I I ·: -·IC; C i ~i;hC~Y!1"E V :,J - ~· ,· ;, ie~:Cd?ss~ll;~*L·- : 1; CTMtC'70li;z-~ E2 ' ·~·;djp%~j!~E~~ c:~::_p~r . -5; " : 3~~~Zq~ *d~t· *~~· ~·- t ~F "R ~ ~~··-r·~5~~ii .~· ~ ~T";:;L ~':3F L~c;~r~` 1~rC ··· pa.-c~( 4.)C '" '·~ ·*~ :·:· "~151~~ .z c ~u ~e, r· --· s ··, r' ·"~*"' 'Z"i~~' i).. I · ~·-· c~ ~,· ,... --IC ···'~'-- '" I a ·: :~··~ :~· I·· .. ~· ·~ ' '· ~aani "' ' L~1· ··~ · .Dab~C~ 1~ ·~i "' $ "' -"· I "· ~-·r.,· i.,~~c~,·~yr*lr · x~ ~.,. L~ · ·~~ '' Y~ I ~~: r " i I, , ·--·~L ~~· ·is,·~~ '' r -"'"·~'" ~P~a~C-·· t ~~'.~· ~" ' ·s. 'F;?:~* iTY ~i : ... ;,,;,,;·i-R ~tiP·jrSJFn l~ifb C~p~ ·;;~dr~~~~~ce "' .i FI~~~~~IC.rR· It "· FS c -··---~---·- · I.bf plMWRT)LWmD IM~ba~~ UT CPUP : *· y~l·I~~Y·k~;rrlkriwl~ !PZ0200i BATCa doclnont br LegJ benl#s : Hlath Calldl 18 rl~ ~990 B.A.T (V.K. &Euport) ltd., Research 8 Development Cenire, SOUTHAMPtON, UCWMF~F~46E A PORTABLE MULTICHANNEL RECORDING SYSTEM FOR MONITORING HUMAN SMOKING BEHAVIOUR REPORT NO: T,203·A 29 November 1988 AUTHOR: U,C, Raines CO·WDRKERS: SECTION LEADER: R,P, Ferris ISSUED BY: Dr. taj Hirji DIS~RIBU'TION: t3r, R, Binns 1 Mr, A,L, Heard 2 Mr, M.L Reynolds 3 Mr. M,L. Reynolds 4 Dr, P,J. Dunn 5 Or. S,R, Massey 6 Mr, 7.I, Wilson 7 Mr, T,I, Wilson 8 Herm E, Rittershaus 9 Dr, E, Kausch 10 Dr, C,J,P, de Siqueira 11 Mr, R.F, Gilderdale 12 library 13 library 14 FI1E; 46E COPYIX. 15 Ist 1988 BATUKE. Do not copy dr show to unauthorised persons. BATCo document for Legal Senriees : Health Canada 19 May 1999 B.A,~ (U,K, &Exporti ltd,, Research 8 Development Centre, SOUTHAMPTON. UCWM~VV~GE 29 November 1988 A PORTABLE MUt~lCHANNEl RECORDING SYSTEM FOA MONITOR[NG HUMAN SMOKING BEHAVIOUR REPORT NO, t,203·A SUMMARY: this report describes the development and validation of a portable multichannel behavioural monitoring system for use in recording the physical actions involved in smoking. Based on a portable Compaq 386 desktop computer (IBM compatible), the system permits the recording of 8 smokers simultaneously and thus is superior to previous equipment where only one smoker could be recorded at a time, ~his improved efficiency and cost effectiveness makes a practical reality of: (a) routine smoking behaviour feedback on product performance for the use of product developers (b) characterisation of smoking behaviour s~les of target consumer segments within the marketplace 'The system is compact in size compared with previous equipment and can be transported easily into the field. Tne system has a fixed disk drive coupled with atape streamerthus permitting mass storage of data and easing the subsequent data processing, data tum around is thus further improved. As the system is menuariven and therefore user-friendly, agency field force personnel jwithout previous computing experience) are able to operate the system effectively following a hail day t~aining sess on. Typica ure of the si~iem can be either with in-house panels or in consumer trials in hail test locations. An b·port puff o duplicator is currently being Constructed, to similarly increase the rate at which P1 records can be used in duplication smoking, ;5 O 1988 BATUKE Do not copy or show to unautt~crised persons. BATCo document fat Legal Services : Health Canada 19 May 1999 INDEX TERMS: Smoking Behaviour Consumer Techniques Consumer Behaviour Human Smoking Parameters Human Smoking Patterns Puff DupBeato, Puff Recorder Smoke Deliveries Nicotine Photographs B 1988 BAIUKE Do not copy or shaw to unauthorised persons. BA~Co document for Legal Services : Health Canada 19 May 1999 1, IN~ROOUCT10N Smoking behaviour research involves measurement of the physical adions people use when puff ing on cigarettes. The most important parameter measured is puff volume jmeasured in millilitres), which is the amount of smoke a smok.er draws per puff as well as the total volume of smoke gained from a cigarette, ii, order to obtain the required smoke volume the smoker exerts puff pressure (measured in cenfmeires waiergauge], as he or she draws on the cigarette. This parameter is responsive to the changing draw characteristics as fl!r: cigarette is smoked. Otherparameters recorded are puff duration (the t$ci !akei· to draw each puff) and puff ioteNal (the elapsed time behreen puffs) bdi: measured in seconds, This recorded information can be subsequently ~!sed kl drive a puff duplicator (1). This is a programmable smoking machine whic!l smokes one or more cigarettes to the same pattern as originally the smokt!r diri by accurately reproducing the previously measured parameters. The re~i;ltlly smoke deliveri is collected on a Cambridge filter pad to be arlitly~~t·d conventionally for nicotine and PMWNF values, The p~me objectives for collecting smoking behaviour information carl he; summarised as fonows:· 1. To provide feedback on product performance for product developers w'~ci are concerned with understanding how particular product ~g:ii~r. propositions interact with the behavioural smoking s~les of t;ige! consumers. 