Tobacco Documents Online

PUFF PARAMETER A'NALYZER' CRITLQUE Puff Parameter Analyzer studies suffer from unavoidable biases and defects : They purport to compare how BARCLAY "functions" in machines and in, smokers' lips but do not attempt to compare how the "functionimg" of other ventilated brands may differ when smoked by humans and by machines. 2. The device fosters occlusion of BARCLAY's ventilating grooves that would not occur in normal smoking. 3'. The studies are highly artifici'all laboratory experiments that tell virtually nothing about how smokers may behave in normal smoking cond i ti ons . Anti-BAR,CLAY'Bias of the PPA The very design of the device forces the subject to subvert the BARCLAY. ventilation system, while preventing him fromi covering the ventilatiion holes of any other test brand with his lips and fingers as-he might normallly do. This design bias makes it appear that BARCLAY alone functions differently when smoked by people and by the machines. Ini fact, Dr Kozlowski (a consultant retained by the FTC), has repeatedly st'ressed'that a substantiali percentage of smokers of all other ventilated cigarettes tend'to bllock the ventilation holes on their brands and reduce their dilution as well. As Dr Kozlowski has noted', the fact that the,Philip Morris Puff Parameter Analyzer ("PPA"), protects the venti llati onisystems of al l brands but BARCLAY: Uw..(n ~4 2: ,---7 1 WICLGCC~it.-mo~ G.~ /i~wtC'

/2 "biases toward finding BAR'CLAY worse than the others, when, ini fact, BARCLAY and the other ultra-low tars might be alike (for the same basic reason) in commonly delivering more tar than the FTC ratings indicate." The Design of the PPA Induces Occlusion with BARCLAY The PPA encourages subjects .to drape their lips over the venGillating, ports of the BARC'LAYfilter.Ihthe USTest'iing study, only about lUinmi of BARCLAY fi1ter was -availiable for insertion in the smoker's llilps.. Even if most smokers ilnsert a cigarette no more than 9mm into, their mouths, a hefty device such as the PPA, placed at 10mm from the filter end, will cause subjects to pull their lips back from the apparatus. In this regard, Drs Louis Fine and Keyoumars Soltani of the University of Chicagp have advised Brown & Williamson that these limitediiinsertio.n dlepths during lip contractiion result in an outward1ly-directedl force on the cigarptte which must'be counteracted by a reactive inward force. ~ exerted by t'he smoker. This artifically-indUced reactive force is very likely -6 foster occlusion of the BA'RCLAY vent outlets by the smoker's contracted llips. The awkwardness of the devices combiinedl with the limited inseltiion depth results in an increased baseline lip pressure. The decreased available surface results in a greater force per unit area, and the location of maximum pressure shifts toward the end of the filter. These design. features are likely to cause blockage of the vent outlets on a BARCLAY. Moreover, due to the weight of the PPA (415& grams),, the cigarettes are apt to be held at an abnormal angile further decreasingi the availlable surface area and distorting the subject's behaviour. In summary, the dlesiign, of the PPA makes occlusion virtually unavoidable.

/3 The Artificiality of the DilutionrMeasuriing Devilces The ventiilation studies show, at most, what people may do when puffing cigarettes attached to cumbersome devices and, when restrictediin their movement and puffing behaviour. Data derive6from highly artificial laboratory smoking can tell' us nothing of rellevance to the;centra1 issue of how people typicallly treat ventilated cigarettes when, smoking freely. Dr Kozl~owski' asserts that the PPA "proviides a liimited, measure of human smoking behaviour," thatt it is likely that these laboratory settings fostered intensive smoking, and that these dilution studies should not be usedias a basis for determining BARCLAY's "tar" delivery. Philip Morris have producedlsome data showing, at best, that BARCLAY is puffedl at lower di'lution levels in laboratory experiments than it is on smoking machines. Twoo explanations can be hypothesiizedlto explain this data: 1) that the BARCLAY filter is.prone to occlusion; or 2) the measuiri ng devi ce used ca~uses occlusi on of tfie BA'R'CLAY, fi l ter vents to occur. If the second explanationi iis true,, there is no reason to concllude that such occlusion typically occurs outside these experimental conditions. Philip Morris contend that a significamt number of'BARCLAY smokers occlude outsiide the laboratory, thereby adoptiing,t'he first explanation., Why they reach that conclusion is no mystery. But no evidence has ever been presented as to whether reduced'dilution occurs at all outisi'de the laboratory or, if it does, what percentage of ./'..

/4 BARCLAY smokers may routinely occlude the BARCLAY' filter vents under normal smoking condiltions, and to what degree. On the other hand, Brown &' Williamson has shown that the PPA fosters occliusion a~nd other di'storted smoking behaviour. The Rlelevant Data - Narrowing the Inquiry To the extent that the US Testing, study purports to show smoking, the only data of a~ny relevance are the data on lit cigarettes. As Philip Morris itself concedes, a smoker's puff characte istics may be quite different when sucking on unlit and lit cigarettes. Moreover, smokingi machines use 1'it cigarettes. Therefore, to examine what the US Testing, data, show, the first step is to treat all unlit values as irrelevant. For the 134 subjects tested on lit cigarettes,, the average dilution rate for BARCLAY on the PPA was 55.6%. The second step is to ignore any data: on BARCLAY by non-BARCLAY smokers,,for two reasons: 11. The question of BARCLAY dilution levels is relevant only to BARCLAY smokers. 2. As Dr Kozlowski suggests, studies of occlusion or holle-blockingg should use subjects who regularly smoke the brand in question. Even Phi l i p, Morris concedes that thi s is the relevant population. For BARCLAY' smokers, the average, liit diilution, rate on all BAR'CLAY cigarettes (Kings and11100s) was 60.2%. The BA'RCLAY Kings and 10'0s have different dilution levels by design even.on machines. By averaging these different BARCLAY cigarettes together, Philip Morris managed: to obtain lower dilution levels for BARCLAY than would otherwise be the case. If BARCLAY Kings are consildered separately, as they s.hiouild be, BARCLAY smokers achieved an average of 63.3% dilution level on lit BARCLAY's in the WS Testing studly, and all smokers achieved an average dilution levelof 56.6%. ../ ..

/5 Thus, even if the Puff Parameter Analyzer were ai flawless replication of normal smokingibehaviour that fairly compared BARCLAY to other ventilated ciigarettes, at most the US Testing study shows that, dumingi this laboratory experiment,, BARCLAY Kings smokers averaged a 63.3% dilution level - far from the 45.6'% claimed'by Phiilip Morris or the % claimed in the Fresenius study.. The Estimated "Tar" Deliveries of Philip Morris are Unsupported Philip Morris: claiimed that UA'RCLAY, "deliivers more 'tar' to smokers than brands listed at seven to eight milligrams." This "deliivery" figure is based' whollly on Philip Morris' dtiillution studies, despite the fact that Philip Morris itself cautions against using laboratory data tolpredict human values. Moreover, the dilution figure of 45.6% which Philip Morris used was derived'from unlit cigarettes. The i'it diilutilon data ini the US Testing report for BARCLAY Kings puffed by BARCLAY smokers was 63.3%. At a 63.3% di'lution level, and otherwise standardlparameters, "ta.r"'del'ivery by machine smoking has been estimated at 5mgiby Phiilip Morris, 3mg by R J Reynoldk, and measured at 4tng by Brown &,Williamson.