EPA-650/1-74-001
July 1973
Environmental Health Effects Research Series
-------
-------
EPA-650/1-74-001
THE USE OF PANELISTS
AS SUBSTITUTES FOR TAXICAB DRIVERS
IN CARBON MONOXIDE EXPOSURE
by
A. Walter Hoover, M. D.
and Robert M. Albrecht, M. D.
Division of Environmental Health Sciences
School of Public Health
Columbia University
New York, N. Y. 10032
Contract No. CPA 22-69-97
Program Element No. 1AA005
EPA Project Officer: J.H.Knelson
Human Studies Laboratory
National Environmental Research Center
Research Triangle Park, North Carolina 27711
Prepared for
COORDINATING RESEARCH COUNCIL, INC .
30 ROCKEFELLER PLAZA
NEW YORK, N.Y. 10020
Project No. CAMP-8-68(l-68)
and
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
July 1973
-------
This report has been reviewed by the Environmental Protection Agency
and approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the Agency,
nor does mention of trade names or commercial products constitute
endorsement or recommendation for use.
11
-------
Acknowledgments
The authors express their appreciation
for the dedicated interest and efforts of
the Division of Environmental Health
Sciences' staff, who obtained the samples,
performed the laboratory analyses, and
provided constructive suggestions at every
stage. These persons included Mr. Richard
Hanauer, Mrs. Elaine Nelson, and Mrs.
Maria Graham.
ill
-------
-------
TABLE OF CONTENTS
Methodology page 1
Results 5
Discussion 11
Conclusions 14
References 15
Appendix A - "Tables and Figures" 17
Appendix B - "Statistical Correlation of Aveolar
Air Carbon Monoxide Concentrations
and Carboxyhemoglobin using Spear-
man Rank Correlation Test" 27
Appendix C - "Data on COHb by Finger Prick
and Alveolar Air COi' 29
V
-------
-------
REPORT
on
The Use of Panelists as Substitutes for Taacicab Drivers in
Carbon Monoxide Exposure Studies
CRC-APRAC Project No. CAPM-8-68 (1-68)
Summary
Analyses of breath and limited blood samples from 30 pairs of taxi
drivers and panelists who drove in New York City traffic fo r 8 hours on two
consecutive days indicated that both panelists and drivers attained similar
COHb (blood carboxyhemoglobin) levels. This was true for both smokers and
non-smokers though smokers had significantly higher concentrations of COHb
than non-smokers. There was no consistent difference between the first and
second day of driving in the levels of alveolar carbon monoxide.
Methodology
Background
Data for this study were collected between June 27, 1969 and November
15, 1970. There were four objectives:
a) To determine whether panelists attain concentrations of
COHb (blood carboxyhemoglobin) similar to those attained
by taxicab drivers with equal exposure.
b) To determine if the levels of COHb for smokers differed
from those of non-smokers.
-------
2.
c) To determine the relationship between concentrations of
carbon monoxide (CO) inside an automobile and the levels
of COHb.
d) To compare the levels of COHb in blood obtained by finger
prick with the level of carbon monoxide in the expired alveolar
air after 20 seconds of breath holding.
The Subjects
Sixty individuals were divided into 30 pairs, each pair consisting of a
taxicab driver and a panelist. Half of the pairs were current cigarette smokers
while the other half had never smoked tobacco in any form or had not smoked
within the previous year. The smoking history is described in table I. All
were males, and pairs were matched by age within 6 years with the exception
of 2 pairs. (See table II).
Before the study, subjects received a complete physical examination
that included a chest x-ray, electrocardiogram, test of 1-second forced
expiratory volume, hemoglobin test, and other indicators of physical con-
dition. Prospective subjects with evidence of pulmonary disease, cardiac
disease, blood dyscrasias, or any other disease that might affect normal cardio-
pulmonary physiology were excluded from the study. In addition, subjects who
were considered incapable of properly providing samples of expired alveolar
air were excluded to facilitate the study.
Materials and^Methods
With his panelist seated beside him, each driver drove through heavily
-------
travelled streets in Manhattan and Brooklyn for two successive days. All
drivers used the same automobile, a rented 1969 Dodge Dart Sedan. It was
not air conditioned, and subjects were permitted to open and close the car
windows freely.
