Document reBNx4ybyVNyMzNqJpm2VagV0

DOW CHEMICAL U.S.A. '1803 Building 4 December 1981 MIDLAND. MICHIGAN 48640 MEMORANDUM TO: M. F. CURRIER POSTON W. E. LEDFORD FROM: G. L. CARLO cc: R. R. Cook Sorry about this, but there are a couple of modifications which should be added to the copy of "A Cross-Sectional Study of Employees With Po tential Workplace Exposure to Ethylene Oxide," which you should have recently received. These are the result of conversations between Ralph Cook and I regarding the draft which was previously sent to you. On page 2 of Abstract (lines 14 and 15): strike out "...and no spe cific future studies were suggested." On page 13 of paper (line 17): strike "Both urine protein and urinary", and also strike lines 18, 19 and 20. On page 15 of paper (line 8): strike "As noted above", and also strike lines 9, 10 and 11. At the beginning of line 12, insert "Elevated,..". Also strike line 15 through the end of the pge. On page 15, also, after line 14, insert: "little Importance relative to potential E0 exposures. Each individual with abnormal urine findings was evaluated clinically. Two of the potentially exposed had sperm in their urine, while one other had acute urethritis. In all others, the findings were transitory in nature and no disease was diagnosed." On. page 16 (lines 7 and 8): strike "...and no specific future studies were suggested." AN OPERATING UNIT OF THE DOW CHEMICAL COMPANY DO 095076 C0N5TDFNTTA/ DOW CHEMICAL U.S.A. 1803 Building 2 December 1981 midland, Michigan 48640 MEMORANDUM TO: M. F. CURRIER P. L. POSTON W. E. LEDFORD FROM: G. L. CARLO' CC: R. R. COOK Attached please find a draft of our report, "A Cross-Sectional Study of Employees with Potential Workplace Exposure to Ethylene Oxide." This, draft reflects the input of the Epidemiology Group whose comments were incorporated after their review of an earlier draft. Please note that Table 3 is only partially complete, since Perry is working on pre paring the missing data. That table, however, has no bearing on our re sults and review without the missing data presents no real problem. If you have no major criticisms, please sign the approval sheet and send it back to me so that I might initiate the CRI review process here. (We need the original signatures here.) If you do have major concerns, please call me and we can discuss them. My feeling Is that, after CRI approval, the paper should be submitted to the Journal of Occupational Medicine for publication. Att. AN OPERATING UNIT OP THE DOW CHEMICAL COMPANY DO 095077 CONFIDENTIAL. APPROVAL FOR OUTSIDE RELEASE OF PROPRIETARY TECHNICAL INFORMATION B * 600 - 1. To be completed by AUTHOR AND sent to SUPERVISOR Manuscript Title: A CROSS-SECTIONAL STUDY OF EMPLOYEES WITH POTENTIAL WORKPLACE EXPOSURE TO ETHYLENE OXIDE Author(s): Carlo GL, Currier MF, Poston PL and Ledford WE. To be presented op^ub 11 shed^>(c 1 rcle one): journal of Occupational Medicine Questions concerning manuscript are to be directed to: Name G. L. CarloDepartment Epidemiology, H&ES Building 1803 City MidlandDate Needed" 18 Dec. 1981 Phone (517) 636-9064 Suggested reviewers: LEGAL ASPECTS Environ. Clearinghouse Product Safety Compliance Patent ______________________ Legal _______________________ Trademarks _________ PRODUCT DEPARTMENT Prod. Dept. _______ Person ___________ Tech. Acq. ________ Person ___________ Other TECHNICAL ASPECTS Tech. Ctr. _______ Person _________ H&ES ______________ Person _________ Other 2. To be completed by SUPERVISOR AND SENT TO REVIEWERS: DEPARTMENTAL APPROVAL FOR RELEASE I have read this manuscript and approve its release at this time as being in the best interests of The Dow Chemical Company. ________________________ _ Immediate Supervisor and Date R. R. Cook, M.D. __________________________________ Laboratory Director and Date 3. To be completed by REVIEWER AND RETURNED TO AUTHOR. Review attached manuscript and" provide comments or recommendations. COMMENTS: ' RECOMMENDATION ( ) Release ( ) Release after changes ( ) DO NOT RELEASE Reviewer's Signature and Date: ___________________________________________________ _________________ J. R. Venable. M.D.