'SAFETY INDEX FOR CIGARETTES
To get a feel for the- assumptions necessary I was asked to
attempt to write a valid' safety index.
Without going into all the thinking the following is the sort -
of thing a serious attempt would come up with:
There would be two indices, one for men arid one for woman but
'the differences would be merely in figures for mortality, loss
of expectation of life etc. Here I will indicate the index
for men.
The index would relate to one cigarette and would be used by
consumers, possibly to compare two types of cigarettes of which
they smoke quite different numbers.
The Index:
Le 27a-lt
TUC, 2) [7o -,co 4- [7"-h@x
VV1
.F7 t-. -/. @ 7
'A,
-j- cited bi
W$-Z M
+ WAILC I-?,
Where W is weight yield per cigarette.
L is average loss of lif a expectancy due to diseases
lung cancer
cardiovascular
-e-bronchitis
;,j.. others
is a relevant test result indicating relative biological
activity of component from one cigarette compared with
the same component in another arbitrarily chosen as
standard.
Is proportion of deaths caused by smoking. (Zen
czar
Contd. "',-J
BATCo document for Province of British Columbia 22 April 1999

2
is the total number of deaths in a given population
over a period - say one year,
is some measure of toxicity per unit weight of
component relative to toxicity of nicotine.
The following figures for U.K. could be used:
Men N total) Women (% total)
Deaths from lung cancer 30,000 (9.4) 7,500 (2.4)
Cardiovascular ( CHD. ) 94,000(29.5) 69$000 (21.6)
Bronchitis 21,000 (6.5) 69500 (2.0)
The following adsumptions are made or implied in the Index
whether justifiable or not:
1) Cardiovascular disease is caused to the extent of 15% of
all such deaths by cigarette smoking and by components
in smoke directly proportional to nicotine and also by
carbon monoxide and further there are linear dose response
relations in all cases and no thresholds.
2) 95% of all lung cancer deaths.-.are caused by smoking
cigarettes and the risk for an individual is directly
proportional to the DPM delivery measured by machine
smoking.
3) Bronchitis deaths are caused in 20% of the cases by
smoking cigarettes and are directly proportional to the
yield of nitrogen oxides in the smoke.
4) Average loss of expectation of life due to smoking related
diseases are as follows:
Lung Cancer 2 years
Cardiovascular diseases 5 years
Bronchitis I year
Simplifying assumptions are necessary whether justifiable or
not as follows:
-1) The only important smoking behavioural assumption is that
nearly all smokers smoke for nicotine and adjust the way
they smoke to compensate for changes in nicotine delivery.
No other differences are important.
2) No other diseases or components of smoke are important-in
this context.
3) There it no interaction between diseases or between causes.
4) 7Fx r4OL
7&6e. /-J rh.,,,,
5) me, /a - 0 To m. e-
74a C 0 'I-'- t Alx-
Contd
BATCo document for Province of British Columbia 22 April 1999

3
6) All components are weight and not concentration related
to disease.
7) All dose relationships are linear without thresholds.
It should be noted that Blwt.could be benzpyrene content.
Furthe promotion and initiation are not considered but
r, pro be taken into account in the calculation of
these co
-e .4. = *
as a rough approximation.
The writer does not consider it reasonable to make all these
-assumptions and would argue some are not even plausible. But
if they were acceptable within some limits a few examples will
show that with any reasonably wide limits the order of cigarette
indices can be changed at will.
-Consider three cigarettes:
A: Tar 10 mgs. Nic. I mg. CO 2 mgs. NO@x 0.1 mg.
S: Tar 20 mgs. Nic. 3 m9s. CO 4 mgs. NO 0.1 mg.
X
C: Tar 5 mgs. Nic. 0.8 mg. CO 1.2 mgs. NO 0.1 mgV
X
and
Case I Assumptions as above:
Case II Assume smoking causes 25% of heart disease.
Index: Case I Case 11
Cigarette A 496,910 590,910
B 496,663 700g173
C, 345,525 453,625
It can readily be seen that equally plausible assumptions could
be made to change the ranking of cigarettes almost at will.
It may be concluded that there is insufficient knowledge at
present to enable any useful combination of smoke constituents
to be made into a single index.
BATCo document for Province of British Columbia 22 April 1999