Document 6E0GrVK8rGqkOqMabeN6bmNd

-ftariosH' lE-r>isnl aSUT, '.oa I I Environmental Health Perspective* Vo1.11, pp. 137-113, mi UTt*rPh -ilWi-Ne* ;ias :: fiv#< ..bifron :!&*****' v.^5tfTT.^*nrin ;>:iofCg, =iS"/.Jv.*c3T -t*- / 7 I Mortality Among PVC-Fabricating; Employees i by Leonard Chiazze, Jr.,* and Lorraine D. Ference* fi>5 >-v? L* ' TKe results of a cross-sectional mortality study of 3647 deaths occurring among current and former (white) employees of 17 PVC fabricators during 1964*1973 are presented. Sex-race-causespecific proportionate mortality ratios (P.MR't) were computed by using tao separate standards: one. the U.S. mortality in 196S; the second, U.S. mortality for the individual years 1964*1973. In addition, a case-control analysis, based upon 44 breast cancer deaths among white female employeesris presented. PMR's are significantly different from unity for all cancers, and for cancers of the digestive system among both white males and white females. Although observed deaths significantly exceeded expectations for cancer of the breast, a subsequent case-control analysis reveals no statistically significant relative risk* for breast cancer. Introduction A follow-up to the cross-sectional mortality in In March, 1974, Organization Resources Coun vestigation, utilizing deaths among white women for whom the underlying cause of death was cancer selors (ORC) was requested by representatives of of the breast, was undertaken (5), and results of the PVC producers who are members of ORC's Occu case-control analysis will be presented. ?Ai*r*; pational Safety and Health Standards Group, to carry out a study of health risks to employees working in the PVC fabricating industry. After considering a number of alternative study designs, Materials and Methods it was decided that a cross-sectional mortality Although alternative study designs were consid 'rV study would best meet the need for providing ered, several factors ied to the derision to carry out information as rapidly as possible. Details regard a cross-sectional mortality study. First, the pri ing reasons for selecting this type of study, the mary study objective was to determine relatively actual study design and results of the inquiry have quickly whether any angiosarcoma deaths could be been reported previously (1). identified among the study group and this was best 1 In the original report,. proportionate mortality ratios were calculated according to the method of accomplished by examining causes of death among relatively recent decedents. Second, it would not Guralnick (2) and by using the sex-cause-age specific distribution of deaths among U.S. whites in 1968 as have been possible to identify clearly and com pletely (if at all) the cohort of workers necessary for the basis for determining expectation (5). For the a historical cohort mortality study within a reason current report, we have analyzed the data using the ably short period of time. As a result, the study was I Mantel-Haenszel procedure both for determination f of PMR's and in order to test whether PMR's differ based upon 4336 deaths which occurred during the years 1964-1973 among active or retired employees <>- significantly from unity (4, 5). In addition, PMR's of 17 companies engaged in PVC fabrication (Table j are presented using two separate standards as the 1). This report will be restricted to the 3S47 deaths basis for expectation. ' : among whites, since the number of deaths for most causes among nonwhites was quite small, making -`W' Division of Bioftatistics and Epidemiology, Georgetown Uni interpretation of cause specific mortality difficult at versity School of Medicine, 3750 Reservoir Road, N. W., bast. Washington, D.C. 20007. Since the population at risk could not be deter- <s*t es October 1981 ucc 062804 137 Table 1. Distribution of deaths among employees of 17 PVC