___________________ To be completed by RESEARCH DIRECTOR and returned to Central Report Index, 566 Building, Midland. Midland CRI should receive A copies of the approved manuscript. ` APPROVAL BY REVIEWER IS ALSO APPROVAL TO ASSIGN OR LICENSE COPYRIGHT IN THE WORK TO OBTAIN PUBLICATION IN A NON-VOW JOURNAL OR PERIODICAL, UNLESS EXPRESSLY STATED OTHERWISE UNDER "COMMENTS." COPYRIGHT TRANSACTIONS MUST BE PREPARED BY THE TRADEMARK/COPYRIGHT DEPARTMENT. COMMENTS: Final approval, to be given by: RECOMMENDATION ( ) Approved for release ( ) Approved after changes ,( ) NOT APPROVED Signature and Date C-L7906 ft-8;8o R. L. DostaD Do D95078 CONF TDfrnttai A CROSS-SECTIONAL STUDY OF EMPLOYEES WITH POTENTIAL WORKPLACE EXPOSURE TO ETHYLENE OXIDE SIGNATURES; PRIMARY INVESTIGATOR; CO-INVESTIGATORS: Mortimer F. Currier (Date) William E. Ledford (Date) DO 095079 CONFIDENTIAL A CROSS-SECTIONAL STUDY OF EMPLOYEES WITH POTENTIAL WORKPLACE EXPOSURE TO ETHYLENE OXIDE Mortimer F. Currier, M.D. Medical Director Louisiana Division Dow Chemical U.S.A. Plaquemine, Louisiana 70764 George L. Carlo, Ph.D., M.Sc. U.S. Area Epidemiology Health and Environmental Sciences Dow Chemical U.S.A. Midland, Michigan 48640 Perry L. Poston, B.A. Medical Department Louisiana Division Dow Chemical U.S.A. Plaquemine, Louisiana 70764 William E. Ledford, B.Sc. Engineering Industrial Hygiene Department Louisiana Division Dow Chemical U.S.A. Plaquemine, Louisiana 70764 DO COA/fr 95080 rOFNTTA( Acknowledgement We would like to extend our sincerest thanks to the following Individuals for their con tributions to this work: Ken Bodner, M.P.H., Dow Chemical U.S.A., Ralph Cook, M.D., M.P.H. Dow Chemical U.S.A., Jed Diem, Ph.D., Tulane University, and Dirk Van Peenen, M.D., Dr. P. formerly of Dow Chemical U.S.A. DO 095081 CONFTDFNTTAl Abstract A Cross-Sectional Study of Employees with Potential Workplace Exposure to- Ethylene Oxide Currier MF, Carlo GL, Poston PL, Ledford WE This paper describes the results of an epidemiological inquiry con ducted among a cohort of employees of a large chemical manufacturing complex. The first objective of this inquiry was to identify, for futyre reference, the group of workers with potential exposure to ethylene oxide (EO), and the group of workers who have never had potential EO exposure. The second objective was to determine whether the group of current workers with potential EO exposure had a higher prevalence of abnormalities of the hematopoietic, hepatic, renal and reproductive systems, as measured by data collected through periodic health examinations, than did an appropriately matched control group. Ninety-six full time employees with potential workplace exposure to EO were identified through personnel records. One thousand three hundred and forty three full time employees not Involved in the EO processes or in Research and Development or Instrumentation were identified and classified as not exposed. In the cross-sectional portion of the inquiry, eighty-four male employees who had been working for at least 6 months in processes involving EO exposure, prior to the time of their 1980 health examination, were matched to individuals not involved in those processes, on age, hire date, race, smoking history, alcohol intake and date of health examination. Results of hematology (Hgb, Hct, DO 095082 qonftdenttai RBC, WBC, percent lymphocytes) and blood chemistry (total bibirubin, alkaline' phosphatase, LDH, SGPT, SGOT and GGTP) tests yielded no statist ically significant differences between the two groups. Results of kidney function (BUN, serum creatinine, urine protein, RBC and WBC) tests revealed a statistically significant excess of individuals in the potentially exposed group with urine protein levels above the reference range; however, because of other medical conditions present among some of those reported as abnormal, this finding was not considered clinically significant. Cursory review of responses to the reproductive and medical history portions of the health Inventory questionnaire, from a clinical perspective, suggested that these data were noncontributory. Within the limitations of this cross-sectional study design, there were no indications of significant problem areas and no specific future studies were suggested. DO 095083 CONFTDFNTTAL Ethylene oxide (EO) Is an Important Industrial chemical used in the manufacture of such products as ethylene glycol, polyesters, nonionic surface-active agents, heavy duty home.laundry detergents, and glycol ethers (e.g., solvents for surface coatings). 1 It is also used as a pesticide fumigant and a cold sterilant for medical equipment. At room temperature and normal atmospheric pressure it is a colorless gas; at higher pressures it may be a volatile liquid. In the United States, where 43 percent of the world production capacity is held, production , of EO has increased from 1.3 billion pounds in 1960 to 5.3 billion pounds in 1979. Acute intoxication with EO at high concentrations has been recog nized as a cause of eye, respiratory tract and skin irritation, and of 2 central nervous system depression. Suspected delayed effects include headache, nausea, vomiting, incoordination, and electrocardiograph ab normalities. Levels responsible for acute effects have been estimated at above 2000 ppm for very short single exposures and above 200 ppm for 3 continuous 8 hour exposures. The American Conference of Governmental Industrial Hygienists lists an 8-hour time-weighted-average (TWA) con