Document 91rBOY72aOjqZwO16p9yYzBb5

p& Reprint & Copyright 8 by Aerospace Medical Association, Washington, DC Environmental Tobacco Smoke (ET) in 2; Airliners-A Health Hazard Evaluation - > _ _* W. A. CRAWFORD,M.B., Ph-D., and L. C. H O L C O ~ Ph.D. .i[h ITS pc .I r .tier $e' :IP LC CRAWFORDWA, HOLCOMBLC. Environmenral robacco smoke (ETS) in airliners-a health ha:ard evaluarion. Aviat. Space Environ. Med. 1991;62580-6. Somo pasrsngors and oiriino attondants havo oxprossod concorn about oxposuro to onvironmontai tobacco smoko (El$)in commorcial airlinos. Jrritotfon of tho .yes ond mspiratory tract, odor, and porronai sonsitivitios a n ofton tho baser of eo plaints. Previous studies domonstrato low concontrationsof 5k r i a l s , including TI.Attondontr fly somo 600 h por ymr, o s d B a d on provious studios and duration of ab- doso of El$ is loss than a cigaretto - yo&.:Nicotino absorption of 0.2 to 0.3 mg in 1 1 0 h flights has 5 S n roportod as uniikoly to h a w phyrloiqlic offocts. Tho rrtainod 'dou' in d a t i o n to tho hypothosor of risks to hooith bawd on tho conflicting nrults of studios on 'pamivo smoking' a m d k u s d . 7lure are additional confounding foe- in aircraft. Wo considor sogrogatlon into smoking and nonsmoking zonos to bo roaronablyoffutivo In mooting tho proforoncos af passongorr. THE LAST 50 YEARS have seen a vast increase in the development of airlines in numbers and qualities, an even greater increase in the number of passengers on a worldwide basis, and an increased number of passengers per aircraft. Relatively recently, jet aircraft have largely superseded piston-engined aircraft. The evolution of long range wide-bodied airliners is another recent development. These developments have taxed the engineering skills of aircraft manufacturers to provide an adequate cabin environment for both passengers and flight attendants. McFarland (12) in 1946 provided the first insight into cabin environmental conditions. Passenger conditions ~ 3 ryeasonzbly be related tc the few standards produced by the U.S. Environmental Protection Agency (EPA) and the American Society of Heating, Refrigeration and Air Conditioning Engineers (ASHIWE). Flight attendants may be reasonably sub- .- From Seaforth, Sydney, N.S.W., Australia (W. A. Crawford); and the Holcomb Environmental Services, Olivet. MI (L.C. Holcomb). This manuscript was received for review in July 1990. The revised manuscript was accepted for publication in December 1990. Address reprint requests IO: Dr.W. A. Crawford, who is a private consultant in occupational and environmental health, 29 Battle Boulevarde, Seaforth, Sydney, N.S.W.,2092, Australia. 580 Aviation, Space, and Environmental Medicine June, 1991 -t.. w,ject to the U.S. Occupational Safety and HealaM.i, istration standards where such are relevant. acceptableconditions are produced for passeogm, the flight attendants will be exposed to si* tions and will differ from passengers in only a fwIb spects: a. They are exposed for more hours per year. b. They are a selected group of workers (unlikep ~ . sengers) who are healthy when first employd, They must pass physical examinations00 ployment. k'.fter subsequent episodes ob s u b tial illness or suspected incapacity, a fitncs fa duty evaluation may be required. Earlier reviews of cabin air quality include those Crawford (1) and Holcomb (5). The trend of sepantig airliners into smoking and nonsmoking zones hps beta reasonably effective in meeting the preferences d pas sengers. However, some passengers have strongly plained of ETS in the areas adjacent to the smolip sections and some cabin staff working in the smdtiq section have expressed annoyance and concern for tkir health and the maintenance of their healthy status. h hibition of smoking is in effect in the continental USA. Canada and Australia and some intra-European flghls. Thejustification for such actions has not been based on scientific air quality monitoring, but on the interpations of the results of certain epidemiologic