fabricators by race and sex. 1964-1973. V' Race Total blale Sex Female Unknown Total White Xonwhite Unknown 4336 3847 198 291 3676 3232 174 231 658 693 24 89 2 0 0 2 mined, mortality rates as measures of risk could not be calculated. Rather, results are summarized in terms of proportionate mortality ratios (PMR) calculated according to the method of MantelHaenszel (5). Two separate standards have been employed. The first is the sex-cause-age specific distribution of deaths among U.S. whites in 1968 (1968 Standard). A second standard employs the sex-cause-age specific distribution of deaths among .U.S. whites in each ofthe years 1964-1973 (1964-1973 standard). Cause-specific PjMR's based upon the 1968 standard are adjusted for age, race and sex. Cause-specific PMR's based upon the 1964-1973 standard are adjusted for year of death as well as for age, race and sex. I The Mantel-Haenszel procedure for assessing the ! statistical significance of a difference between the { reserved and expected numbers of deaths for a > cific cause requires tjie construction of a series * ol 2x2 contingency tables as shown in Table 2. Separate tables are constructed for each level of the faetoris) to be adjusted for. If, for example, we wish to compute a PMR for all decedents for cause X, adjusted for age (five age groups) and year of death (10 years: 1964-1973), i.e., 1964-1973 stan dard, we would have 50 2 x 2 tables: age < 35, year of death 1964; age < 35, year of death 1965; . . . ; age < 65, year of death 1973. Observed deaths and j expected deaths are summed over the 50 separate i 2x 2 tables to produce a PMR for cause X, adjusted ! for age and year of death. The resulting age-year of death-adjusted PMR compares the number ofdeaths ; from cause X observed in the study group to the 1 number of deaths from cause X expected in the ! study group, if the proportion of total deaths 1 Table 2. Conttniction of 2 x 2 contingency tablet. Number of deaths in group 1, attributed to Specified cause Other causes Total deaths Study group A; C, N Standard B, D, Nr. Total Af,, Af*, T, ascribed to cause X was the same, age group by age group and year by year, as for the comparison group. The statistical significance of the PMR (a test of the hypothesis that the PMR is 1.0 against the alternative that the PMR is different from 1.0) is evaluated by using the Mantel-Haenszel continuity corrected chi-square with one degree of freedom, Le., 9 (|SA,- - 2E(Ai) | - Vt? Xmh " 2 VarAj where (1) Var Ai T\<Ti - 1) E(Ai) Expected value of A,- (MUNU)I r, The case-control analysis is based on 44 deaths among white females in the study with cancer of the -Vbreast as the underlying cause of death. Controls were drawn from deaths due to diseases of the circulatory system and accidents, and matched by age plus or minus five years and company, if possible. Forty of the cases were matched on both criteria and four on the basis of age alone. The basis for the case-control analysis, then, is the 44 breast cancer deaths and 134 matched controls distributed among eight of the 17 companies. Results Proportionate Mortality Ratios ' Cause specific PMR's adjusted for age are pre-sented for white male and white female employees 4 in Tables 3 and 4. For Table 3, expected numbers of4 deaths are based on the cause-sex-age specific dis-' tribution of deaths among U.S. whites applied to^ the study group. For Table 4, PMR's are adjusted^ for year of death as well as race, sex and age;' Statistically significant departures from unity (i.e.