A member of the study team accompanied the test vehicle, and he
collected ambient air and alveolar air specimens while seated in the back
seat. The driving schedule extended from 8:00 a.m. to 5:00 p.m. with a
coffee break around 10:00 a.m. and lunch from noon to 1:00 p.m. After
their dinner, subjects were requested to return for a final alveolar air
specimen at 8:00 p.m. Tests -were scheduled as follows:
Hours Alveolar Air Ambient Air
X
X
X
X
X
X
X
X
X*
*Ambient levels in the laboratory where the specimens were obtained.
Range was constant bet-ween 5-10 parts per million CO.
Breath and air samples were collected in bags of aluminum foil covered
by polyester film and equipped with valved closures. When filled, the bags were
shown to lose less than 10% of their volume after three days, and they withstood
shipping.
0800
0900
1000
1100
1200
1300
1400
1500
1600
1700
2000
X
X
X
X
X
X
-------
4.
To provide a breath sample, a subject exhaled completely and then
inhaled rapidly. After holding his breath for 20 seconds while being timed
by a stop-watch, he discarded the first 1400 cc. of breath and the remaining
portion was collected and analyzed. Ambient air was pumped into a 7-liter
bag over a 15-minute period. Alveolar and ambient air specimens were an-
alyzed on the same day using a Hilger and Watts Infra-Red Gas Analyzer,
type S. C. /L. C. This is a single-range instrument with a linear response
and a range -- for the purpose of this study -- of 0-200 ppm of carbon mon-
oxide. Carbon dioxide and water vapor are removed prior to analysis by a
drying tube containing soda lime and magnesium perchlorate.
While driving, smokers were allowed to smoke at will except when
specimens were taken. Since the opening of windows would affect panelists
and drivers equally, it was not considered to affect the comparability of
carbon monoxide levels between panelists and drivers. The number and time
of cigarettes smoked would affect COHb but the study was designed to measure
the comparability of carbon monoxide in smoking panelists and drivers rather
than the possible reasons for differences in carbon monoxide levels.
Determination of Carboxyhemoglobin
Determination of carboxyhemoglobin levels were made by analyzing
finger-prick blood and using a procedure that was a slight modification of
one described by B. T. Commins and P. J. Lawther (1). About 0. 025 ml. of
blood from a finger prick was dissolved in 25 ml. of 0. 04% ammonia solution.
This solution was divided in half, and oxygen was bubbled through one half to
convert any carboxyhemoglobin into oxyhemoglobin. The spectra of the two
-------
5.
halves was then compared in a spectrophotometer. The difference between the
spectra was used to estimate the carboxyhemoglobin content of the blood and
was compared with the difference between the spectra of pure oxyhemoglobin
and pure carboxyhemoglobin samples of the same blood. The formula for the
percent of carboxyhemoglobin was:
_h_
% COHb = h0
Where h = optical height above the mean of the absorbances at 414 mu (b) and
426 mu (c) with a peak at 420 mu (a). A value was obtained by h = 2 - (b+x)/2.
h was obtained by the same procedure from the two pure samples.
Results
Ambient Air Concentrations of Carbon Monoxide
As described in Table 3, the carbon monoxide levels varied from 2 ppm
to 48 ppm. The average reading for each hour, however, varied from a mini-
mum of 16. 1 ppm of carbon monoxide at 1000 to 23. 7 ppm. at 1500. Comparing
the carbon monoxide levels in the cars of non-smokers to smokers, the averages
were the same at 1100 and the maximum difference between averages occurred
at 1500 when levels in non-smokers' air reached an average of 3. 7 ppm higher.
These concentrations are generally similar to those recorded by Jaffe
(2) for the commuter exposure in New York City. Using an automatic infra-red
analyzer and 20-30 minute integrated samples, he found the following concen-
trations in the "Center City" in 1966 and 1967:
Maximum: 52 and 48 ppm
Average: 32 and 27 ppm
Minimum: 14 and 9 ppm
Values in 14 other American cities: 8 to 70 ppm
-------
6.