studies; e.g.. International Agency For Research on CYwer 1987 ( 6 ) .National Research Council (17). This paper describes the results of recent intermtional studies and discusses the hypothetical thnruto 9-the healt of passengers and flight attendants. WC w i l note thatshe scientific data on which decisions bw been made is, at best, meager, and medical and Cpidc mnoionleoxgiisctednatt.a\on passengers and flight crew appearmb W Health Factors Impacted by Air Quality? Passengers and flight attendants have comPl*e about eye irritation, headaches, nose and throat mu- SNVIRONMENTALTOBACCO SMOKE-CRAWFORD & HOLCOMB ,ion and breathing discomfort. Crawford ( 1 ) and HolZomb ( 5 ) have reviewed the literature and reported a Janet7 of factors that may alone or in combination :ause these complaints. These factors are low humidity, )zone, ETS, contact lenses, reduced oxygen pressure ind chemicals such as NOz. aldehydes and polycyclic vomatic hydrocarbons. Events such as fires initiated 3y smoking passengers are rare, but do occur, and the ase of fire resistant or firesafe materials is standard practice in modem airliners (2). Weber (22) reported on the imtative effects on eyes of those exposed to relatively high concentrations of ETS. Using CO as a surrogate, they postulated a tenta- tive threshold limit value of ETS-generated CO of 1.7- 2.3 mg CO/m3. As discussed in greater detail later, recent studies by Malmfors (IO)and Geomet Technologies, Inc. (4) show monitored levels in a range of 0.8- 1.1 ppm. Lategola et al. (7) reported eye discomfort as the most frequent ozone symptom in conditions simulating flight conditions. In descending order they reported headache, nasal irritation and throat irritation. At values less than 40% relative humidity, symptoms include dry mucous membranes and irritated eyes. All fight reports summarized by Crawford (1) and Holcomb (9,as well as the newer reports reviewed, give relative humidity values ranging from 8.5-25%. tistically significant, the 95% confidence interval must lie entirely above 1.0. It is generally accepted that the relative risk must be above 2.0 to be considered impor- tant. Fig. I illustrates significant and nonsignificant ranges of relative risk. Ofthe 13 studies reviewed by the NRC. only 2 demonstrated statistically significant increases in relative risk (Fig. 2). These studies have also been severely criticized by Uberla (20) and Layard (8). among others, for problems with confounding,-misclassification, and bias. As a result of the widespread criticism and lack of consistent results between the studies, the NKC (17) performed a meta-analysisof the studies. A similar anal- ysis also has been performed by Wald (21). These rnetaanalyses have produced combined relative risks of 1.35 for all of the studies and I . 13 for the U.S. studies. Let- zel (9) has evaluated these meta-analyses, the assumptions behind them and their conclusions, and found that the criteria necessary to perform a proper meta-analysis have not been met in these studies. He has further found that of the 1,023 possible meta-analyses, only 24 would produce an overall significant increase in relative risk and that these were dominated by the poorer quality studies. Thus, even the attempts to combine the avail- able studies to help correct for their weaknesses does nor produce a credible estimate of risk due to ETS exposure. Airliner Air Quality Data Several studies of airliner air quality performed since McFarland's have previously been reviewed by Crawford (1) and Holcomb (5). They found that, with few exceptions, low humidity and ozone were the most likely causes of the most commonly reported symptoms. The levels of tobacco constituents that were measured were too low to have caused the reported symptoms. However, few parameters were measured in these studies. Because of this, exposure to many of the con- stituents of ETS as well as other chemicals that may have been present could not be evaluated. Studies on Nicotine and Cotinine Crawford (1) and Holcomb (5) have reported on earlier studies regarding measurement of nicotine in airliner ambient air. Malmfors (10) and Geomet (4) monitored nicotine and reported mean levels between 0.04 f Epidemiological Studies: Environmental Tobacco Smoke During the same period that airline air quality data began to appear, reports released by the U.S. Surgeon Gened (19) and the U.S. National Research Council WRC) (17) suggested that exposure to ETS may result an increased risk of lung cancer. As a result of this, exposure to ETS oii airliners was fia longer a questicn Of irritation, but now represented a possible risk of conhcting a serious, chronic disease. The conclusion of these reports was based on reviews Of the 13 epidemiologic studies that had compared lung Qncer rates between nonsmoking women mamed to mnsmokers with the lung cancer rates of nonsmoking women mamed to smokers. The difference in rates of contracting lung cancer is calculated as a relative risk. %e relative risk calculated from these studies is not a Precise value. As a result, it is often more accurately h t r a y e d as a 95% confidence interval, or the range of wues within which the true value has a 95% chance of I' . = :bnegi included. For a relative risk to be considered sta- I LA BI ,c D, .- SIGNIFICANT NOT SIGNIFICANT IMPORTANT NUT INCONCLUSIVE TRUE IMPORTANT NEGATIVE RESULT Cis. 1. lllurtratlon of rtatirtlcal rigniflcanco. Studios A and show rtatlrtlcaily ri~nlflcant95% Confidonco Intervals. Study b howover, has an I n c r o a d RR that Is small onough to bo consid- wed unimportant. Studies C and D, bocausa tho 95% Confidonco intorval door not excludo tho Null Hypothesis(RR = 1.0), OW not considorod rtatlrtically significant ovon though study C door show a modest immaro In risk. Aviation, Space. and Environmental Medicine June. 1991 581 I ENVIRONMENTAL TOBACCO SMOKE-CRAWFORD & HOLCOMB ODDS RATIO 20 10 TABLE 1. AIR QUALITY STUDIES ON 48 SAS FLIGHTS ( 8 ) (WEIGHTED MEAN VALUES) - Seating Category 8 Parameter BNS BS TNS TS 5 Nicotine (&m') 5 41 21 32 RSP (&m') 60 250 160 220 Carbon Monoxide (ppm) 0.6 1.1 0.8 I 3 Carbon Dioxide (ppm) 1310 1310 1270 1430 -2 Relative Humidity (46) Temperature ("C) 25 25 25 2s 23.9 23.8 23.6 23 1 0.8 0.5 0.3 0.2 0.1 1 STUDY NUMBER 2 3 4 5 6 7 8 9 10,11 12 13, ~~ CASE-CONTROL STUDIES PROSPECTIVE STUDIES Clg. 2. ETS exposure by nonsmokers and lung cancer. Tho 95% Confidence lntonalr from the 13 epldomlologic studies review4 by the U.S. NRC Study (17): 1. Chon and fung, 1982; 2. T r i c k poUl05, 1983; 3. Correa, 1983; 4. Kabat and Wynder, 1984; 5. Buffler, 1984; 6. Oarfinkel, 1985; 7. Penhagen, 1987; 8. Akiba, 1986; 9. Koo, 1984; 10. Loo, 1986; 11. Oarfinkel, 1981; 12. Oillis, 1984; 13. Hlmyama, 1984. pg/m3in the middle of a nonsmoking section to 32 pg/m3 in business class smoking sections. Foliart (3) monitored flight attendant blood nicotine on transoceanic flights. Blood nicotine increased from a mean of 1.6 n g / d to 3.2 nglml in five of six women. The authors concluded that physiologic effects were unlikely. Mattson (1 1) demonstrated that nicotine in airliner air can be detected as urinary cotinine in the most highly exposed nonsmoking passengers and flight attendants. This confirms the previous data from Foliart (3). Recent Airliner Studies: Is There a Health Risk? In spite of the widespread criticism of the methodologies used to produce long term health claims resulting from ETS exposure, anti-smoking groups and the popular press have continued to fuel the controversy. In part because of this continued controversy, two recent studies by Malmfors (IO)and Geomet (4) of airline air quality have increased the information on ETS exposure on airliners. They also monitored a wider m y of chemicals and uther eilviionmeniai conditions IO wnich one may be exposed during flight. The key findings of these studies are summarized in Tables I and 11. In addition to collecting basic data, Geomet attempted to calculate a lung cancer risk resulting from ETS exposure for various classifications of airline attendants and passengers. This risk estimation, based on measured respirable suspended particulate (RSP)levels Might and assumed exposure durations, assumed a causal relationship between ETS exposure and lung cancer and a relative risk of I .3 for this relationship. As has been shown earlier, both the assumption of a causal relationship and a magnitude of 1.3 for that relationship BNS = Business Nonsmoking; BS = Business Smoking; TNs = Tourist Nonsmoking; TS = Tourist Smoking. TABLE 11. AIR QUALITY STUDIES ON FLIGHTS BY GEOMET (3). 