Tobacco Documents Online

This is a draft copy of a paper desijned to be published in a trade journal such as the Tobacco Rsporter rather than a seientific journal. I have tried to sake it easier to read than standard journalese rould be. The idea is to get the aasa=e to the inde:stry rather than to the scientific eoamlinity. The topic is finjer occlusion. ,-. iiw. . _.- _ i fhts dncument belon9t; to F'htlip Morris U.S.A. it cannot be photoc.,pied. it cannot be transterred to a coileague. h f ' ' er use or it. s Central File agZariD,ti as you have no turt ceturned to R&D 1 la rnuat be

------- -w-r-----s--------r-a ------r DRAlT COPY ---u--..---r--------'- o--------------- TAR DELIVERIES WHEN FINGERS COVER DILUTION HOLES It has been recently suSSest.d(*) that smokers who cover dilution holes with their fingers will receive markedly more tar than FTC smoking machine procedures i.ply. To investigats this issue, we designed a series of fact-finding studies on the effects of dilution hole coverage on FTC tar delivery. We sought answers to several basic questions. They weres 1. How many holes are covered when fingers are placed directly on dilution holes? 2. What does this amount of hole coverate do to FTC tar delivery? 3. During what proportion of pufft do smokers cover holes? 4. What is the net effect of this coverage on delivsryt and finally, S. How i.portant is the problem to us as an industry? * Koslowski, I. T. TreckMr, Il C. Khouw, V. topa, M A. The sisuse of "less _- ~. . . . -- - --- -- - , has~~do~en+,~t#~k~~tr:dr''di~d "dste~~:#~o# 1wlii-ibiaok#p~ bt''rFdi~11 ltil~i~-'tC1iiilrb 6 n118Hi,Ue~ 111 rnUtit tret reiturRed to R&D s GCFltral ~-/Ie sb bb0l1~8S you heVe no turl7+tlr uStl /or st.. ~ Amlriaan Journal of PuMlic Hs+t#th 1f$Ot 701 12O2-S• i ~ M M" - f~_.M•

~_._..._.. -- -'--- -2- -1. In addressin; the questions we gathered some informatton at our own Research Center and had some data collected by an isipartial non-industry laboratory accustomed to measuring consumer behavior scientifically. Our answers are based on the combined research from the two sources. Paint Smears To ai~swsr the first question we Interviewed 31 smokers from among our R&D personnel. We covered their fingers with paint and then asked thsm to place their paint smeRred finaers directly over the holes of five types of dilutad filter cidarettes. These cigarettes were chosen from amona the many packinss and brands available to consumers because they wre of considerable interest to Philip Morris. They do not cover all lensths and flavors, but are probably otherwise representative. They had nominal deliveries of trom 1 to 16 ms !TC tar and had differing amounts of dilution achieved by different procedures: laser, mechanical, and electric spark perforation. The extent of the paint smears waa carefully saasured under the aicroscope, which revealed that an average of 43 to 60 X of the holes could be covered when fin=ers were intentionally placed on the holes, depending on the cigarette (s.e illustration and Table I.) aven the intentional coverage of holes cannot block all holes when only fingers aee used as the obrLruetion device and th. cigarette is held normally. (Insert i1luMtration about here) Illustration shows (a) finisr0 on cigarette shawind tisat only top and bottom surfaces arR in contact with skin and .~.~.. . _... ___... __ . _._ _. _ . .~ ... _ . __s i his docutr{.~1)k~,~la~nt A~I#Ir 0R9 *corked 11p61*arMtte• "t „~ otr transierred to a colleac2ue. [3t rnust be returrzed to R24D s Centra/ Ffie as soon as you have no turthOr ube tor it. ~ _ i .; i .: 4

-3- To answer the second questioa, we applied a coating of ilue to cover the holes of other sample ciSarsttes to the saa sxtent that our samples had b*en paint- seeared and then smoked the cigarettes in standard 20-potrt smoking machines according to the FTC procedures. Chemical analysis of the swoke condensates from these cigarettes was performed in our laboratories using standard procedures (see Table I). As would be expected, we observed some slight delivery increases in tar with the magnitude of the increass depending upon the extent that the en;ineerinS of thk cigarette had relied upon dilution to achieve the brand's nominal delivery. Thus the slightly diluted brand A increaoad its delivery only from 16.3 to 16.5 sg FTC tar, whereas for the highly diluted brand 1R the increase was from 1.0 to 2,,5 m,d 1TC tar. Nodsrately diluted brand C increased from 7.9 to 8.7 .y tar. -4 ~ (Insert Table I about here.) We must recosnise that to the extent that consumers happpen to thwart the intent of cigarette desi=nerr by the way that they handle the ciSarettes, there will be some inevitable mliSht changes in the deliveries of diluted products. The real issue is the extent to which such behavior occurs and the extent to which this sisleads the consumer (and the manufacturer) about the probable tar deliveries involved. his document helonAs to Philip Morris U.S:A, tt cannot be photocopted. It aannot be transferred to a coileague. [it rnust be r4turnecs to R&D's Centrui Fite as soon as you have no further usa t4_r it. ~