Johnson, Dworetzky and Heller (3) recorded carbon monoxide levels
continuously within a trailer parked in midtown Manhattan during 1967. The
median hourly concentration on week days was 12-13 ppm. Colucci and Begeman
(4) found median hourly averages of 7. 5 and 10. 5 ppm on week days at two sites
in midtown Manhattan during 1962-64.
Comparison of Drivers and Panelists
As shown by Figure 1, the average hourly alveolar air measurements for
drivers and panelists showed that carbon monoxide concentrations for the two
groups varied almost identically for non-smokers. For smokers, slight
differences can be discerned. This lack of conformity for smokers can be
attributed to the broad range that carbon monoxide measurements spanned,
varying from below 25 ppm to over 45 ppm. By contrast, the average carbon
monoxide readings for non-smokers had a minimum of about 12 ppm and a
maximum of less than 20. When each day was investigated separately, as shown
in Figure 2, the same general observations could be made.
Comparison of Smokers and Non-Smokers
The striking contrast in the study occurred when smokers, both drivers
and panelists, were compared with non-smokers. As shown in Figure 3,
smokers consistently exhibited a considerably higher concentration of carbon
monoxide than non-smokers. The difference was greater at 5:00 p.m. in the
case of drivers and at either 3:00 p.m. among panelists, depending upon
whether the first or second day is being considered. However, the gap already
existed at 8:00 a.m. and remained through 8:00 p.m. three hours after driving
had ceased.
-------
7.
Interestingly, the level of carbon monoxide in non-smokers at 8:00 a. m.
was slightly lower among panelists than for taxicab drivers. This raised the
question whether the groups differed in exposure at home or on the way to work.
An occupational effect could have existed, but this does not seem likely since
the effects of occupational exposure presumably would have worn off by the
next morning.
Analysis of Variance
All carbon monoxide readings were averaged by four categories: (1)
smoking drivers, (2) non-smoking drivers, (3) smoking panelists, and (4)
non-smoking panelists. When these categories were averaged, the difference
between drivers and panelists was -2. 9 ppm for smokers and +1. 5 ppm for non-
smokers. Statistical tests were then performed to see whether a significant
difference existed between any of the four comparable groups. Tests were
performed on the log CO concentration, and the model was a three-way in-
complete analysis of variance. But the difference between drivers and panelists
on both the smoker and non-smoker categories was not statistically significant.
Trend Over Time
As illustrated by Figure 1, the change in carbon monoxide levels among
non-smoking drivers during each day was minor. Among panelists, the
levels rose slightly to peak around 3:00 p.m. or 5:00 p.m. For smokers, on
the other hand, the increase was considerable, and the peak occurred at 5:00
p.m. A consistent difference in alveolar carbon monoxide between the first
-------
8.
and second days of driving did not appear to exist.
Comparison of Ambient Air and Alveolar Air
To determine the effect of carbon monoxide levels in ambient air upon
corresponding levels in alveolar air, correlation coefficients were calculated
using the logarithms of carbon monoxide measurements in ppm for each type
of air. Since 60 persons -- either drivers or panelists -- were involved for
two days, each correlation was based on 60 pairs of numbers. Both the
morning and daily averages of carbon monoxide levels in ambient air were
used since it seemed possible that morning values might reveal a time-lag
relationship. As shown in Table 5, however, the data did not confirm this
hypothesis, and the morning figures did not appear to have any advantage
over the figures of the entire day. In general, the correlation was stronger
with the 5:00 p.m. alveolar air carbon monoxide than with the 8:00 p.m. con-
centrations. In general, the correlation was higher among non-smokers than
smokers, a predictable observation since non-smokers had ambient air as
their only source. The maximum correlation coefficients were relatively
weak: 0. 60 and 0. 55 for non-smoking panelists and drivers respectively at
5:00 p.m.
The only other recorded study comparing ambient air within a car and
concentrations of carbon monoxide by the car's occupants was by Clayton et al
(5). Two individuals drove for 8 hours continuously in a police scout car,
presumably entirely within Detroit. One smoked cigars, and the other did
not smoke. The carbon monoxide was monitored continuously, and readings
were recorded every five minutes. Using 50 readings, the average value was
-------
17 ppm and a peak of 120 ppm occurred when the engine was idling. The
COHb rose from 0. 8 to 1. 2% in the non-smoker and 3. 1 to 3. 9% in the smoker.