2 To provide additional information regarding the delivery potential of products, referenced in human rather than machine smoking terms, These two objectives impinge directly on the maintenance or improvement of BATs competitive market position. Underlying these prime objectives is a central concern to develop an explanatory model of the effort·reward relationship involved in smoking behaviour ji.e. the relationship of puff ing effort to the various motivating stimuli: taste, impact, nicotine, satisfaction etc.). Historically, smoking behaviour research has been limited by Pne complexity of monitoring technology (2), which in tux demanded highly timeicost inefficient data collection. The study of consumers or the integration of such work with seasofl(panel) testing has not been possible hitherto since the demand far rapid data collection has not been fulfilled, since behavioutal testing has been limited to single equipmen~subject labcralc~ situations. Since 1985 a new generation o suite of microcomputer (Apple 2E) based behavioural monitoring equipment was P, commissioned, this being a developmental stage prior to the latest system o described in this report. For the first time a commercially available computer [13 O 1988 BATUKE Oo not c~py or show to unaulhPrised persons. BATCo document for Legal Services : Health Canada 19 May 1999 controlled the entire data aquisitio~recording and duplication stage. The advantages of the Apple 2E based system were that ii greatly reduced the amount of electronics by using commercially available components, the system could be programmed to make t easier to operate, and its computer based design greatly simplified the building of puff duplicators, The disadvantage was that the Apple 2E based system was dedicated to only record the smoking behaviour of one smoker at a time, so that sample sizes were limited to achieve realistic timingsl costinqs. Fu~hermore, if a number of Apple 2E systems were employed, the bulk of equipment was extremely difticult to transport, quite apart from the difficult~ in providing a natural relaxed smoking environment in which the smoke rs could feel at ease, The key difference regarding the latest behaviovrai monitoring equipment is that up to eight smokers can be recorded simultaneously, using a single portable computer, this permitting consumer sampling on a large scale. jTo test the turn round capability of this system, 600 srrokers were recorded in 2.5 days, whilst it would have taken 30 working days to record the same number of tests on the Apple 2E based system). Based on a Compaq 386 portable computer (IBM compatible), basic concepts and approaches to data collection are identical to the Apple 2E based system but the speed and storage capahili~ of the Campaq 386 are considerably superior. When ail eight smoking stations are to be used, the Complete system can be packed into 2 cases of the dimensions 50 cm x 30 cm x 50 cm, each case weighing approximately 24 kg. The new system is thus easily portable and considerab y less bulky than the number of Apple 2E systems required to achieve the same data collection objectives. 2.HARDWARE Photographs of the behaviaural monitoring equipment are exhibited in the Appendix as Figure 1. The basic components of the system are: Cigarette holdei(s) Transducer box(es) and power supply unit Compaq 386 potable computer (IBM compatible) At maximum capaci~, 8 transducer boxes can be connected to the Compaq computer which permits the simultaneous recording of 8 smokers. A diagram of the cigarette holder and transducer box is shown as Figure 2 in the Appendix. 2.1 Cigarette Holder (Ref, to Fig, 2 in the Appendix) o To record be smoking profiles, pressure and flow are measured whilst the smoker is puffing on the cigarette. The cigarette is placed in the cigarette '3 holderwhich contains an onfice plate thus providing the necessary impedance ·p ~ 1988 BATUKE Do not ropy or show to un~ulb3rjsed persons, BATCo document for Legal Senrices : Health Canada 19 May 1999 as the smoice is drawn through ii, The pressure difference caused by this impedance is measured to calculate the rate of flow on the basis of calibration tests, The cigarette holder is made from polypropylene and is injection-moulded for accuracy, so that Vie oriEce is 2mm Y3.02mm in diameter, The draw resistance of a typical holder is 0,3cm W.G, at 175 mVsec, flow rate. The holder is completed by a mouthpiece which is 25mm in circumference, in conformation with most U.K, cigarettes (otherdimensions can and have been produced to ower), Light plastic tubes Im long are! connected to the sidearms on either side of the orifice plate. 