-,'* PMR * 1) are evaluated by using the Mantel^ Haenszel continuity corrected chi-square with onei degree of freedom. Causes of death for which the 1 calculated chi-square indicated a statistically sig-i nificant excess or deficit at the a * 0.05 level ofi significance are indicated in the table. It is impor tant to remember that when carrying out a larged number of significance tests, some number will turn' out to be significant on the basis of chance alone.1 When dealing with a large number of tests for4 PMR's, it may be more appropriate to consider the" Mantel-Haenszel test as a screening device rather^ Environmental Health Perspectives? than givir With th 1968 stanc tically sig cancer, res Table 3. Ob* Cause of dead All causes All cancer Buccal ci' Digestive Stomach Large ir Rectum Liver Pancmu Other an Respirator, Larynx Trachea, Other Breast Female geo Cervix Corpus Ovary Other Male genital Pteauie Other Urinary oro Bladder Kidney Brain xad an Lymph u re Leukemaa Other aid aft Diabcttft Disease* ed <*e Heart Rhiiui hn lid. Vyaa^Mi Ccrehr****** Other mm*11*? Disease* d CholrtiMa** ** uddftdamS* ArodccO. Aoddeea^ AD octetepm ` V..V joov . aEe comparison life test of agunst the from 1-0) is jitontinuity ridom, i.e., *. * yt. - Id) fc H _ ,ni> ,'i* . g- sl'JlTi sfc: i. 3*4 deaths mcerofthe ikr Controls toes of the touched by esmuany, if hed oh both seTT iasis 44 oreast ttfeiributed r,.. 1 fai lv -v >* ^ are preifcr.plovees Ittinbere of pedfic dis4applied to *** adjusted t?;and age. (i.e., tbe Mantel- with one Jf-yhich the firJly sig- level ot " i-c impor*K a large "ill turn alone. * lests for )!?' -U>* *l,c -ether -ervpectives than giving it a strict probabilistic interpretation. With these caveats, we see that based upon the 1968 standard, among white males there are statis tically significant PMR's for all cancers, digestive cancer, respiratory cancer and other and unspecified cancers (Table 3). Among white females, PMR's are significantly different from one for all cancers, di gestive cancers, breast cancer, urinary cancer and other and unspecified cancers. In contrast to white males, the PMR for urinary cancer among white Table 3. Observed deaths and cause specific proportionate mortality ratios (PMR) for employees of 17 PVC fabricators by sex (-white only), 1964-1973 (expectations based on U.S. mortality, 196SIT Cause of death ICDA 8th Revision Male Observed deaths PMR Female - Observed deaths PMR All causes All cancer Buccal cavity and pharynx r Digestive organs and peritoneum Stomach Large intestine Rectum __ Liver Pancreas Other and unspecified digestive Respiratory system Larynx Trachea, bronchus and lung Other Breast ' Female genital organs Cervix Corpus Ovary Other Male genital organs Prostate I Other Urinary organs Bladder Kidney Brain and nervous system Lymphomas Leukemias Other and unspecified cancers Other causes Disbetes Diseases of circulatory system Heart disease Rheumatic heart disease Hypertensive disease Ischemic heart disease Myocardial infarction Cerebrovascular disease Other circulatory Diseases of respiratory system Cirrhosis of fiver Cholelithiasis, cholecystitis and cholangitis Accidents, poisonings and violence Accidents Suicide All other causes 000-999 140-209 140-149 150-159 151 153 154 155 157 152, 156,15S, 159 160-163 161 162 160, 163 174 180-184 180 181,182 183 184 185-187 185 186, 187 188-189 188 189 191-192 200-203, 208, 209 204-207 170-173,190,193-199 250 390-458 (390-398), 402, 404, (410-429) 893-398 400-404 410-413 410 430-438 450-45S 460-519 571 674-575 800-999 800-949 950-959 Residua] 8252 670 15 212 41 72 26 6 87 80 206 9 194 8 0 _ ___ -- -- -- 42 41 1 87 20 17 16 44 19 79 43 1930 1495 16 36 1381 898 290 40 144 42 7 187 134 39 229 1.0000 1.1905* 0.8382 1.8058* 1.3000 1.3688* 1.3230 1.4306 1.1307 1.3861 1.1647* 1.0695 1.1704* -- ___ -- -- -- -- 0.7936 0.8223 -- 1.1153 1.0053 1.2800 1.1499 1.3560 0.8117 1.6132* 0.E655 1.0516* 1.0625* 0.6373 1.0575 1.0599* 1.1040* 0.9913 1.5294* 0.6344* 0.6675* 1.1018 0.7007* 0.7013* 0.7427 1.0410 595 1S1 .3 53' . . 8 44 -8 0 7 6. 