The correlation coefficient for ambient carbon monoxide and COHb was not
provided.
Comparison of Finger Prick Blood and Alveolar Air
Blood was obtained by finger prick six times a day for the two-day test
from four drivers and their panelists. Half of the pairs smoked. Table 6
illustrates the correlation between COHb and the alveolar air carbon monoxide;
but, because of the small number of subjects, additional statistical evaluation
was indicated because a plot of all values for alveolar air carbon monoxide and
the simultaneous COHb on a scatter diagram indicated that the comparison did
not approximate a normal distribution. The Spearman Rank Correlation Test
was then performed. When smoking habits were disregarded, an r of 0. 5531
was found between the carbon monoxide alveolar air and the COHb readings.
A test of significance gave the T -2 as 6. 26, which indicates statistical sig-
nificance at the 1% level. The second test using non-smokers resulted in
values of r = 0.4002 (N=48) and Tn-2 = 2. 96 significant at the 1% level. The
final procedure testing only the values of smokers resulted in values of r =
0. 0795 (N=43) and Tn_2 = 0. 51 with no statistical significance indicated.
According to the article, "Post Exposure Relationship of Carbon Mon-
oxide in Blood and Expired Air" by J. E. Peterson (6), "Alveolar breath
analysis can be used to accurately estimate the percent COHb saturation in
adult white males provided the ambient carbon monixide concentration is near
zero. " In our study, a definite difference appeared to exist between
-------
10.
correlations among smokers and non-smokers, perhaps suggesting that part of
the ambient air exposure in smokers is the carbon monoxide present in the res-
piratory exchange necessary for smoking cigarettes.
*
Further study has been conducted at the request of the authors by another
*.
laboratory to correlate measurements between single samples of expired alveolar
air concentrations of carbon monoxide and single finger prick blood specimens as
•well as venous blood specimens for determination of COHb. Carbon monoxide
concentrations of below 2 parts per million were maintained in the ambient
atmosphere. The only major variable was the smoking habit of the individual.
Under no circumstances were non-smokers tested in the same laboratory where
a smoker had previously been tested. Subjects consisted of five adult males
aged 22, 25, 30 and 34 years with smoking histories of at least two packs daily
during recent years. Examinations and medical records indicated no demonstrable
physical disease. Using the Spearman Rank Correlation, the relationship between
expired alveolar air carbon monoxide and finger prick COHb determinations was
found to be r = 0. 568, which indicates no statistical significance.
Then six 1 -pack-per-day male smokers of ages 22, 22, 35, 44, 38, and
21 were tested for carbon monoxide concentrations in alveolar air and COHb
in blood. Correlations by the Spearman Rank Test were found to be almost
identical with the smoking group of the cab drivers in this study. The r was
0. 0821 and the T -2 = 6. 84, which gave statistical significance at the 1% level.
Correlations between the finger prick and the venous blood methods indicated
a correlation coefficient of 0. 87, which was considered to be highly significant.
-------
11.
Discussion
Use of Panelists as Substitutes for Taxicab Drivers
In general, this study found that panelists attained concentrations of
COHb similar to those attained by taxicab drivers with equal exposure. Some
questions remain though, on some points, such as the generally higher average
carbon monoxide readings in alveolar air for panelists in the smoking category.
But to investigate this subject further, more information would have to be ob-
tained. Contributing factors could be the air-pollution concentrations at the
residence of subjects, the method of transportation from home to the study site,
and the length of time for travel from home to the site of the study. Seasoning
(i. e. , long or heavy experience in driving) and age may possibly affect levels
of carbon monoxide in subjects.
Nelson and Hasselblad (7) examined four taxi drivers and four passengers
at 9:00 a.m., 11:00 a.m. , 1:00 p. m. and 3:00 p. m. on two successive days.
The design paired one non-smoking driver with one non-smoking passenger
and one smoking driver with one smoking passenger. While no gross differ-
ences between drivers and passengers were observed, the sample was ob-
viously minimal in size. This seems to be the only published study comparing
drivers and passengers.
