulting from the addition
of the SM at the lower levels of energy expendi-
ture used in this investigation would not be detected
in humans as compared with using the same FFR
without an SM. These are the same levels of energy
expenditure that occur with a significant portion of
activities conducted by healthcare workers.
STUDY LIMITATIONS
Although the ABMS is an accurate, reproducible,
functional, and useful tool to characterize the meta-
bolic responses that can be produced by the use of
respiratory protection, there are limitations to its use.
For negative pressure respiratory protection, such as
FFRs, elastomeric air-purifying respirators, and gas
masks, the ABMS measurements for the respira-
tor’s dead space are affected primarily by the minute
ventilation, more specifically, tidal volume. As the
normal user’s tidal volume decreases, the effect from
respirator dead space becomes greater. Conversely,
the opposite occurs as tidal volume increases, such as
that in normal larger persons and exercise. In a field
study, smaller healthcare workers (e.g. women) were
more probable to experience intolerance for wearing
FFRs before the end of the shift (Radonovich et al.,
2009). The limitation, therefore, is characterizing
respiratory protection with a tidal volume specific
to the human data used to program the metabolic
parameters of the ABMS, or a subset of subjects with
a body size of 85–92 kg.
Another limitation for the ABMS is that it does
not respond, that is, respiratory protection for the
ABMS does not cause changes in breathing times,
breathing volumes/depths, or breathing rhythmic-
ity. Humans respond to the changes in the breath-
ing zone from the use of respiratory protection.
However, those stimuli produced by the results
of using respiratory protection are masked by the
human response. The human response was similar
to the ABMS measurements in a previous inves-
tigation (Sinkule and Turner, 2004). The stimuli
from using various forms of respiratory protec-
tion, or types of FFRs and treatments affecting
FFRs (e.g. SM), will vary in magnitude. These
effects were characterized in this investigation.
Some human participants are hyposensitive to
CO2 and metabolic acidosis, and do not respond
normally to increased CO2 concentrations until
hyperventilation occurs at exhaustive workloads
(Whipp et al., 1989).
CONCLUSIONS
Approximately 10% of commercially available
NIOSH-approved FFR models were examined with
and without SM using the ABMS to characterize
metabolic responses in an attempt to understand the
implications of the recommendation to apply an SM
over the FFR to extend the respirator’s useful life for
healthcare workers. Conclusions for this investiga-
tion include the following:
1. generally, average inhaled CO2 decreased and
average inhaled O2 increased with increasing
oxygen consumption in FFRs and FFRs with SM;
2. peak exhalation pressure and peak inhalation
pressure increased with increasing oxygen con-
sumption, but more so in FFRs with SM;
3. as compared with FFRs without SM, higher
average inhaled CO2 were observed in four of six
workloads among FFRs with SM;
4. the addition of the SM to horizontal flat-fold
FFRs at
VO2 of 1.0 l·min−1 and 1.5 l·min−1
caused a reduction in average inhaled CO2 and an
increase in average inhaled O2 due to the effects
of the (horizontal flat-fold) SM on the FFR dead
space;
Table 8. Representative peak flow and peak pressure ranges at each level of energy expenditure (n = 30).
Oxygen
consumption,
(l·min−1, STPD)
Approx. peak flow
(l·min−1, BTPS)
Peak exhalation pressure
range for (mmH2O)
Peak inhalation pressure
range for (mmH2O)
FFR only FFR + SM FR only FFR + SM
0.5 45.0 3–12 4–14 −2 to −8 −4 to −10
1.0 96.6 6–15 7–17 −8 to −19 −14 to −23
1.5 149.5 8–23 10–26 −13 to −31 −14 to −37
2.0 243.8 13–40 18–45 −25 to −55 −26 to −68
2.5 209.3 exhaled;
241.5 inhaled
11–34 16–38 −25 to −55 −26 to −68
3.0 246.1 exhaled;
276.0 inhaled
13–41 18–45 −30 to −67 −31 to −90
by guest on August 8, 2016http://annhyg.oxfordjournals.org/Downloaded from
Evaluation of N95 respirator use with a surgical mask cover 397
5. within matched models of FFRs with and with-
out EVs (without SM), the delta average inhaled
CO2 and inhaled O2 concentrations were low-
est at V
.O2 = 0.5 l·min−1 where the lowest flows
appeared insufficient for opening the EVs; and,
6. FFRs (matched models with and without EVs)
with an SM produced small and similar delta
average inhaled CO2 and O2 concentrations
across the spectrum of energy expenditures
suggesting an insufficient pressure differential
between the inside of the FFR and the space
between the FFRs and SM, which prevented the
EVs from opening.