12 0 12 0 44 19 5 6 7 1 ___ _ 11 4 7 4 9 1 25 10 278 188 7 3 158 93 65 8 17 8 8 46 37 2 57 1.0000 1.3266* ___ 1.5134* 1.6226 1.5785* 2.0822 1 1.1612 1.3680 1.0119 1.0788 1.3710* 0.8232 0.6327 1.2911 0.6714 ,, 2.9296* 3.4535* 1.1900 1.9584* 0.6540 0.90S7* 0.9009 0.7502 0.8644* 0.9SS8 0.9133 1.23&S 0.5804* <p 1.1364 1.3369 1.1500 PMR significantly different from one at a * 0.05. October19S1 ucc 062806 139 females is high and, although based upon only 11 low, but the corresponding number of observed *deaths, is statistically : gnificant. On the other deaths is quite small. hand, in contrast to the observation in white men, Among both white male and female employees, the P.MR for respiratory cancer among white women diseases of the circulatory' system account for a is very close to unity. Similar to the observation in large percentage of total deaths. In each case, Exposure white men. mortality from cirrhosis of the liver is observed numbers of deaths are close to expected, No exposu Impmbabl Possible t: Table 1. Observed deaths and cause specific proportionate mortality ratios (PMR) for employees of 17FVC fabricators by sex (white only), 1964-1973 (expectations based on U.S. mortality, 1964-1973). Definite e? Unknown Cause of death ICDASth Revision ' All causes All oncer Buccal cavity and pharynx Digestive organs and peritoneum Stomach Large intestine Rectum Liver Pancreas Other and unspecified digestive Respiratory system Larynx Trachea, bronchus and lung Other Breast Female genital organs Cervix Corpus Ovary Other Male genital organs t Prostate Other Urinarv organs Bladder Kidney Brain and nervous system Lymphomas Leukemias . Other and unspecified cancers Other causes Diabetes / Diseases of circulatory system Heart disease ' Rheumatic heart disease Hypertensive disease Ischemic heart disease Myocardial infarction Cerebrovascular disease Other circulatory Diseases of respiratory system Cirrhosis of liver -- Cholelithiasis, cholecystitis ar.d cholangitis . ------------ Accidents, poisonings and violence Accidents Suicide All other causes 000-993 140-209 1401149 150-159 151 153 154 155 157 152.156, 158,159 160-163 161 162 160. 163 174 ISO-154 180 181,182 183 184 185-187 185 186,187 -- 1SS-189 IBS 189 191-192 200-203, 208, 209 204-207 170-173, 190,193-199 250 390-458 (390-398), 402, 404, (410-429) 393-398 400-404 410-413 410 430-433 450-458 460-519 (71 574-575 600-999 800-949 950-959 Residual PMR signifiesntiy different from one at e 0.05. 140 Male Observed deaths PMR 3252 670 15 212 41 72 26 6 37 30 206 9 194 8 0 -- -- _ _ --- * 42- 41 1 87 20 17 16 44 19 79 1.0000 1.1667* 0.8340 1.2667* 1.3084 1.3239* 1.3164 0.9801 1.1202 1.3139 1.1155 1.0563 1.1266 * -- -- -- 0.7687 0.7964 1.0942 0.98S4 1.2517 I.IZ67 1.3525 0.8019 1.5975* 43 0.8993 1930 1.0538* 1495 1.0600* 16 0.7030 86 1.1541 1381 . 1.0512* 898 1.U4S* 290 1.0011 40 1.5062* 144 -0.68S3* 42 0.643S* 7 1.1536 187 0.6991* 134 0.7084* 39 0.7034* 229 1.0246 Female Observed deaths PMR (95 181 3 53 8 24 8 0 7 6 12 ,--o 12 0 44 19 fi 6 7 1 -- -- *-- 11 4 7 4 9 1 3Sl 1.0000 1.2956* -- 1.4813* 1.6541 1.5490* 2.1165 -- 1.1461 1.4548 0.9100 -- - 0.9787 -- 1.3448* 0.8161 0.6869 1.2738 0.6597 -- -- -- 2.8365* --* 8.4206* --* 1.1944 -- 1.9564* 10 0.6743 278 0.9132* 188 - 0.9035 7 0.8009 8 --w 158 0.8522* S3 0.9062 65 0.9184 8 1.2256 17 0.6546 3-- 8 46 1.1173 87 1.3266 2 57 1.1197 Environmental Health Perspectives? "'V. although Result in gener dard fT: exceptio 196MK upon Us respinti nifieanth 1964-19; PilR k standard CaseC ;itirilfrT' trtd Mb; A." - .Vfc. ;~W* -isSlp! mss- observed . Table 5. Distribution of 44 breast cancer deaths and 134 matched controls by exposure category. f npk. -es, ^: Breast cancer, cases Matched control mu for a " f Exposure category Number * Number ach case, expected, No exposure Improbable exposure 27 61.4 67 72.4 6 13.6 17 12.6 .tors hr sex Possible exposure Definite exposure Unknown exposure 00 4 2 4.5 6 _9 20.5 10 3.0 4.5 7.5 44 100 134 100 r.MR 1 ("00 1.2956* 1.4M3* l.t-741 1..