Smokers vs. Non-Smokers
There is abundant evidence that cigarette smoking has a more potent
effect on carbon monoxide concentrations than ambient city air, and the
-------
12,
ambient air concentrations in this study were not extraordinary. But is there
interaction between smoking and ambient air? Our data suggest that there may
•*
be. The ambient air carbon monoxide has only a small effect alone. But the
«
large rise for smokers, as shown in Figure 1, could be due to cigarette smoking >
alone, except that the sharp fall between 5:00 p.m. when driving stopped and
8:00 p.m. implies a significant effect of ambient air. Of course, alternative
explanations for this drop could also be found.
Bartlett (8) stressed the point that carbon monoxide from smoking and
carbon monoxide from ambient air are not additive. If one has a high level of
COHb from smoking, and exposure to a low level of carbon monoxide in ambient
air will not increase the COHb. But Bartlett adds, "this conclusion is modified
by the fact that smokers' carbon monoxide excretion bet-ween cigarettes is
slower in a carbon monoxide polluted environment than in pure air. Thus,
their long-term average COHb concentrations are slightly higher in the presence
of environmental carbon monoxide than in its absence. "
Ambient Air and COHb
The relationship between ambient carbon monoxide concentration -within
the automobile and COHb is positive but weak. But this correlation is inherently
different to ascertain for subjects driving an automobile and is more suitably
investigated in the rigidly controlled conditions obtainable in a laboratory
chamber. We were not able to operate the infra-red analyzer in the car becausf
the shocks and jolts could not be softened sufficiently. We then used bag
-------
13.
samples integrated over 15 minutes, taking 8 samples over the 9-hour day.
In view of the potentially radically different concentrations encountered in
driving, even from one block to another, and the fact that subjects escaped
exposure to high levels of ambient carbon monoxide during the one hour lunch
period, it is not surprising that the correlation was not great.
Alveolar Air, Finger Prick and Venipuncture Blood
The correlation between carbon monoxide in alveolar air and the COBh
in finger prick blood was -weak. But a significant correlation can only be
expected -where the ambient air concentrations of carbon monoxide are
relatively low; and, in this study, the ambient carbon monoxide concentrations
were elevated in all collections of expired air with the exception of the evening
specimen at 8:00 p.m.
Another explanation for the insignificant correlation is technical. In
any micro method using finger puncture techniques, the possibility of dilution
by tissue juices is always regarded as a valid technical disclaimer. Additional
venous blood and finger tip comparisons were conducted by one of the authors
(A.W.H. ), and the significant correlation (0.82) between COHb as determined
by finger prick and by venous blood methods was obtained for 15 subjects,
seven smokers and eight non-smokers. Because the test subjects for the
panelist study were in a vehicle, venipuncture was not feasible and the micro
finger prick method had to be used.
-------
14.
Conclusions
The data of this study indicate that panelists are a satisfactory sub-
stitute for taxi drivers in field studies of carbon monoxide concentrations.
Use of alveolar air to indicate the probable levels of COHb, however, did
not prove to be a viable assumption since CO concentrations in alveolar air
did not correlate acceptably with COHb under the conditions of this study.
More satisfactory for future studies, the authors believe, will be expanded
application of the finger-prick method.
This study suggested several promising areas for future investigation.
The influence of high levels of ambient carbon monoxide on the validity of
alveolar air measurements must be further explored. Correlation of COHb
and expired air from heavy smokers should be studied. And there should be
a more intensive effort to establish a standardized method for field deter-
minations of COHb.
A. Walter Hoover, M. D.
Robert M. Albrecht, M. D.
-------
15.
References
(1) Commins, B. T. and P. J. Lawther, "A Sensitive Method for the
Determination of COHb in a Finger Prick Sample of Blood, "
British Journal of Industrial Medicine 22 (1965), p. 139-143.
(2) Jaffe, Lewis S. , "Ambient Carbon Monoxide and Its Fate in the
Atmosphere," Journal of the Air Pollution Control Association
18 (1968), p. 534-540.
(3) Johnson, Kenneth L. , L. H. Dworetzky, and Austin N. Heller,
"Carbon Monoxide and Air Pollution from Automobile Emissions
in New York City, " Science 160 (1968), p. 67-68.