69 Smoking Flights 23 Non-Smohq Flights Parameter Smoking Non-Smoking Section Section .Middle ROW Nicotine (&n3) 13.4 0.04 RSP ( d m ' ) 176 31 Carbon Monoxide (ppm) 1.4 0.7 Carbon Dioxide (ppm) I562 1568 Relative Humidity (%) 15.5. Temperature (T) 24.3. Ozone (ppm) 0.01 0.01 Bacteria (CFU/m3)** 163 131 Fungi (CFUIm'Y 5.9 s.0 0.00 40 0.5 1756* 21.5. 24.1. 0.02 131 9.0 Represents value for entire aircraft. ** These were taken at the end of the flights and may not be reprt sentative assays. are based upon inappropriate evaluation of the studie involved and are not widely accepted within the scier tific community. In their calculation, Geomet also ne glected to consider retained dose and used instead er posed dose. They also used incorrect values whe calculating respiration rates and employed a phenomc nological risk assessment model not generally accepte in the scientific community. Geomet calculated est mated lifetime risks of lung cancer per 100,OOO nor smoking attendants for domestic flights at between 12 17 persons. Geomet calculated 48.9 pg/h per person of RSP. U. ing a respiratory rate of 0.62 for women, this is correct1 1438 pglh of RSPs. Assuming 11% retention results i a dose of 1.6 pglh. This recalculation reduces the po tulated risk by 31 times, giving between 0.40-0.57 cancl cases. This would not be discernible wit-hh a populatio, In an erratum, Geomet also estimated the lifetime ri: of deaths due to lung cancer that can be attributed I radiation exposure on domestic flights at between % 1,026 per 100.0oO persons exposed. In summary, flight attendant lifetime risk of lung ca cer deaths per 100,OOOexposed on domestic flights wa ETS risk calculated by Geomet Corrected ETS risk values Radiation risk corrected in erratum 12-17 0.40-0.57 90- 1026 582 Aviation, Space, and Environmental Medicine June, 1991 ENVIRONMENTAL TOBACCO SMOKE-CRAWFORD & HOLCOMB One can use the correct values for respiration rates and exposure and, as is toxicologically correct, calculate retained dose of the constituent of concern. This will provide a much more accurate estimate of the magnitude of the exposure one may receive under the conditions measured in the Geomet study. The results of bese calculations are in Table 111. ' TABLE 111. CALCULATION OF ABSORBED DOSE OF RSPs FROM ETS. , Calculation of absorbed dose utilizing the data on RSPs (23.2 I &m3 on average) from the Geomet study based on the percentage of time that flight attendants were exposed to different i concentrations of exposure in different aircraft sections. i TO CALCULATE AB!WRBED DOSE RSP x Respiratory Rate x Absorb. (%) x Hours Per Year = Absorbed Dose WOMAN: 23.2 pg/rn' x 0.62 m3 x 0.11 x 960 = 1519 pg = 1.52 my MAN: 23.2 &rn x 1.08 m3 x 0.11 x 960 = 2646 pg = 2.65 mg TO CALCUWTE CIGAJWITE EQUIVALENT: Tar DeliveryiCigarette to Smoker = 10.6 mg (Woman) 13.2 mg (Man) Absorbed Dose Nonsmoker Absorbed Dose Smoker = Cigarette Equivalent -WOMAN 1.52 = 0.14/Yeat 10.6 -UAN 2.65 = 0.uYYeat 13.2 500 OSHA-USA; AND UK STANDARD 0 E \ a0, w 2 F -0 0 z 50 I.40 30 20 410 - MURAMATSU OLDAKER SAS GEOMTT; (13,141 AND 1989 U.S. STUDY CONRAD (10) (18) (4) ria. 3. Nicotino moasuroments cor;pard to standards. Air Quality Standards: Ventilation Several health-related organizations both American Od international. have developed exposure standards b protect the health of those exposed. The most com- F n parameters measured in the airline air quality stud- ?that have been performed are