-490* 2.1163 1.1461 1.4^8 0.!*1<X> 0.!'7t7 l.:H4S* 0.-161 C.<*C9 1.27 rsaca* 3 4306* 11944 I :<564* i< 0743 <*132* t'9035 II *009 a s22* * M2 * 9154 1 2236 ' 6W6 11173 13266 1 1197 ,,.vs although statistically significant. necessary to combine cases of similar age within Results based upon the 1964-1973 standard are, the same set in order to have at least one control It in general, consistent with those on the 196S stan dard (Table 4). There are, however, some notable per matched set. The Mantel-Haenszel procedure was used to derive a summary estimate of relative 1 exceptions. In general, the PMR's based upon the 1964-1973 standard are lower than those based risk and to test for significant departures from unity U, 7). In this procedure, each of the 38 sets upon the 1968 standard. Further, the PMR for can be viewed as a 2 x 2 contingency table. Thus, V respiratory cancer among white males is not sig the ith set can be represented as in Table 6. nificantly different from one when based on the Relative risk is defined as the ratio of the proba 1964-1973 standard. Among white women, the bility of dying from cancer of the* breast among PMR for other (noncancer) respiratory deaths women exposed to PVC to the probability for women is significantly below unity based on the 1964-1973 not exposed. Estimates of these individual proba standard. bilities are not available from a case-control study. . However, a measure of estimated relative risk from ' I1 Case-Control Analysis case-control studies as suggested by Mantel and Haenszel has been calculated as in Eq. (2). T Comparison of cases and controls on a variety of assess whether the departure from unity of an variables where information was available, includ observed relative risk is too great to have occurred ing lenglh of employment,, ever manned versus by chance alone, a summary chi-square test cor never married, and child-bearing history, reveals rected for continuity was performed using the no statistically significant differences (p > 0.05) Mantel-Haenszel procedure. The calculated chi- between cases and controls for any of these vari square must be 3.84 or larger in order to conclude ables. with 95% assurance that the observed relative risk Review of employment histories for cases and did not differ from unity by chance alone. controls revealed a wide variety of jobs ranging from office and clerical work to production jobs such R-li (AA7\)/i (B,C4Tt) (2) as bench inspector, press operator, trimmer, as sembler and sweeper. The wide range of job con Combining the five exposure categories into two tent and location made it impossible to determine may be accomplished in a variety of ways, resulting precisely whether there was PVC exposure in every in several possible relative risk measures as shown case, or the precise length of that exposure. There in Table 7. fore, a subjective ranking system was developed to None of the calculated relative risks, including classify exposures. Work histories were reviewed the second shown which treats anything other than with plant personnel and PVC exposure potential no exposure as definitely exposed, are statistically was categorized into five classes--no exposure, improbable exposure, possible exposure, definite exposure and unknown exposure. This classification Table 6. Representation of /th set. scheme enabled some definitive exposure statement > in 80% of the cases and 92% of the controls (Table S). PVC exposure Yes No Total After matching by age and company, 38 matched sets of cases and controls were developed from the I 44 cases and 134 controls. There are fewer matched sets than cases because, in six instances, it was Cases Controls Total -4. B; -Vit c, D, Si, J/,, A/-, T. October 1981 141 ucc J 062808 I* * significant: i.e.. they may have occurred by chance alone. Similar analyses were canned out on a company-by-companv