(4) Colucci, Joseph M. and Charles R. Begeman, "Carbon Monoxide
in Detroit, New York and Los Angeles Air /'Environmental Science
and Technology 3 (1969), p. 41-47.
(5) Clayton, George D. , Warren A. Cook, and W. G. Frederick, "A
Study of the Relationship of Street Level Carbon Monoxide Concen-
trations to Traffic Accidents, " Traffic Safety (December I960),
p. 25-31.
(6) Peterson, J. E., "Post Exposure Relationship of Carbon Monoxide
in Blood and Expired Air," Archives of Environmental Health
(August 1970).
(7) Nelson, William C. and Victor Hasselblad, "Carbon Monoxide Levels
of Taxidrivers and Passengers, " Division of Health Effects Research,
National Air Pollution Control Administration, Durham, N. C. mimeo-
graphed, undated.
(8) Bartlett, Donald, J. R. , "Pathophysiology of Exposure to Low
Concentrations of Carbon Monoxide," Archives of Environmental
Health 16 (1968), p. 719-727.
-------
-------
17.
APPENDIX A
TABLES AND FIGURES
-------
TABLE 1
Smoking History
Smoking
Never
Ex-smokers
Currently under 1 pack of cigarettes per day
Currently 1 to 2 packs of cigarettes per day
Currently over 2 packs of cigarettes per day
Total
Drivers
5
10
4
7
4
30
Panelists
7
8
4
9
2
30
-------
TABLE 2
Age Difference Between Driver and Panelist
Number of Years Number of Pairs
0-3 21
19.
-------
20.
TABLE 3
Ambient Air Concentrations of CO in ppm by
Time of Day
Hour
0800
0900
1000
1100
1300
1400
1500
1600
MINIMUM
8
3
6
4
4
2
6
5
AVERAGE
Non-smokers Smokers
22. 5
20. 6
18.1
21. 7
18. 9
19. 2
23. 7
19. 4
19. 3
18. 8
16.1
21. 7
16. 4
19. 3
20. 0
20. 2
MAXIMUM
41
42
36
48
42
43
44
43
-------
21.
TABLE 4
Average CO Concentrations in ppm by-
Time of Day
NON-SMOKERS
8:00 A.M. 10:00 A.M. 12 NOON 3:00 P.M. 5:00 P.M. 8:00 P.M.
Drivers,
Drivers,
Day 1
Day 2
Avg. Drivers
Panelists
Panelists
, Day 1
, Day 2
Avg. Panelists
Drivers,
Drivers,
Day 1
Day 2
Avg. Drivers
Panelists
* Panelists
, Day 1
, Day 2
Avg. Panelists
18.
15.
16.5
13.
12.
12. 5
22.
23.
22. 5
29.
23.
26. 0
18.
18.
18. 0
15.
14.
14. 5
29.
26.
27. 5
29.
31.
30. 0
15.
16.
15. 5
14.
16.
15. 0
SMOKERS
30.
23.
26. 5
22.
25.
23. 5
18.
20.
19. 0
16.
21.
18. 5
38.
37.
37. 5
45.
40.
42. 5
17.
21.
19. 0
18.
21.
19. 5
41.
39
40. 0
44.
44.
44. 0
15.
16.
15. 5
14.
15.
14. 5
30.
30
30. 0
33.
36.
34. 5
-------
22.
TABLE 5
Correlation Coefficients
Correlation with:
Avg. all day
Smokers Ambient air CO
Drivers 5PM, CO 0. 38
Drivers 8PM, CO 0. 26
Panelists 5PM, CO 0. 36
Panelists 8PM, CO 0. 45
Non-Smoke rs
Drivers 5PM, CO 0.55
Drivers 8PM, CO 0. 53
Panelists 5PM, CO 0. 60
Panelists 8PM, CO 0. 17
Avg. morning
Ambient air CO
0.40
0. 34
0.29
0.40
0. 51
0.49
0. 56
0.23
-------
23.
TABLE 6
Correlation of COHb obtained by
finger-prick and alveolar air CO
Pair number Smoker Driver Panelist
1 Yes ,_15 0
2 Yes ._65 . 08
3 No J50 . 54
4 No .71 .23
-------
24.