nicotine, carbon mon- nride, ozone, and carbon dioxide. The two parameters h may be related to ETS are nicotine and carbon Ponoxide. Fig. 3 and 4 show that the data derived for C s e parameters represent but a small fraction of the Rposure standards. Of the two remaining parameters, @one concentrations have exceeded the exposure stan- s in the earlier studies (FAA, in NAS (16); Nastrom but not the Geomet (4) study (Fig. 5). dioxide has been snown to exceed both the and the Japanese exposure standards in all ies where it was measured (NRC, (16); Malm;Geomet (4); Fig. 6). This standard has been ped, not as a health-based standard due to the ty of carbon dioxide, but as a surrogate for proper ation. ASHRAE recommends 20 cfdperson of air for buildings. It is obvious from Fig. 7 that as air ventilation decreases on airliners, Cot in- Airlines are beginning to reduce fresh air quanRC p. 141) (16) and recirculation of filtered w is available in modem airliners. Overall air y inevitably suffers as a result (this may be due to the density of passengers in airliners being greater than that of workers in offices). Geomet investigators(4) attempted to use a tracer gas technique for estimating ventilation rates, but had technical problems since the technique, while adequate for aircraft with recirculation systems, was inadequate for other aircraft. If one calculates fresh air based on the C02 concentration inside and outside the aircraft and the CO, generation rate, the results demonstrate that fresh air was provided at less than 50% of the US. ASHRAE standards; Le., less than 10 cfm/person. Proper ventilation is the primary factor in controlling exposure to other constituents in an indoor environment. Theoretically, the potential for increased air exchange rates is 2-4 times that in an office. This should provide good air quality, but in practice this was not found. Carbon dioxide levels rose to some 1,300 ppm, indicating tnat fresh air was not provided at suificient rates to provide recommended air quality. Discussion Additional studies of cabin air quality clearly need to be conducted. There are variations between aircraft types and variation within types caused by differential operation of the ventilation systems. Therefore, much more data need to be acquired to understand overall cabin air quality. With such information air quality standards for airliners could and should be established. In addition to augmenting the type of data so far acquired, there is an absolute need to obtain data on other chem- Aviation, Space, and Environmental Medicine June, 1991 583 I ENVIRONMENTAL TOBACCO SMOKE-CRAWFORD & HOLCOMB 50 OSHA-U.S. WORKPLACE STANDARD ( STEL 400 * * ) 0.6 35 e.g. ASS-SWEDEN 0.5 0.4 I c 0.3 9 EPA-U.S. ENVIRONMENTAL STANDARD 0.2 1.5 1.o 0.5 `Suggested TLV for CO as surrogate for ETS; 1.7-2.3ppm 0.1 .I0.0 FAA; GEOMET; NASTROM SAS WEIGHTED MEAN (10) GEOMET MEAN; U.S. STUDY (4) 11% EXCEED U.S. STUDY (15) 0.25 NRC (16) (4) Fig. 5. Ozone mearuromonts compared m standards. `EXPRESSEDASEXCESSCO "Short Term Exposure Limit Fig. 4. CO mwsures compord m standards. icals of perhaps a more potentially hazardous nature. The potential hazard to health is related to the type and concentration present, the durations of exposure, and the retained dose. A disturbing absence of epidemiologicinformation on flight attendants and passengers also exists. These data should be supplemented by objective eye initation studies such as those of Weber et al. (22) in nonsmoking and smoking sections and also in nonsmoking flights. Flight attendants are mainly concerned about the studies on the lung cancer status of nonsmoking spouses of smokers. The results of these studies are highly contentious, contradictory and largely not statistically significant. All 23 studies so far reported have been confounded, as were those selected for study by the NRC (17). Significantly, the extrapolations from these confounded epidemiological studies are being used to estimate the hazard to airline staff. Equally