centration of 10 ppm for the E0 threshold limit value (TLV) in its Notice of Intended Changes for 1980. The current OSHA permissible ex posure limit (PEL) is 50 ppm TWA. The first epidemiological inquiry into the possibility of adverse chronic effects resulting from EO exposure was reported in 1964. This cross-sectional study involved 37 EO operators engaged in manufacturing operations, and age-matched control operators working with chemicals other Ethylene Oxlde/Currier et al DO 09&084 CONF tdfntiai than EO. The highest air level of EO recorded was 127 ppm; however, the average levels were less than 10 ppm for the working day. In their study, the EO group reported fewer symptoms and less absenteeisms than the controls. In addition, the frequency of "abnormal" laboratory findings in the study and control groups was equivalent. In more recent years, however, there have been epidemiological studies conducted which have suggested positive associations between EO exposure and human illness. For example, a cross-sectional study of employees involved in cold sterilization using EO, revealed varying frequencies of symptoms such as sore throat among 12 exposed employees, versus no symptoms of any kind in a control group.Cytogenetic studies by these same investigators reportedly showed more frequent sister chromatid exchanges (SCE's) among the 12 employees than among the 11 controls (average SCE's 8.65 versus 6.37). In addition, another investigation reported peripheral neuropathy with decreased nerve g conduction velocity among 4 men after overexposure to EO. Concern also has been expressed regarding the possible carcinogenicity of EO. Results of an industry-sponsored inhalation study showed an In creased, dose-related incidence of mononuclear cell leukemia in male and female Fischer 344 rats.^ Effects were noted at exposure levels of 100, g 33, and 10 ppm. Because of the chemical's alkylating agent properties, 9 and because it is mutagenic in several test systems including Neurospora, plants,^ and the Salmonella typhimurium strain TA 1535 bacterial assay,^ carcinogenicity is biologically plausible. Ethylene Oxide/Currier et al DO 095085 CONF rOfTNTTAI. Recently, there have been two reports from Sweden regarding human malignancies allegedly associated with EO exposure. The first was of a case each of acute myelogenous leukemia, chronic myelogenous leukemia, and Waldenstrom's macroglobullnemia, diagnosed between 1972 and 1977, among employees of a small factory which used EO for sterilization. 12 In a subsequent retrospective mortality study of employees in an EO production facility, the same author reported more deaths from leukemia than would have been expected (3 observed versus less than 0.5 expected) using Swedish national death rates as a standard. 13 Two of the leukemia deaths were caused by chronic -lymphatic leukemia and one was caused by acute myeloid leukemia. The significance of these epidemiologic findings are limited, however, by the small number of observed deaths and the uncertain ty of worker exposure information. In addition, a more recent cohort mortality study by Morgan et al. 14 revealed no deaths from leukemia among EO workers. Because the results of previously reported epidemiological studies of EO exposure were conflicting, the development of an additional occupational epidemiology data base was seen as an important contribution to the under standing of the relationship between EO exposure and possible adverse health effects. This paper describes an inquiry conducted among a cohort of employees in a large U.S. chemical manufacturing complex. The first objective was to identify the group of workers with potential exposure to EO, and a group of workers who have never had potential EO ex posure. The a priori definition of these groups was seen as important in expediting any future EO studies. The second objective was to determine whether current workers with potential EO exposure had a higher preval- Ethylene Oxide/Currier et al DO 095086 CONFIDENTIAL ence of .abnormalities of the hematopoietic, hepatic, renal, and repro ductive systems, as measured by data collected through periodic health examinations, than did appropriately matched current workers who never had potential exposure to EO. The results of this prevalance study were seen as valuable in generating hypotheses regarding specific questions in need of further investigation. Ethylene Oxide/Currier et al - O9607 r&Fio*"1 MATERIALS AND METHODS A. Cohort Identification Process description and exposure information. Within the chemical manufacturing complex, there are two groups of employees who may have exposure to EO. The first group is comprised of those individuals in volved In the production of EO and ethylene glycol. The second group, considered mainly as users