2.2 Transducer Box and Power Supply Unit (Ref. to Fig, 2 in the Appendix) The tube in front of the orifice is connected to a FC040 Furness Controls ltd. transducer which has a range oi 0-1000mm W,G, The other inlet to the transducer is open to atmosphere, The tube behind the orifice is conrlec.ted to one inlet of a second transducer identical to the first except for its range of 0·1 OOmm W.G. The second inlet to this iransducer is connected by a T p eee to the tube leading to the first transducer, The first transducer thus measures the pressure behind the cigarette but in front of the orifice plate relative to atmospheric pressure, The second ttansducer measures the difference in pressure across the orifice plate. The pressure drop across the orifice is related to the flow through the orifice plate. The power supply unit provides the necessary voltages for the transducers and associated amplifiers. Zero and gain controls for all 8 channels are situated on this unit. The pressure transducer gives an output voltage of 0-IV directly proportional to the applied pressure over the stated range, The pressure signal therefore needs no further linean'satiao. The relationship between pressure drop and flow through the orifice plate is not linear but exponential. The voltage to flow translation is software controlled, the range of 0-IV corresponds to a now range of 0-110 muse, 2.3 Campaq Computer The voltage output from each ttansducer is connected to the Compaq via a 32channel DI·AF~ Micro Systems DMS 541 data aquisition card. The Compaq E4 386 is equipped with one megabyte of 32-bit random access memory, An N imwnat clack permit be me~ruring oleapsed lime, a 10 me3abile lued ON disc drive is able to store data from approximately 600 smokers, p Q 1988 BATUKE Do not copy or show to unauthorised persons. BAT~o document for Legal Services : Health Canada 19 May 1999 3, SOFTWARE Data a0uisitioo is controlled by the Labtech Notebook proprietary software packagd. The operation of a Notebooic data aquisilao run can be divided among three distinct functions: Data input from the hardware interface device Data storage to dise files Data display on the Compaq monitor These functions are controlled automatically by Notebook via a series of set·up files. Figure 3 in the Appendix shows a block diagram of the data aquisiiion stage. For each smoker or channel, four set·up files are required. Fiie 1 continually monitors analog pressure. When a puff occurs, the analog pressure rises past a preset threshold value. On exceeding this threshold, files 2 to 4 sarr,pie time, pressure and fiow voltages respectively at a sampling rate of 25 Hz. Sampling continues until the pressure threshold has been exceeded on the downv!ard slope of the puff, Thus a data file is created for each smoker which contains times, pressure, and flow voltages, for each puff as the cigarette is smoked. When all smokers have finished smoking their test details and file names are entered (this process is described in detail in the section headed 'operaiions). A series of custom developed programs interfacing with labtech Plotebook uses the data in each file to calculate actual pressures and flows for each pu3, First the data files are extended by linear inierpoiation of a further data·point between each pair of sampled pressure and flow voltages. The delay at be start of the puff, incurred whilst the pressure threshold is exceeded and before sampling starts, is overcome by linear extrapolation of both pressure and flow voltages to zero. As already mentioned in section 2.2, the pressure voltages are directly proportional to the applied pressures and need no further conversion. The flow voltages are converted to actual flows using a formula which describes the relationship be~een pressuredrop and flow. Actual pressures and flows are subsequently integrated to produce integrated puff pressure (also called puffing efforti and puff volume for each puff, The time readings are used to caiculate the duration of each puff as well as the inte~vai be~een puffs. In su mmary, for each smoker a file containing actual pressures, actual flows, integrated puff pressures, puff volumes, puff intervals and puff durations is stored on be Compaq 386 fixed disc drive. O 1988 BATUKE. Do not copy or show to unauthorised persons, BATCo document for Legal Services : Health Canada 19 May 1999 If a printer is connected to the Compaq computer a further program can be used to obtain a hard copy of the summary van'abies: Puff Intervals (seei Puff Durations (sec) Puff Volumes (ml) Puff Pressures (cm W.G.) O 1988 BATUKE Do not copy or show to unauthorised persons, BATCo document for Legal Services : Health Canada 19 May 1999 4, OPERAflON ihe recording system is started by typing the command "S"M", fhis command loads the necessary programs and the main program ent~ screen is displayed on the monitor. T"ne system is entirely menu·driven and prompts the operator to select each option displayed at the top of the screen, whilst indicating at the bottom of the screen how these options can be selected. The options to be selected are: Session Channels Process and the choices are made by either pressing