significantly, erroneous dose data are being used in such extrapolations and risk assessments. One questions Hmether the potentially low risks found in spousal studies can be applied to estimate a lower potential risk to flight attendants and the lowest potential risk for passengers. Crawford ( I ) reviews the conditions and possible risk that flight attendants and passengers face in their nonflying hours at their homes and in overseas resting periods between flights. In aerospace medical practice over the last 50 yea the progress of knowledge and application of th knowledge has been based on medical physiology ar the sciences. With regard to the health effects hypot esized for exposure to ETS,scientific data are quan tatively and qualitatively inadequate to estimate haza and, equally importantly, to justify prohibition of smo ing in airliners. In our view the classical `Type 1` error (Le.. false concluding that a health hazard is present when in fz it is not present) is being committed through the pr mulgation of regulations based on unscientific juc ments. Conclusions Airline cabin air quality is determined by a combir tion of chemical and physical factors, not the least which is proper ventj!ation. Low humidity, ozone, E l and reduction in partial pressure of oxygen levels hl: all been implicated in the effects that are felt in 1 upper respiratory tract and eyes. Except for carbon oxide and, sometimes, ozone, airline cabins meet . standards for most occupational settings. However. air quality of airline cabins could be improved by creased ventilation and filtration. The very low levels of ETS in airliners do not aPP to pose a measurable risk to health of passengers flight attendants. If one looks 5all the studies pertaining to ETS posure on airlines, the following conclusions can drawn regarding health: 584 Aviation, Space, and Environmenlal Medicine June. 1991 ENVIRONMENTAL TOBACCO SMOKEARAWFORD & HOLCOMB 5000 [ OSHA-USA-8 HRS. STANDARD E P P \ 0" 0 1000. ASHRAE-U.S.; JAPAN U.S./NRC (16) SAS GEOMET; (10) U.S. STUDY (41 Pig. 6. C 0 2 moarunmonts compared to standards. 1700 1500 E P fL1300 0" 0 1100 900 25 2 0 15 10 5 cfm/PERSON 1 8SAS STUDY (10) GEOMET STUDY (4) pig. 7. Ventilation impact on C 0 2 Iovols. 1. As has been shown previously by Crawford (1) and Holcomb (5). the levels of ETS constituents are not high enough to result in the symptoms most commonly reported by airline attendants or passengers. These symptoms are much more likely to ': result from elevated levels of ozone or abnormally low relative humidity, neither of which is uncommon on long flights. 2. The scientific data do not support a causal reIationship between ETS exposure and the develop ment of chronic health effects. The operational value of 1.3 for an overall relative risk is likewise the result of misapplication of the available data. 3. The ETS health risks calculated by Geomet (4) are based upon faulty assumptions and inaccurate data. Thus, the ETS health related conclusions reached by that study must be considered highly suspecr. 4. The lack of data does not allow evaluation of exposures to a wider variety of environmental agents, and prevents any evaluation of potential health risks due to exposures resulting from air travel. We suggest, at a minimum, that the follow- ing parameters be evaluated: benzo(a)pyrene, kerosine, fuel vapors, aldehydes, other polynuclear aromatics, bacteria, fungi, viruses. 5 . None of the epidemiologic studies conducted on ETS exposure evaluated exposure scenarios that are relevant to the airline environment. We suggest that epidemiologic studies of airline flight attendants be carried out as a more direct measurement of the health risks incurred by this professional group. ADDENDUM In Aviarion, Space, and Environmental Medicine June 1990; 6115.3142, Drake and Johnson report on six flights in 1987. Their data on nicotine and respirable particulates preceded the results reported herein but were not available before this paper was written. The title of their report is "Measurements of Certain Environmental Tobacco Smoke Components on LongRange Flights." Their data do not conflict with conclusions of our paper. REFERENCES 1. Crawford WA. Environmental tobacco smoke io airliners-health issues. Aerospace 1989; 16:(7):12-7. 