-4- Observations of srokers To find the degree to which such thwarting occurs in the normal smoking population, we asked NeuroCoamunications Resoarch Laboratory, Dsnbury, Connecticut, to make careful and unobtrusive observation of a large group of consumers smoking cigarettes taken from their own pockets. WCRL is an independent laboratory best known'for its research in brain hemisphere response to advertising copy. It h.s no tie to Philip Morris, and was chosen for its ability to make scientific consumer observations carefully and ispartially. The MGRL consumers sat in a chair next to an ashtray and watched local television programs and commercial advertisements, which they were then asked to evaluite. When they smoked a cigarette their behavior was videotaped by three hidden video ca.er•ss which provided different views of their head, mouth, and cigarette. it is difficult to measure the placement of fingers on cisarettes in any unobtrusive fashion. Not the least of th. problem lies in the fact that many smokers grasp the cigarette from the ashtray, bring it to their lips, let =o of the ciRarette, take a brief puff, and then end the puff by regraspins the cigarette and removing it from the mouth in a well practiced motion which lasts but a few seconds. VideotapinR proved a good, but not perfect, wuy to record this behavior for later close examination during replay. We 4sked NCRL to measure the extent to which full flavor, low delivery, and ultra-low delivery smokers covered the dilution holes of their cigarettes with their finjerA, paying particular attention to the duration of such aov+trIse This document belongs to Phitip Morris U.S.A. It cannot be photocopied. It cannot be transferred to a}cofieague. it rnuat be returned to R&D's Central File as soon as you have no further u.Pe for it.

during the actual puffins. Industry critics, we reasoned, would predict that smokers would Set increased tar and nicotine from 16w delivery and ultra-low delivery cigarettes by being more apt to cover holes for greater proportions of time than would smokers of full flavored hisher tar and nicotine brands. Such a prediction is based upon th4. idea that smokers can (and do) compensate for low tar by behaving in ways that increase tar delivery. If, on the other hand, smokers do not cover the holes in'order to increase the delivery of their ci=arettes, but instead only cover the occlsiotially and randomly as they Srasp and regrasp the filter during puffing, then we should ses similar incidence and duration rates of hole coverkse among the three groups of smokers. In either case, given the proportions of total smoke intake taken while holes are uncovered and covertd, we could calculate the net effects of the coverage upon the !"tC tar delivery of the five cigarette models of iamediatt interest to us. The video tapee were reviewed and scored by three NC1tL judges who watched thes in slow wotion .ode and ti.ed the duration of hole coverage when it occurred within each puff's duration. The puff durations werQ inferred from changes in the `low characteristics of the coa1R. The judses were in general ajreesarnt about the occurrence and durations ot finger coverage, although there were of course some puffs taken by some ssokers on which they disagreed. To lessen thq likelihood that their evaluations could be seriously challenged by a critxc, we asked that when a judge had reasonable doubt about whether hol.s were covered or nut the observation should be classed as coverage. The dack was therefore stack«d in favor ofe r.portinS hole rov#rajf_in thie_ . ............... i hts docurrient belnngs tu Phthp Morris U.S.A. It cannot be photocopied- it cannot be transterred to a colteagtte. reseoicnh,..~ t.fis;tdid,aot~4ntlrottcw,.tkif_Paoeeco"f:dx4' b'ils,.Wit4es wi felt that

critics making similar measures would be biased in making their scoring decisions, but because we wished to evoid the criticis^ that our judges had an anti-coverage bias. From the industry's point of view, the data so obtained would represent a"worst possible case" interpretation of the amount of coveraXe. The judges observed 113 people takin= a total of 1105 puffs. The majority of these puffs (59X) were initiated and completed without the fingers over coming in contact with the dilution holes. On most of the occasions when fingers did come in contact with the holes they did so at the and of a puff when the cigarette was re=rasped to be removed from the mouth. Classifying any puff which had its holes covered at any time ao an incidence of coverage, thera was no differRntial coverage among the smokers of different tar deliveries. On the average, 432 of the puffs on full flavored brands, 45X of the puffs on low delivery brands, and 33% of the puffs on ultra low delivery brands had at least some coverage during their duration accordinS to the worst case criterion. In toras of the proportion of the smoking which occurred with holes covered and uncovered, 26.7% of the total duration of the full flavored puffs, 24.4X of the total duration of the low delivery puffs, and 21.01 of the total duration of the ultra-low delivery puffs were taken with holes covered, so that 73.32, 75.6X, and 79.0% of their durations were uncovered, again using the worst case standards. AlthouRh the observed incidence of coverage was opposite to that which a compensation hypothssis would pr*lict, we will not belabor the potnt. It is sufficient to note that there was no evidence that the low and ultra-low i his d{cument beiGnqs to f'hiiip Morris U.S.A. it cannot be photocopied. it cannot be transierretl to a colleague. it rriu::t be returned to Fi2k0's C,entral He as soon as you have no further use tot it.