At the lower levels of energy expenditure, this
investigation provided evidence to suggest that the
IOM recommendation of adding an SM over FFRs
in order to extend the daily duration of FFRs and
reduce the consumption of FFRs during a pandemic
would produce clinically small changes in inhaled
breathing gases and breathing pressures resulting
in a minimal effect on physical work performance,
and the amount and direction of change is affected
by the type of FFR and shape of the SM. In addition,
the evidence also indicates possible improvements in
inhaled breathing gases caused by the effects in the
dead space characteristics of the FFRs by the shape
of the SM. If the FFR is equipped with EVs, the evi-
dence suggests that the SM prevents the opening of
the EVs because of the pressure change in the space
between the FFRs and the SM. The ability of the SM
to change the pressure characteristics in the space
between the FFRs and SM occurred across the spec-
trum of energy expenditures.
Acknowledgements—The authors wish to express their grati-
tude to Ron Shaffer, Ph.D. (CDC/NIOSH/NPPTL, Technology
Research Branch) for sharing this idea; Kathleen M. Stabryla,
MLS (CDC Public Health Library and Information Center/
Information Reference, Research, and Education Services
Team (Pittsburgh Branch)) and her staff of professional librari-
ans; Heather Shoup, M.S. (CIGNA) for her technical expertise;
Doug Landsittel, Ph.D. (Department of Medicine, University
of Pittsburgh) and Linda J. McWilliams (CDC/NIOSH,
Office of Mine Safety & Health Research) for their statisti-
cal expertise; and Stefan H. Constable, Ph.D. (Deputy & Chief
Scientist, Human Performance Integration Dir., AFRL, Brooks
AFB), David Hostler, Ph.D. (Emergency Responder Human
Performance Lab, University of Pittsburgh), and Nicholas
Kyriazi, B.S. (CDC/NIOSH/NPPTL, Technology Evaluation
Branch) for their thoughtful reviews and suggestions. Finally,
the lead author wishes to acknowledge the work with Nina
Turner, Ph.D. (CDC/NIOSH/NPPTL, Technology Research
Branch), for her mentorship and the many hours of her exper-
tise with the ABMS prior to the initiation of this project.
Disclaimer—The findings and conclusions in this paper
are those of the authors and do not necessarily represent the
views of National Personal Protective Technology Laboratory
(NPPTL) or Centers for Disease Control and Prevention
(CDC)/National Institute for Occupational Safety and Health
(NIOSH). Mention of commercial products or trade names
does not constitute endorsement by NPPTL or CDC/NIOSH.
The authors do not have any financial interests in the present
research.
REFERENCES
Ainsworth BE, Haskell WL, Whitt MC et al. (2000) Compendium
of physical activities: an update of activity codes and MET
intensities. Med Sci Sports Exer; 32: S498–516.
Amann M, Romer LM, Pegelow DF et al. (2006) Effects of
arterial oxygen content on peripheral locomotor muscle
fatigue. J Appl Physiol; 101: 119–27.
American Conference of Governmental Industrial Hygienists.
(2008) Occupational exposure values. Guide to occupational
exposure values. Cincinnati, OH: American Conference of
Governmental Industrial Hygienists.
Approval of Respiratory Protection Devices. (2006) 42 CFR
§ 84.97 Test for carbon dioxide in inspired gas; open- and
closed-circuit apparatus; maximal allowable limits.
Approval of Respiratory Protection Devices. (2011) 42 CFR §
84.2(y) Definitions.
Bennett ED, Jayson MI, Rubenstein D et al. (1962) The ability
of man to detect added non-elastic loads to breathing. Clin
Sci; 23: 155–62.
Campbell EJ, Freedman S, Smith PS et al. (1961) The ability
of man to detect added elastic loads to breathing. Clin Sci;
20: 223–31.
Checkoway H, Pearce N, Kriebel D. (2004) Characterizing the
workplace environment. Research methods in occupational
epidemiology. New York, NY: Oxford University Press, Inc.
ISBN 0-19-509242-2.
Compressed Gas Association. (1999) Carbon dioxide. Norwell,
MA: Kluwer Academic Publishers.
Consolazio WV, Fisher MB, Pace N et al. (1947) Effects on
man of high concentrations of carbon dioxide in relation to
various oxygen pressures during exposures as long as 72
hours. Am J Physiol; 151: 479–503.
Cotes JE, Chinn DJ, Miller MR. (2006) Basic terminology
and gas laws. Lung function: physiology, measurement
and application in medicine. Malden, MA: Blackwell
Publishing, Inc. ISBN 0-6320-6493-9.
Deno NS. (1984) Automatic breathing and metabolic simula-
tor: the respiring robot. J Int Soc Respir Protect; 2: 38–51.
Harber P, Tamimie RJ, Bhattacharya A et al. (1982) Physiologic
effects of respirator dead space and resistance loading. J
Occup Med; 24: 681–9.
Hogan MC, Cox RH, Welch HG. (1983) Lactate accumulation
during incremental exercise with varied inspired oxygen
fractions. J Appl Physiol; 55: 1134–40.
Institute of Medicine, Committee on the Development of
Reusable Facemasks for Use During an Influenza Pandemic.