FIGURE. 1
45
I 35
CL
UJ 25
>
UJ
O
o
AVERAGE CO READINGS
FOR SMOKERS AND NONSMOKERS
BY TIME OF DAY
(AVERAGED OVER 2 DAYS)
NONSMOKERS
.^•*"
8a.m.
45r-
10a.m. I2noon 3p.m. 5p.m.
35
CL
a.
ui
LU
O
O
25
15
-•
SMOKERS
8a.m. 10a.m. I2noon 3p.m. 5p.m.
. DRIVERS
^ PANELISTS
8p.m.
^PANELISTS
'• DRIVERS
8p.m.
-------
25.
FIGURE 2
45
1.35
Q.
25
LU
o 15
o
AVERAGE CO READINGS
FOR SMOKERS AND NONSMOKERS
BY TIME OF DAY
NONSMOKERS
8a.m. lOa.m 12 noon
45r
3p.m. 5p.m.
1.35
Q.
_i
UJ
25
O 15
o
SMOKERS
8a.m. lOa.m 12 noon 3p.m 5p.m.
PANELISTS, day I
v. DRIVERS, day 2
*" DRIVERS.day I
PANELISTS,day 2
8p.m.
•>.PANELISTS,day I
PANELISTS, day 2
^ DRIVERS.day I
DRIVERS, day 2
8p.m.
-------
26.
FIGURE 3
45
I 35
QL
UJ
25
O 15
o
AVERAGE CO READINGS
FOR DRIVERS AND PANELISTS
BY TIME OF DAY
,X DRIVERS
SMOKERS, day I
SMOKERS, day 2
...^NONSMOKERS.dayl
"NONSMOKERS,day2
8a.m. 10a.m. I2noon
45r
3p.m. 5p.m.
I 35
OL
UJ
UJ
_J
O
o
25
15
8p.m.
•SMOKERS, day 2
SMOKERS, day I
PANELISTS
NONSMOKERS,day2
8a.m. 10a.m. I2noon 3p.m. 5p.m. 8p.m.
-------
27.
APPENDIX B
STATISTICAL CORRELATION OF ALVEOLAR AIR
CARBON MONOXIDE CONCENTRATIONS AND CAR-
BOXYHEMOGLOBIN USING SPEARMAN RANK
CORRELATION TEST
-------
28.
Statistical Correlation of Alveolar Air Carbon Monoxide
Concentration and Carboxyhemoglobin Using Spearman
Rank Correlation Test
All values of the simultaneous readings of CO* and COHb* were
roughly plotted on a scatter diagram. It was found through observation that
the comparison between the two values did not approximate a normal distri-
bution. It was, therefore, decided not to accept the correlation coefficient
test since the assumption of normality of distribution must be made for the
test to be valid. Conversion of the CO and COHb values to logarithms still
left doubt as to the normalcy of distribution, although normalcy was being
approached.
Thus, the Spearman Rank Correlation Test was applied since normal
distribution of values would not be a limiting factor. Three tests were per-
formed. The first was to correlate all CO and COHb readings taken as a
single group, without regard to smoking. The formula for the test was:
2
r = 1 - —-—^2—L where r = correlation, d = difference of
n3 -n
the ranks, and n = 91. This gave a value of
r= 0. 5531. A test of significance gave t ,, = 6. 26, which indicates a statistical
significance at the 1% level.
The second test, using non-smokers only, resulted in values of
r = 0. 4002 (n = 48) and tn_2 =2. 96 significant at the 1% level. The final test
using only smokers resulted in values of r = 0. 0795 (n = 43) and t ?= 0. 51,
showing no statistical significance.
#CO concentration is measured in ppm (parts per million) and COHb in
% (per cent)
-------
APPENDIX C
DATA ON COHb BY FINGER PRICK AND ALVEOLAR AIR CO
-------
30.
Levels of COHb (%) and Alveolar Air CO (ppm) for 4 Panelists
and 4 Drivers
Non- Smokers
Time
Day 1
Day 2
Day 1
Day 2
8 a. m.
10 a.m.
12
3 p. m.
5
8
8 a. m.
10
12
3 p. m.
5
8
8 a. m.
10. 30
12. 30
3. 15
5. 30
8
8 a. m.