and holding down the "ALT" key together with the initial letter of the option you wish to setect, or by pressing the right and left arrow keys until the cursor highlights the chosen option. As this selection is made, mufipie·choice menus appear on the screen, From these menus, choices can 5e made by pressing the "UP" or "OOWN" anew keys to move the highlight bar to the option required and then pressing the "RFTURN" key to enter the choice. i Session channels Ge Process F~ for advert 'PITt' ~ I~nitial to select Quit in Process O menu to puit pr~m O t 1968 BdTUKE, Do not copy or show to unauthorised persons, BATCo document for Legal Services : Health Canada 19 May 1999 4.1 Session This option displays a choice of "AII-day", "Conference" or'Set·tlp". The "All~ay' version of the recording system is used in a panel room·type set-ui; where the recording can be enabledidisabled using a switch located on the transducer box, With in·hcuse panel testing smokers arrive at their own convenience throughout the working day and the switch can be operated by the smoker him or herself tt;us freeing the operator to attend to other matters. The "Conference" version is used to simultaneously collect data from up to 8 smokers and is therefore primarily suitable for consumer testing. H is important to note that recording does not have to occur at the same time for all 8 smokers, the system merely stays in record mode until the test smoker has finished and the operator has reset the equipment. The "Set-up" option is used for calibrating the system, Session Channels Ga Process All·day conference Set-up O 1988 BA~UKE Do not copy or show to unauthorised persons. BATC6 document for Legal Services : Health Canada 19 May 1999 4,2 Channels ~his option displays the different configurations of smoker testing stations available, i.e, 1, 2, 4 or 8 smokers can be recorded simultaneously; Session Cha3ncl~ Go erocess O 1988 BATUKE Do not copy or show to unauthorised persons, BA~Co document for Legal SerJices : Health Canada 19 May 1999 4,3 Go This ootion ds~ap a choee of'Yed4' o~'Run~ Ssbelng Verg' ~1 dsp$l me choices made so far, ~HC~ selecting 'Run' will start the date aoouisition system for calbr~ion or recording. session Chatls 60 PEoesss Verifying ~oiee of session type and ndsr of chaanels for net run sassta ~6 ~S Conftrenct ~fB OP SDfIIG C~gLS ~S F1 for advert 'Ar t initial to select Odt in Process menu io quit prgm, 0 1988 PArUKE. no not copy or show to unauthorised persons. BATCo document for Legal Seniees : Heath Canada 19 May 1999 __ rme SYStem is in e~hdon scds ka'Sau~ ha3 been cdcd~ ]mm the 'Seabd monuj up to 8 r~da are displayed on he SCIP depending upon the n~mbw of smoking stations selecfed ~om lhe "Channd~ menu. in esCh WindWl Ca CLllrent ·bes d the ~nsdra Vo/lagoS Cr )hPI~ and dor are displayed Id PpdY at a trsqllency d 2 HL chaMel IO 1 lo a C1 Flow 0.003 C2 Flow 0.000 CI Pressure 0.002 Pressure 0.002 C~lnner No 3 No I C3 Flow O,OOt ( Flow 0.003 C3 Pressure 0,001 C4 Pressure 0.002 Channel No 5 No 6 C5 Flow 0.001 C6 Flow 0.001 CS Pressure 0,001 CS Pressure 0.003 Channel No Ho 8 C1 Plow 0,002 CB Plow 0,002 C7 Pressure 0.000 C8 Pressure 0.003 O 1988 eATUKE Do not copy or show b unaulodsed persons. BATCo daeulcnt br Lsgd Slnkss : Heslth Caladl 18 YII 181) If the system is in record mode Ii,e, "Canierence" has been selected from the "Session" menu) and 8 smoking stations have been selected from the "Channels" menu, the micro processor speed is insufficient to also display transducer voltages an the screen. the display "press ESC to quit" then informs the oper~ior that the system is in record mode and to use the "ESC" key to end the recording session. If i, 2 or 4 smoking stations have been selected from the "Channels" menu 1,2 or 4 windows are displayed on the screen. In each window a highlight bar displays the current pressure and flow voltages for each smoking station up·dated at a frequency of 25 H2. This display is secondary to the recording function which at times results in brief pauses before the display is up-dated. Press ESC to QUIT C1 - Pressure Flow C2 - Pressure Flow C3 - Pressure Flaw C~ - Pressure Flaw As soon as the Esc key has been pressed, the system prompts for smoker details to be entered. These details must be entered as otherwise a filename for the data cannot be created and the data is lost. Tne system will prompt fore test number (e.g, C0071] followed by a prompt forthe experimental code (e.g. SB44]. From these entries the filename is cons~ucted (e.g. C0071 B~A.DAT for the raw data file, C0071B44.BAT for the processed data file). Other prompts are for replicate number, date, time, sex, subject name, subject number, cigarette length code, brand name and butt length, Not all fields are compulsory and some data entry validation is perlormed, b ~ 1488 BATUKE. Oo not ropy or show to unauthorised persons. BATCo document for Legal Services : Health Canada 19 May 1999 ___ Session Channels Go Process Input Data for Smoicer 1 Test Number IC00ill E~eriment N~ber lS838 1 Replicate Number 1007] Date \DDMMYY' 12201871 Tine \HHMM' 115001 Sex 'M' or $ InitiallMAI Subject Name IWICKS subject Number IMOa61 Cig, LenFth 'A' to \H' IEI Gig. Product IPGL Butt Length 130.5 Input between bars; Rtn to enter When all smoker details have been en~red and the data has been stored on the fixed dise drive, the system returns to the main program entry screen (ii a po'nter is connected and enabled, the summary analysis information is printed for each smoker). 