(2006) Use and reuse of respiratory protective devices for
influenza control. Reusability of facemasks during an
influenza pandemic: facing the flu. Washington, D.C.: The
National Academies Press. ISBN: 0-309-66000-9.
Kyriazi N. (1986) Development of an automated breathing and
metabolic simulator. Pittsburgh, PA: U.S. Department of the
Interior, Bureau of Mines, IC 9110.
Li Y, Tokura H, Guo YP et al. (2005) Effects of wearing N95 and
surgical facemasks on heart rate, thermal stress and subjective
sensations. Int Arch Occup Environ Health; 78: 501–9.
by guest on August 8, 2016http://annhyg.oxfordjournals.org/Downloaded from
398 E. J. Sinkule, J. B. Powell and F. L. Goss
Lim ECH, Seet RCS, Lee KH et al. (2006) Headaches and the
N95 face-mask amongst healthcare providers. Acta Neurol
Scand; 113: 199–202.
Louhevaara VA. (1984) Physiological-effects associated with
the use of respiratory protective devices – a review. Scand J
Work Environ Health; 10: 275–81.
National Institute for Occupational Safety and Health. Certified
equipment list. Available from http://www.cdc.gov/niosh/npptl/
topics/respirators/cel/default.html (accessed 13 May 2011).
National Institute for Occupational Safety and Health. CET-
APRS-STP-CBRN-0454: Determination of Human Subject
Breathing Gas (HSBG) Concentrations (Carbon Dioxide
and Oxygen) for Chemical, Biological, Radiological and
Nuclear (CBRN) Air-Purifying Escape Respirator. Available
from http://www.cdc.gov/niosh/npptl/stps/pdfs/CET-APRS-
STP-CBRN-0454.pdf (accessed 25 August 2006).
Nieuwenhuijsen MJ. (2006) Introduction to exposure assess-
ment. In Nieuwenhuijsen, MJ, editor. Exposure assessment
in occupational and environmental epidemiology. New York,
NY: Oxford University Press, Inc. ISBN 0-19-852861-2.
Patterson JL, Jr., Heyman A, Battey LL et al. (1955) Threshold
of response of the cerebral vessels of man to increase in
blood carbon dioxide. J Clin Invest; 34: 1857–64.
Radonovich LJ, Cheng J, Shenal BV et al. (2009) Respirator
tolerance in health care workers. J Am Med Assoc; 301:
36–8.
Raven PB, Dodson AT, Davis TO. (1979) The physiological
consequences of wearing industrial respirators: a review.
Am Ind Hyg Assoc J; 40: 517–34.
Roberge RJ. (2008) Effect of surgical masks worn concurrently
over N95 filtering facepiece respirators: extended service
life versus increased user burden. J Public Health Manag
Pract; 14: E19–26.
Roberge RJ, Coca A, Williams WJ et al. (2010) Surgical
mask placement over N95 filtering facepiece respirators:
physiological effects on healthcare workers. Respirology;
15: 516–21.
Sayers JA, Smith RE, Holland RL et al. (1987) Effects of car-
bon dioxide on mental performance. J Appl Physiol; 63:
25–30.
Schneider EC, Truesdale D. (1922) The effects on the circu-
lation and respiration of an increase in the carbon dioxide
content of the blood in man. Am J Physiol; 63: 155–75.
Schulte JH. (1964) Sealed environments in relation to health
and disease. Arch Environ Health; 8: 438–52.
Sinkule E, Turner N, Hota S. (2003) Automated breathing and
metabolic simulator (ABMS) CO2 test for powered and
non-powered air-purifying respirators, airline respirators,
and gas masks. American Industrial Hygiene Conference
and Exposition (Abstracts); 54.
Sinkule EJ, Turner NL. (2004) Inhaled carbon dioxide and
oxygen concentrations during rest and exercise of three air-
purifying escape hoods. Med Sci Sports Exer; 36: S245.
U.S. Department of Health and Human Services. (2009)
Authorization of emergency use of certain per-
sonal respiratory protection devices; Availability, 74
Federal Register 38644. Available from http://www.
federalregister.gov/articles/2009/08/04/E9-18570/
authorization-of-emergency-use-of-certain-personal-
respiratory-protection-devices-availability#h-9
(accessed 2 November 2010).
Vojtko MR, Roberge MR, Vojtko RJ et al. (2008) Effect on
breathing resistance of a surgical mask worn over a N95 fil-
tering facepiece respirator. J Int Soc Respir Protect; 25: 1–8.
Whipp BJ, Davis JA, Wasserman K. (1989) Ventilatory control
of the 'isocapnic buffering' region in rapidly-incremental
exercise. Respir Physiol; 76: 357–67.
Yang Y, Sun C, Sun M. (1997) The effect of moderately
increased CO2 concentration on perception of coherent
motion. Aviat Space Environ Med; 68: 187–91.
by guest on August 8, 2016http://annhyg.oxfordjournals.org/Downloaded from