-7- delivery smokers were more apt to cover the holes with their fingers than were the full flavor smokers, and that at a minimum it is scientifically correct to reject the compensatory covering hypothssis. It is sore likely that finger coverage is un:elated to delivery and is tha consequence of wore-or-less random sraspinSs. We note in passing that there havs'been alternative compensation hypotheses proposed, such as the suggestion that smokers take bigger and longer puffs on low delivery cigarettes. It is not yet possible to measure puff volums unobtrusively, but we were able to measure the smokers' puff durations with some accuracy. It was plain that the three Zroups had th. same average puff duration, and that the lowar tar smokers did not take longer puffs than people who smoke full flavored cigarettes. Coabinine the Data ue th.fn calculated the probable worst case effect of such covera=e using the combined laboratory and consu4sr data, tor cigarette A, for exaaple, if 26.72 of the smoke was being taken at a delivery rate of 16.5 s of MTC tar (holet covered) and 73.3x was being taken at a delivery rate of 16.3 .S of FTC tar (holes uncovered), then the total FTC tar delivery of the cigarette Mould be about 16.6 s>;. tuch ccl,cul.ct-;ns were snds for each of the five aodwls tested in the laboratory, using thn observed lRborstorr deliwtrisa and the general proportiou data f Qr th~~~;~ks~Upi~pr~,~t~i°h6 ~~ftt p.~!~i~R ![~,~.l_~I:i,~;~~rG. n Us p v I !t rnubt be -urneci to R&D's CentrEii File as soon as you have no further use for i!. _ .~ _......,._. ._. ._.. Th• co*.rs6e st'...cts !n no case exceede$ .3 MIS M tter#xhjah taoet th Cn CC

-a- (Insert Table II about here.) In the USA advertised tar deliveries of ci;arette bronds are presented as whole numbers, not decimals. In the case of the five brands examined here, the minuscule tar inereases would have - left the advertised tar deliveries unchanged had the effects of finger coverage been taken into account. We do not suggest that all brands would have been so unaffected. Given a slightly different set of consumer and laboratory observations applied to a different set of cifarettee we would still expect to see trivial increases in delivery but would not be amaxed to set advertised deliveries increase to the next larger whole number for some brnnds. The brands we would expect to see so affected would be those low delivery products (a) whose nominal FTC tar deliveries are principally created by dilution rather than filter efficiency and (b) whose deliveries are just above, rather than just below, a whole nusber. We conclude that finger coverage is of questionable ioportance. The calculated delivery changes are trivial, and even these are artificially high because of the bios in interpreting any coverage as occlusive coverage. Given the high cost of gathering such data and the unlikely probability of seeing any practical consequences of the effort, we consider such research as i.portant only insofar as it sheds li=ht upon the ways people smoke cisarettes. =itnger coverage has relatively little bearing upon the question of h ow much tar any given brand delivers compared to any other brand. ; This document belonps to Phihp Morrrs U.S.A. It cannot be photocopieci. It cannot be transterred to a colleague. ~!t rnubt be rt~turned to R&Q's Centrai Ftie as soon as you have no furth©r use tor it.

Percentaje of Holes alocked uhen tiasers Are Placed Directly on Th.a and the Eltects of Elock&Ee on FTC Tar DeliY.ry lrand Nerber of Rowe of X Holes overlaid tTC Tar Deliveries (.E/ciE) Deliberately Clue Qsoa! Perforation Holes by tinEers Unblocked Dlockad A Electric Spark 1 60 16.3 16.5 D Mechanical 1 55 10.2 10.7 C Lser 2, 48 7.9 8.7 D Laser 2 43 4.5 7.0 E Laser 4 51 1.0 2.5 _ ._ . .. __.-~...._~ _..._._... i his document t;elongs to PhiVlp Morris U.S.A. it ca nnot be photocopied. it cannott be transterred to a colleaflue. i it rnust be returned to ii&D's Centra! File as soon as you have no further use tor it. `