4.4 Process This option displays a choice of "Analysis","Printer ONIOA" or "Ouii", Selecting "AnaIysis" enables the operator to recall a file previously stored to either change erroneous smoker details or to obtain a hard copy of the sLlmma~ analysis information, providing "PRINTER ON" was also selected. Seieciing'Quit" returns the system to the DOS operating system. e 19ee BATUKE. Do not copy or show to unauthorised persons. ~O BATCo document for Legal Services : Health Canada 19 May 1999 5. CALIBRATION Transdocers are affected by climatic conditions and regular calibration is therefore necessa~ to preserve the accuracy oi the recording system. As a rule, weekly checks on the pressure calibration are deemed to be adequate whilst the flow calibration should be checked daily, Although the cigarette holders are injection moulded to very precise specifications, some variability is inevitable, To avoid calibration for individual holders, the recuired number of holders are matched at various flow rates, The use of one "caiibration holder" eases the calibration procedure when all 8 smoking stations are in use. 5.1 Set·Up The upper end of a rotametei (range 0·10 Vmin) is connected to a tube dimensioned to hold a cigarette which in turn is connected to the cigarette holder, The cigarette is thus completely encapsuled, with the cigarette holder protruding from the tube. The cigarette used for calibration should reflect the pressure~drop of theeigarettes to be measured. Theciga~ete holderis connect ed via a nestle valve to a vacuum supply so that the air drawn through the set·up can be controlled. The lower end of the rotameter is open to atmosphere. Great care is required to ensure that the sei·up is leakproof. 5.2 Pressure Calibration The plastic tube which is connected to the cigarette holds r in front of the orifice plate is removed from the holder and connected to a T piece, The other arms of the T piece are connected again to the holder and a water manometer, Zero is set on the power supply unit before the vacuum supply is switched on. Different air flows are then drawn through the system and the readings of the water manometer compared to the Compaq display Inote that a pressure of e,g, 56 cm W,G, is displayed on the Compaq as 0,56), Tne gain control on the power supply unit is adjusted so that the readings on the display and manometer agree, 5~3 Flow Calibration With both ~ansducer tubes connected, zero is set on the power supply unit before the vacuum supply is switched on, Different airflows are then drawn through the system and compared to the Compaq display. The flow rates used, and their corresponding voltages are shown in Table i, o 8 1988 BATUKE Do not copy or show to unauthorised persons, BATCo document for legal Services : Health Canada 19 May 1999 TABLE 1 Conversion Values used for Ftow Calibration Flow Rate Compaq Display (vmin) (Voits) 1,5 0.04 3.0 0.18 4.5 O.d~ 6,0 0,81 6,6 1,00 G,ACCIJRACY The accuracy of the Compaq based recording system has been compared with:· · an 8-port Fjlttona model 302 smoking engine, · a Borgwaldt soapbubble flow meter and · the Apple 2E based recording system The main emphasis of this validation is placed on accurate puff volumes and puff pressures, Puff durations and intervals are "ca[clllated" rather than "measured" quantities (compared to pressures and flows). Sampling frequency and consistent time incrementation have been checked and were found to be accurate, When comparing two computers, some time van'ation is to be expected as the clocks are not running in syechronisation. There is strong evidence to suggest that the occasional puff interval difference (difference greater than 0.1 sec., but smallerthan 1 ,O sec.) is due to the inaccuracy of the Apple 2E "mountain" clock, as the same occasional differences occur when the "mountain" clock is compared to the "proclock" which supersedes it. in either case, the practical significance of this effect is negligible, 6,1 Compaq vs Smoking Engine The a channels of the recording system were connected to an 8·pon smoking engine calibrated to simultaneously draw 2 second duration, 35 mi volume bell·shaped puff s, 95 consecutive puffs were recorded on each channel, The test was repeated with 2 ports linked together, so that a bubhiesticked mean volume of 63 mi was drawn on one smoking station. Table 2 shows the results ;5 of the puff volume comparison. o 8 1988 BATUKE. no not copy or show to unau~orised persons, BATCo document for Legal Services : Health Canada 19 May 1999 TABLE 2 Puff volume Results of Compaq vs Smoking Engine Channel I Chnnpll Channel I Channel I Channell Channell Channell channel I liri~ed 2 3 1 5 6 7 8 Channel 35.08 36.16 35,35 34.51 35.4 35.22 35.06 35.61 64.16 Oeiadwl 0,714 ~,6p 0.695 0.656 0.564 0,729 0.697 0,633 0,518 33,30 31.70 33,40 3250 33.80 33,t0 33.40 34.20 8290 36.40 3;10 36.90 35.70 36,80 37.30 36.60 37,00 65.60 3693· 1 36.06· 1 3521· 1 34.38· 1 35.15· 1 35.01 ·13(.92·135.