basis. None of the relative risks so calculated is significantly different from ^ . unity. Discussion and Summary The cross-sectional mortality study was designed with two objectives. The first was to determine if any angiosarcoma deaths had occurred among em ployees of the PVC fabricators under study. Since no angiosarcoma deaths were found among the employees studied, the first question has an un equivocal answer. A secondary objective was that of examining the distribution of deaths by cause among the employees under study. Implicit in that objective is the question of whether or not that distribution is, in some sense, unusual. There is no unequivocal answer to the latter question. Whether or not an observed distribution of Causes of death is unusual clearly relates to the standard or compari son population as well as the analytic methodology (8), This is illustrated by the observation that there are PMR's which are significantly different from unity on the 196S standard and not on the 1964-1973 standard. However, it seems much more important to focus on the large area of agreement between the two standards rather than on the few areas of disagreement. On the basis of the PMR analyses, there are statistically significant excesses in total cancer mor tality among both .white males and white females when compared to the distribution of deaths for the total United States specific for color and sex and adjusted for age. Excesses in cancer mortality ap pear concentrated in cancers of the digestive sys tem and, in particular, in cancers of the intestine for both men and women. In addition, there is a suggestion that mortality from cancer of the breast and urinary organs among white women employees is higher than that for the total U.S. There are, however, several reasons why definitive interpre tation is difficult. Factors meant to suggest that proportionate mortality analyses must be interpre ted cautiously have been reviewed previously {9-13). However, while these results must be interpreted with caution, they appear to be consistent with previously studied workers and suggest the need for some continuing investigation (7). One such follow-up investigation has been pre sented here in the form of a case-control study involving the 44 breast cancer deaths. On the basis of a case-control analysis, estimates of relative risk were derived but none of these relative risk esti mates is significantly different from one, although such results must be interpreted with caution. Absence of a statistically significant relative risk does not demonstrate that there is' not an excess risk of death from breast cancer among women employees with PVC exposure. In fact, when no statistically significant relative risks are found, it is pertinent to ask what the chances are of detecting an increase of a given magnitude from the available data. Using the method described by Walter in), we have subjected each of the relative risk esti mates (Table 7) to a least significant relative risk analysis under the conditions that we desire 95% assurance that a risk of such magnitude, if ob served, did not occur by chance alone and 80% probability of detecting the least significant relative risk if it exists. Even in the case where all but no exposure are counted as exposed, the smallest relative risk which could be detected from these data is nearly 3:1. Nearly 200 cases and 200 controls would have been necessary to detect a true doubling of the risk under the specified conditions. Given the sample size in this study and the percentage of controls exposed (a percentage which was unknown at the .i ! i v. l h \ start of only ve reasons) large in< exist an stateme: in risk. 1. Chii amo 623 2. Gut. Ami Stat Prir 3. Nat the War 1404. Mar ana] Nat 6. Li, W.i 43: Table 7. Relative risk (RR) and least significant relative risk (LSRR) estimated