2. Department of Transportation. Report to Congress: Airline cabin air quality. Washington, DC: Department of Transportation: 1987. 3. Foliart D, Benowitz NL, Becker CE. Passive absorption of NCotine in airline flight attendants. N. Engl. J. Med. 1983; 308:1105 (letter). 4. Geornet Technologies Inc. Airliner cabin environment: contaminent measurements. health risks and mitigation options. Washington: U.S. Department of Transportation, 1989. 5 . Hokomb LC. Impact of environmental tobacco smoke on airline cabin air quality. Environ. Tech. Lett. 1988; 9 5 W 1 4 . 6. International Agency for Research on Cancer. In: O'Neill IK. Brunneman KD, Dodet B, Hoffman D. eds. Environmental carcinogens: methods of analysis and exposun measurement. Vol. 9 , Passive smoking. Oxford, UK: Oxford University Press, 1987:6-4. 7. Latqola MT,Melton CE. Hisgins EA. Effects of ozone on symptoms and cardiopulmonary function in a flight attendant sumogate population. Aviat. Space Environ. Med. 1980; 51:237-46. 8. Layard MW.Environmental tobacco smoke and cancer: the cp idemiologic evidence. In: Echobichon DJ,Wu JM,eds. Envi- ronmental Tobacco Smoke: hweedings of the International Symposium at McGill University 1989. Lexington: Lexington Books, 19%9:99-115. 9. Letzel H, Blumner E, Uberla K.Meta-analysis on passive smok- Aviation. Space, and Environment01Medicine June, 1991 585 ENVIRONMENTAL TOBACCO SMOKE-CRAWFORD & HOLCOMB ing and lung Cancer: effects of study selection and miscbsifi- cation of exposure. Roc. Indoor Ambient Air Qual. Lond. 1988:293-302. IO. MMors T, Thorburn D,Westlia A. Air quality in passenger cabins of DC-9 and MD80 aircnft. Environ. Tech. Lett. 1989; 10:613-28. 1 I . Mattson ME, Boyd G. Eyar D, et ai. Passive smobng on commercial airline flights. J.A.M.A. 1989; 261:367-72. 17. National Research Council, Committee on PPssive smob Board on Environmental Studies and Toxicology. mental Tobacco Smoke. Measuring exposures sad nsWssrn health effects. Washington. DC: National Academy pres 1986. IS. Oldaker GB,Conrad FC. Estimation of effect of e n v k n t tobacco smoke on air q d t y within passenger cabins of cor medal aircraft. Environ. Sci. Technoi. 1987; 21:%9. 12. McFarland RA. Human factors in air transport design. New YO^: McGmw-Hill. 1946. 19. U.S. Surgeon General. The health consequences of ~ v smoking. Rockville. MD. U.S. Department of Healtb and H 13. Mutamatsu M. Umemura S. Okada T,Tomita H. Estimation of man Services, 1986. personal C X ~ ~ S U tRo tobacco smoke with a newly d e v e W d 20. Ukria K.Lung cancer from passive smoking: bypathesis or co personal nicotine nonitor. Environ. Res. 1984; 35:218-r]. vincing evidence? Int. Arch. &CUP. Environ. Health 1% 14. Muramatsu M. Umcmura S, Fukui J, Arai T, Kira S. Estimation of personal exposure to ambient nicotine in daily environment. Int. Arch. Occup. Environ. Health 1987; 59545-50. 15. Nastrom GD. Holdeman JD, Perkins PJ. Measurements of cabin and ambient OZOM on E747 airplanes. J. Aircraft 1980; 1 7 246-9. 16. National Research Council. Committee on Airliner Cabin Air Quality, Board on Environmental Studies and Toxicology. Commission on Life Sciences. The airliner cabin environment: aPirresqsu, a1li9t8y6.and safety. Washington, DC: National Academy 5 9 4 ? 1-37. 21. Wald NJ. Nanchahd K,Thompson SG. CUckle HS.Docs brea ing other poplc's tobacco smoke I X U S ~ lung cancer? Br. Me J. 1986:1217-22. 22. Weber A, Grandjean E.Acute effects of environmental tobac smoke.In: O'Neill JK.BNnncmann KD,M e t B,Hoffma. D,eds. Environmental carcinogens: methods of analysis a exposun measurement. Vol. 9. Passive smoking. Oxford,L' oxford University Ress, 198759-68; International Agency Research on Cancer. .- 586 Aviafion.Space, and Environmental Medicine a June, 1991