~9· 1 s40~· Interval35P 36.29 35.49 34.65 35.58 35.37 35.20 35.74 6~~2i~ 6.2 Bubblestick vs Compaq vs Apple A Compaq channel was linked to the Apple 25 system using a T piece to connect the plasSe transducer lubes. Both systems are calibrated n th s set-up, f~e soapbubb e ilowmelerwas connected to the cigarette holder via an airtight tube, thus encicsing the cigarette. 30 consecutive large puffs were drawn by a smoker, iahie 3 shows the results of this comparison, individual measuremeo!s are presented in Table t in the Appendix. TABLE 3 Comparison oiCompaq and ApF~e System tothe Soapbubble Flowmeter Mssn MIM SL 01(, d Maximum Yn ~broktl CD~du~ Cb~luto Prcm~i~ DinwmFI Oiblnwal Dillmne, OinHcnee PIR YQLUME (ml) Bub~es~neomw a,42 052 1.60 PS8a lut~Jckvs~ppkH 251 aoo 130 3411: CPpalnqPo~~2E 209 O~11 130 326 PUR PeESSOREln W,G.L e~apnCPB 1.67 010 220 1,6(1: C~mp~pn~PU GM) Rd6 R10 RMY, WFEIMERVIIL 1MKl C~mpaqvlCpk2E an ale 080 1.Bar, O O N 0: IU \D IV Q 1 og8 BATUKB. Do not copy pr show to unauthorised persons. BATCo document for Legal Services : Health Canada 19 May 1999 _I 6.3 Compaq vs Apple 2E A Compaq channelwas again linked to Vie Apple 2E system using a T piece and bol systems were subsquentiy calibrated. A number of different types. oi puffs, ranging from "short, small" puffs to "bg, large" puffs, were drawn by a smoker to test Vie performance of the Compaq ender extreme conditions. Table 4 shows the results of this evaluation, Individual measurements are presented in Title 2 in Ihe Appendix. Figure 1 presents a ~picalhuman puff profile recorded on both systems. TABLE 4 Comparison of Compaq vs Apple System IOn MOM SL Dsr, al Wrrtlm LhPn Ib60luh WWIII Mwidl PICBntags oinerenoO D1Kmcl OleronFD Ole~e~s Shorl~Upldh O,11 025 0.60 R17~ Shen~pputh O,M 093 lJO Rtm: Avwagepufls o.n 025 I,lo l~ssr, Long large put~ as~ l.00 328K WR~E~S~LRElnW,GI~ She~tsrrallp~~ 0.30 a12 0,se tim Shat~slg,puns 069 035 130 11WI Avdrb3'pHa 0f9 016 1.00 Long brg~ plfls 1.66 01~ 1,90 3.62~ P11FFeURAn_ON~p~F~ Shwrllrallpu(fs 0.01 006 0.10 1.1311 Shpn~plpl~t~ 0.01 aOL P.IO 069K Avwsgl puHr 0,00 003 0:0 a17~C Long hrgP puM 0.01 003 0.10 0262 Sh~rl~~Up~fs 0.01 0.03 0.10 0.655 Shon ~ge pffs 005 RW 01P 0.731: Av~n;l puns 006 P.11 020 3.31X LDng Brgl pns 0.OB 0I2 0.20 0.9h G 1988 BATUKE Do not Eopy or show to unauthorised pnons~ BATCo document for legal Services : Health Canada 19 May 1999 Fit, 1 T. 203 R A TYPICAL PUFF PROFILE RECORDED ON BOTH SYSTEMS --- APPLE ZE FLOW ··---- COMPAQ 386 FIOW ~ APPLE 2E PRESSURE ·--··· COMPAO. 386 PRESSURE FIOW PRESSURE (mllser) Irm Wgl 1 B 46 "i 1 35~ (j 34 26 '"ii ,in I Is 11~11 io ill ~ ie b ill' 2 L..-_ i 2 ' -% n O Z O 2 ~o TIME Isecs) TIME Isetsl BATCo document for Legal Services : Health Canada 19 May 1999 7. DATA FORMAT For each smoker ~o files are stored on the Compaq's fixed dise ddve~ The raw data file indentified by the fiiename extension ,OAT, contains time, pressure, and Ilowvoltages, ior each puff, The data are recorded in thefollowing sequence: Header · to identify smoker, cigarette test etc Date of test Time of test Sampling frequency Channel number Smoking data Trailer · test number, butt length The smoking data are in the form: "T~me" "Plessore" "now" i,e, 22.00000 0.00700 0,00000 22,02000 0.01 756 0.OW74 22.04000 0,02812 0.00747 22.06000 0.03868 0,01120 22,08000 0,04924 0,01 494 22.1 0000 0,05980 0,01 86f 22,12000 0,07062 0.02~50 22.14000 0.11359 0.04281 24,40000 0,06938 0.00500 24,42000 0,04844 ·0,00·438 4224000 0,00700 0,00000 4226000 0,01933 0.00456 4228000 0.03165 0.00913 etc. Thus the duration of the puff in Vie above example is 2.42 seconds, followed by an interval of 17.82 seconds to the next puff, An occasional negative voltage inN the flow channel at the end of a puff is caused by transducer bounce and is interpreted as zero, Q 1988 BATUKE, Do not copy or show to unauthorised persons. BATCo document for Legal Services : Health Canada 19 May 1999 The processed data file identified by the filename extension .BAT contains actual pressures and flows. The data are in the form: Header · to idenii~ smoker, cigarette test elc. i.e, 10000 P0007 F0000 P0018 F0081 P0028 F01t7 P0039 F0142 P0049 F0162 P0069 FOD96 P0048 F0000 D0024 V0694 H0478 10178 etc, Trailer · test number, buft length where = Intervals behreen puff (sec " 10) D = Duration of put jsec ' 10) V = Volume of puff(ml ' 10j H = Total pressure used to draw puff (cm W.G.' 10) P = Pressure(cm W.G.' 10) p F =now(mVsee'lO) o R = End of puff te~minata~ P, ~ 1988 BATUKE Do not ropy or show to una~thon'sed persons. BATCo document for legal Services : Health Canada 19 May 1999 the data are arranged in:his way because the first information that the duplicator (the machine which smokes cigarettes to the same parameters as the smoker did) mist receive is the intenal between puffs, the interval preceeding the first puff is set to zero so the duplicator may smoke the first puff as soon as it is ready. the next interval figure, formatted at the beginning of puff 2, refers to the interval bebeen puffs 1 and 2. Similarly volume, duration and pressure formatted at the beginning of puff 2 refer to the volume, duration and pressure oi puff 1. The inte~al after the last puff is discarded, it being the time between the last puff and the stopping of the recording system, If a printe, is connected to the Compaq computer, the summary variable data can be obtained. the data are displayed in the form: Header Interval Duration Volume Pressure (sec) (seci (ml) (cm W.G.) i.e. 0,00 2,42 69.38 47.77 17,82 2,32 73,67 50,97 Total Means Puff Number Trailer 8, DATA STORAGE AND RETRIEVAL Raw and processed data files are stored on the Compaq's 40 megabyte fixed dise drive, Approximately 600 smoke tests can be stored, the exact number depending upon the size of each file (i.e. the number of puffs the smoker has drawn from the cigarette). Further storage space can be created by deleting the raw data files. When the fixed dise is full, the following options are available; ·Data transfer to the Cyber mainframe computer (a local operating requirement) - Data transfer to magnetic tape cartridges Data transfer to floppy discs oi o 1988 BAUKE, Do not copy or show to unauthorised persons, BATCo document for Legal Services : Health Canada 19 May 1999 If the Compaq system is on-site, the preferred option is to transfer data to the U.K. Cyber mainframe using the Connect interfacing package. A file is created on the Compaq (using a redirected DIR command) which contains all tile names of tf:s data to be transf erred, ~he actual trsnsf er is automatic and can be carried out overnight. On the Cyber, the data are stored in a data base for ease oi retrieval and statistical analysis. if the ComFaq system is oif-site,·data can be down·loaded onto magnetic tape cartridges. It is advisable to dov~n·lodd regularly, not just to create more disc space but also ~ create file back-ups should the fixed dise crash. Data on the magnetic tape cartridges can subsequently be reloaded onto the Compaq and transferred to the Cyber mainflame~ The Compaq's single disc drive can be used to copy files onto floppy discs, Approximately 15 tests can be stored per disc, which makes this type of sto,ageiback·up device most suitable when smaller sample numbers are envisaged. S~atisticai packages forthe Compaq, to enable the userto statisticaliy analyse the collected data, are commercially available and can be advised. 9~ DISCUSSION The advantages of the new Compaq based recording system are manifold, with the physical size and weight of the system making it suitable for worldwide portability, end it only needing one mains power point to became operational. The entire B·stdon system can be packed into two cases with the dimensions 50 cm x 30 cm x 50 cm, each case weighing approximately 24 kg. The processing speed of the computer enables up:o 8 smokers to be recorded simultaneously, but even when the system is used in 8 channel mode the number of smokers (1·8) to be recorded is optional, The system is entirely menu·driven and therefore very user·friendly, During a validating behavioural study in the U.K. (report pending), 600 smoker tests were recorded in 2.5 days [equivaient time for traditional single monitors would be circa 30 days). The system was mainly driven by agency field force personnel without any previous computing experience. The field force staff were trained on location for a period of 0.5 days which proved to be the minimum training pen'cd required to operate the system effectively. Smoking data recorded on the Compaq based system can be duplicated on the o Apple 2E based puff duplicator by converting real pressure and flows back to voltages. An 8 port Compaq based puti duplicator is currently being constructed, o Q t 588 B~UKE Do not copy or show to unauthorised persons. BATCo document for Legal Services : Health Canada 19 May 1999 to align the work throughput of the recording and duplication facility, In this system the duplica~on ports will be capable of independently programmed operation if necessary, ~e disadvantages of the C~mpaq based recording system are not of afunctional nature, but are concerned with the intrusion which any measuring system poses on normal human behaviour. It is evident (3) that smoking is a