for contrasts of various exposure category combinations. Contrast Number of cases RR (Relative risk) estimate1 LSRR (Least significant relative risk) estimate* I . i Definite exposure vs. no exposure Definite -r improbable possible + unknown exposure v*. no exposure Definite + possible exposure v*. no + improbable exposure Definite + possible + unknown exposure ve. no + improbable exposure 29 44 85 44 1.81 1.94 0.624 2.73 6.77 2.92 6.02 3.41 None of the relative risk estimates are statistically significantly different from unity (p > 0.05). *The true relative risk would have to be at least this large to have an 805 assurance of detection (i.e., power 0.80) with a type I error of 0.05 (i.e., o " 0.05). i 142 r * Environmental Health Perspectives Octobt - jmployees There are, interprejgest that interpreisly (9-13). iterpreted ;tent with .the need been pretrol study n the basis Native risk 3 risk esti>, although aution. dative risk an excess >ng women t, when no found, it is if detecting ie available 4V -(14), 1 . esti"llative risk desire 95% ude, if obte and 80% ant relative xposure are e risk which nearly 3:1. d have been of the risk the sample of controls nown at the >ur* category significant estimateb i.77 2.32 '.<12 1.41 ith a type I Vtspcctivcs start of the study), it would be possible to detect only very large increases in risk. There seems reasonable assurance, therefore, that such very large increases in the risk for breast cancer do not exist among these PVC fabricators, but no such statement can be made for possibly smaller increases in risk. REFERENCES 1. Chiazze, L., Jr., Nichols, W. E., and Wong, 0. Mortality among employees of PVC fabricators. J. Occup. Med. 19: 623-628 (1977). 2. Guralnick, L. Mortality by Occupation and Cause of Death Among Men 20-64 Years of Age: United States, 1950 (Vital Statistics-Special Reports, Vol. 53, No. 3). U.S. Govt. Priming Office, Washington, D.C., 1963. 3. National Center for Health Statistics. Vital Statistics of the United States, 196S, Volume II, Mortality, Part A. Washington, D.C., U.S. Govt Printing Office, 1972, pp. 140-205. .4. Mantel, N. and Haensiel, W. Statistical aspects of the analvsis of data from retrospective studies of disease. J. Naif. Cancer Inst. 22: 719-748 (1959). 6. Li, F. P., Fraumenj, J. F., Jr., Mantel, N., and Miller, R. W. Cancer mortalitv among chemists. J. Natl. Cancer Inst 43:1159-1164 (1969). 6. Chiazze, L., Jr., Wong, O., Nichols, W. E., and Ferenee, L. Breast cancer mortality among PVC fabricators, J. Occup. Med. 22: 677-679 (1980). 7. Pike, M. D,, and Morrow, R. H. Statistical analysis of patient-control studies in epidemiology: factor under inves tigation an all-or-none variable. Brit. J. Prev, Soc. Med. 24:42-44 (1970). 8. Chiazze, L,, Jr. Problems of study design and interpreta tion of industrial mortality experience. J. Occup. Med. 18: 169 170 (1976). 9. Redmond, C. K., and Breslin, P. P. Comparison of methods for assessing occupational .hazards. J. Occup. Med. 17: 313-317 (1975). 10. Monson, R. R., Peters. J. >!., and Johnson, M.>N. Proportional mortality among vinyl-chloride workers. Lan cet ii: 397-398 (1974). 11. Lloyd, J. W., and Ciocco, A. Long term mortality study of steelworkers: I. Methodology'. J. Occup. Med. 11: 299-310 (1969). 12. Redmond, C. K., Ciocco, A., Lloyd, J. IV., sr.d Rush, H. W. Long term mortality study of steelworkers: IV. Mortal ity from malignant neoplasms among coke oven workers. J. Occup. Med. 14: 621-629 (1972). 13. Tabershaw, I. R., and Gaffey, W. R. Mortality study of workers in the manufacture of vinyl chloride and its polymers. J. Occup. Med. 16: 509-518 0974). 14. Walter, S. D. Determination of significant relative risks and optima] sampling procedures in prospective and retro spective comparative studies of various sizes. Am. J. Epidemiol. 105: 387-397 (1977). Octohcr 1981 ucc 062810 113