10
12
3
5
8
Driver A
% COHb
2. 7
2.2
2.6
4. 3
2.6
1. 5
1. 5
2. 3
5. 3
3. 5
1. 5
. 1
Driver B
2. 2
0
2. 1
3. 5
2. 9
3.6
3.8
3. 0
3.4
3.2
3.0
1. 3
CO (ppm)
16
18
21
22
21
18
20
21
19
22
17
12
15
15
12
22
18
35
22
23
23
25
22
21
Panelist A
%COHb
1. 5
7. 5
2.6
1. 8
2.8
2. 9
1. 3
3.6
2.4
2. 2
4. 7
. 1
Panelist B
2.6
2.8
2. 8
2.4
2.8
2.8
1. 1
2. 5
3. 7
3. 1
2. 3
2. 5
CO
10
10
10
20
21
11
11
14
13
18
17
7
16
20
11
16
22
23
14
18
18
19
20
17
-------
31.
Levels of COHb (%_) and Alveolar Air CO (ppm)
V
•4
Time
for
4 Panelists and 4 Drivers
Smoke
Driver C
% COHb CO (ppm)
8
9
1
2
5
8
8
10
11
2
4
8
a. m.
. 35
1.45
. 30 p.m.
a. m.
. 45
. 30
. 30
6.
8.
11
17
12
12
7.
13.
11.
13.
22.
12.
2
1
. 0
. 8
. 5
. 8
9
1
5
4
0
2
29
41
40
35
36
30
35
23
31
38
23
Driver D
8
10
12
3
5
8
8
12
2
5
8
a. m.
. 45
. 30
. 15
a. m.
. 45
2.
5.
7.
5.
6.
6.
3.
3.
6.
4.
-
0
7
7
5
5
6
3
2
3
6
12
50
50
13
50
45
38
22
32
37
19
rs
Panelist C
%COHb
6.
6.
8.
7.
10.
8.
10.
7.
16.
14.
8.
17.
9
5
6
4
4
9
2
9
2
0
5
3
CO
28
45
15
45
-
31
8
24
32
34
37
44
Panelist D
1.
3.
3.
6.
4.
7.
6.
3.
1.
1.
4.
1
0
1
5
3
8
4
1
7
7
9
33
50
39
50
45
30
10
39
39
19
-------
32.
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-650/1-74-001
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
The Use of Panelists as Substitutes for Taxi cab
Drivers in Carbon Monoxide Exposure.
5. REPORT DATE
July 1973
|6. PERFORMING ORGANIZATION CODE
t
V
7. AUTHOR(S)
A. Walter Hoover, M.D. & Robert M. Albrecht, M.D.
8. PERFORMING ORGANIZATION REPORT NO.
|9. PERFORMING ORGANIZATION NAME AND ADDRESS
Division of Environmental Health Sciences
School of Public Health, Columbia University
600 West 168th Street
New York, N.Y. 10032
10. "PROGRAM ELEMENT NO.
1A1007 21AFU
11. CONTRACT/GRANT NO.
CAMP-8-68(l-68) and
CPA 22-69-97
12. SPONSORING AGENCY NAME AND ADDRESS
Coordinating Research Council, Inc, 30 Rockefeller
Plaza, New York, N.Y. 10020 and The Environmental
Protection Agency, Research Triangle Park, North
Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final Report
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Analyses of breath and limited blood samples from 30 pairs of taxi
drivers and panelists who drove in New York City traffic for 8 hours on
two consecutive days indicated that both panelists and drivers attained
similar COHb (blood carboxyhemoglobin) levels. This was true for both
smokers and non-smokers though smokers had significantly higher concen-
trations of COHb than non-smokers. There was no consistent difference
between the first and second day of driving in the levels of alveolar
carbon monoxide.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS |c. COSATI Field/Group
carbon monoxide exposure
smoking
driving
blood carboxyhemoglobin level (COHb)
CRC-APRAC Project No.
CAPM-8-68 (1-68)
Plumonary
Physiology
Environmental
Toxicology
18. DISTRIBUTION STATEMENT
19. SECURITY CLASS (This Report)
N/A
21. NO. OF PAGES
35
Unlimited
20. SECURITY CLASS (Thispage)
22. PRICE
EPA Form 2220-1 (9-73)
------- |