social habit and that smokers are more relaxed (a) in the presence of other smokers and (b) outside of the confines of a laboratory situation, The abiii~ to record in groups compared with a monadic test design is therefore preferable in terms of logistics and repiesentativeness of the data collection situation, (b 1988 BATUKE Do not copy or show to unauthorised persons. BA~Co document for Legal Services : Health Canada 19 May 1999 REFERENCES f. BA.T Report No. R0.1490·R, 14.6,1977 2 B~A,T Report No~ RD,1199·R, 28A,1915 3. Smoking Psychology and Pharmacology, Heather Ashton and Rob Stepney, Tavistock Publications, 1982 6 t988 BATUKE Do not coW at show to unauVlorised persons. BATCo document for Legal Services : Health Canada 19 May 1999 APPEHDIX 8 1988 BATUKE Do not copy or show to unauthorised ps~~~ BATCo document for legal Services : Heath Canada 19 May 1999 25 RGURE1 rl ~1 ,, h t ,, t rI r c r E~L~ta~~ IsP~ .IlCT h I s Y I 'I LII1' 11 ii a I b O r O O N r O 3 19~ BATUKE Oo nd mpy ~ shaw ta unautharised parsans. BATCo document for Legal Services : Hoe~h Canada 19 May 1999 3co T b O C-u (r] 2d i W a PI Q L e w C r u ZLL~ 3 O O r o L E g 391 w r a NT 3 O g or c o U3 Z 1 Qn t W BA'TCo document for legal Services : Health Canada 19 May 1999 FIG. 3 T, 203 R OATA ACaUISITION BLOCK DIAGRAM COMPUTER OATA O1SK STORAGE INS1RUMENT OATA INPUT OAT`A SCREEEI OISPIAY HARDWARE INTERFACE BATCo document for Legal Services : Health Canada 19 May 1999 I TS-91Shqohq??Nqqqqq?~?~q~?Nqqq[~$ PNOhnQBnOOTlt~~~~~nQrOPPnaONnr ~f +~nc~~~~+~~~~ ~~Rlp~~n·~on~ -cm-osnean~*e*oa~noho~*h ?P .,.,.., ...........~... ~~`~~~)ON '~ :d:::8::isl!BC;:P13:~::S::5::Si~~ i > 1 :U oo n~~pOODPlrdONO~rtQOI~~~~nOOMc~ e: -:-: · · · · ,-: ·-I · ·-:·1·1·:-:·; · ' · · · · · · · · · · nnrtrrrnrn*ronnn~Rrrrtl*rrrrrr r", oP·P001N00nNn~n~~r*N~nPINnOn 10 ~~P*R*~0cN00m~NI~OIOnAnNN~ P Nr ICNPrrC~CCNl~trrrrrrrt~frr~~rr UI tr I I I I I ," V) 1 I I I I ) I I-I - I I CO I I i 1 I Wd In ~PrnODNDIPN~~tNlltRPInOhnlNPOr o, ..,,,....,,,,:j..0-SO.L*I.*.NO ~% VJ teau·~a~tn ua~·o~*1*nt*nn* r 3 a I I f I P a rn ~a i. o a ?D II *~P*OP~~O~*~PNlln~rlI~Oh~0 L1 i!~~~ias%~ 8 r~n nhr) E o > O Z" E CP o I O Q w ,, w a E fl tnPDIPloqlq,alolo,,,O~YlhPIIOON~~nrlP ., ~ ~u N~rr,r,,,N;~i~i~~~;;;;;:r;rr~rla_ I " D O m O m O I I I ~I I Y n~ O '" OI~Nr~rOtnltllPNrOnON~nOllO~r~ ;oo OI o~O~?f!P=?BO~NN"'O*OlllOOnNPleBPld JL F *r+rrrr~*rrre+*,,,, *r Y a 'o ODOQdQ·00d00mmlOQlrnO~nOn00h0 nh IL $~ dn~·nc·o~so~~*~o4s·?Cc~9h~An~ ~f oh~ChP~h~olhhhhh~~ooD~~~*0~·1 O 1> o ol I II o to, ,, o r a I I I h) O W a I·· N ~ m oO o ul BATCo document for legal Services : Health Canada 19 May 1999 TABLE 2 SHQRIUIAIIWRS: ibp~2EBeredSys~Wn CompwBawSlaum Pun Numkr Incmi Durb6Pn YOLMP Flown ]IYlrvpl Ouls~iPn Wml Preaaul (999) M Imll tan ~V.Qi ~oc~ (s~l (m~ km W.GI 20 Os 12 9,5 RO 26 164 92 2 18 0I 37,9 156 19 27 2&0 151 3 21 0,7 262 21 27 216 #.O 1 II R't 227 121 1.7 27 228 11,8 5 1.7 Of 2(.3 129 1,7 27 127 6 13 O.B 27.3 Y.B If Os 2.1 7 If 07 a.l w.s In c8 14 111 6 1.7 09 308 16) IL 09 ]08 161 9 1.6 Os 307 167 1,6 On 3R3 16,( 10 1.6 09 21.5 161 1,6 P9 29.3 118 tOT~LS 15,8 1,6 254,7 143 1U 71 263,3 113 HECHS 1,is 076 2547 1l.33 r3l on 2630 #.m S(ORI WICE PIFFS: I\PP92E%I4SYIIM C~mpaqB~PdS~wn hn Humbw lruonol Ou1110on YO~m, R*wn hPrvd aursa'on Yaiumo Pnaun gpcl (IoC) cmn (anW.G,) (wfl (wl Iml ten W,G.1 1 P.O as 521 M1 (LO as 5U 5].3 2 88 as 529 34,7 69 oB 53,6 3 122 of 5a7 #3 04 a7 527 126 7,6 1,0 54,1 321 77 l,d 53.1 316 7.6 0,8 523 326 75 as 508 #B 6 51 0.8 482 31.3 57 0.8 4; 30.1 7 20 on M.S 27.0 5?1 0.7 ~ZB 17 8 52 0.8 ~55 21.0 52 as (59 a3 9 22 as 37.6 52 09 54.8 31.4 10 22 as 56,1 n.7 53 Os 51.3 n,3 XmLS 61,5 8.3 508.5 331.6 620 &Z 509.9 3U,7 ~E~NS 683 083 5295 3216 6B9 182 50,99 3247 I~UGEPUFFS: ~paBandSyslwn Co~lB~ldS$wn Pun Hunmw Inlpnpi turadon Vo4An, Rsaun htwvJ DuniPn YoCml RPsswo Isle) loocl (rnl) lem W.GI Is~e~ Isocl Iml (em W.G) O.d 1.0 428 2V QO 09 43 350 2 23 1.0 2P9 13 1.0 1.6 290 3 20 1.1 80 21,( 22 ~I (61 265 13 1,O 47.0 27,9 25 1,0 21.1 3 25 12 49,6 23,6 24 It 49.3 ~.8 6 21 11 521 30,3 13 ~t 11.~ 29.5 t 11 12 21,6 21 13 50.5 28.6 B 20 tJ 557 3Pl LO If 528 29,4 9 1.9 12 lL2 228 !1 12 252 10 Is 12 46,8 2110 19 1~ 420 27.1 MTALS 19.0 11.4 486.0 281.6 19.11 11.4 1,3 25,7 M~C~E~S 211 1.14 48.80 2&36 218 1,14 4601 p.Si LOHG UAGE ~UR9: ~oplo2EBaodSy#om e~pPqB~odS~slwn I~WHumb* Intmd Dunak, Yok~m, P~un l~1p3 Dvr~n Vo~wn~ Pnwr [soc) rml (anW,G,I (w) (soe~ II~I ion w.B) 0,0 51 91.0 41 RO 3.1 117,9 15 2 &9 33 957 SQ8 68 33 924 49.1 3 12 9R3 146 27 U Bb9 QZ 1 93 id Bli 111 1.5 16~1 (3.1 b 5 oa 21 86,6 rep 11 84,3 (3.8 O 6 127 22 91,6 425 127 32 90.6 41,7 T OS JJ 81 so ,A 13 aa2 14.1 O 6 73 3.3 825 7,4 20 80.4 102 N 9 12 83.3 427 9.1 81.0 112 lp at 51 oil at as u are 4e O XlhUS 821 33.7 890.8 4Sl,f 629 33.8 861.i MERNS 9.12 317 89,CB 4597 Od 338 Cc,J1 C~ O O 1988 BATUKE Do not coW or show to unaulhdrised persons, BATCo document for Legal Services : Heath Canada 19 May 1999