of EO, is comprised of those employees in volved in the production of the DOWANOLS* and ethanolamines. Personnel' sampling for EO exposure has been done at various times for both groups of workers by collecting breathing air zone samples on activated charcoal. Samples are desorbed with carbon disulfide and analyzed by flame ioniza tion gas chromatography. All measured eight-hour TWA EO exposures for those employees involved in EO and ethylene glycol production have been well below the TLV. The loading operating technicians have had the highest exposures with a range of yearly average TWA exposures from <1.0 ppm in 1977 to 1.7 ppm in 1980. Among this group the range of Individual measurements has been <.1 ppm to 5.7 ppm. All other j obs have had yearly average TWA exposures below 1.0 ppm. Grab-samples taken since 1973 to determine peak exposure levels have indicated excursions for the loading job as high as 235 ppm (in 1976), while all other jobs had peak exposures below 100 ppm, with most being below 20 ppm. Since the loading job requires wearing a supplied air respirator, the exposure measurements for this job represent only potential exposures. Other chemicals to which EO and *Traderaark of The Dow,Chemical Company Ethylene Oxide/Currier et al DO 095088 C0N5IDFNTTAL 5. ethylene glycol production employees might be exposed include ethylene dichloride (EDC), DOWTHEEM* A (biphenyl and biphenyl oxide), and ethylene glycol. Exposure levels for all of these have been well below theirrespective TLVs. The only one of any substantial toxicity, EDC, is used in very small quantities (less than one 55-gallon drum per month), and is usually handled by only one job classification for approximately one minute every other day. The highest recorded EO exposure among personnel involved in the production of the DOWANOLS* and ethanolamines was 0.1 ppm during the bleeding down of some EO containing equipment in 1980. Other chemicals to which these employees might be exposed include propylene oxide, methanol, butanol, ammonia, the mono-, di- and tri- ethanolamines, and the DOWANOLS* EB, DB, TBH, PM, DPM and TPM. For the most part, 8-hour TWA exposures to these chemicals are generally below 1.0 ppm. Definition of the populations potentially exposed and not exposed to EO. Figure 1 outlines the procedure through which the potentially exposed and unexposed populations were identified. The potentially exposed group was comprised of the full-time employees ever involved in the production of EO, ethylene glycol, the DOWANOLS* or ethanolamines as identified through per sonnel lists. The group of individuals who did not meet the work history requirement were considered as candidates for the unexposed group. Elimi nated from this group because their overall exposures would likely be ill-** **Throughout this report, reference is made to the "potentially exposed" population. Exposure measurements were not carried out on every individual Included in this study; therefore, actual exposure could not technically be documented. Ethylene Oxide/Currier et al 00 095089 confidential defined, were those who were not full-time employees, and those who were involved in either Research and Development or Instrumentation. Thus, the total number of employees considered unexposed was 1343. Ethylene Oxide/Currier et al 0950^0 DO C,ONF B. Cross-Sectional Health Inventory Study The major analytical portion of this research involved the assess ment of whether a differential prevalance of certain abnormalities of the hematopoietic, hepatic, renal and reproductive systems existed between the workers potentially exposed and the group of appropriately matched unexposed workers. A readily accessible data base suited for this cross-sectional analysis was that maintained as part of the health inventory program, which is a voluntary annual health examination offered to all production employees of the manufacturing complex. It includes both a medical history and a laboratory work-up. Potentially exposed and Unexposed groups. Extracted from the over all list of potentially exposed (as defined previously), were all em ployees who had been working for at least 6 months in the processes involving EO exposure prior to the time of their 1980 health examination. Employees with previous service in these areas were not included in the potentially exposed study group, but neither were they eligible to become matched unexposed controls. Females were not included because there were so few, and it was also necessary to exclude those for whom complete 1980 health examinations data were not available. Thus, from the 96 individuals In the original pool of potentially exposed, six females and six males with Insufficient health inventory data were eliminated. The 84 employees remaining represented 93.3 percent of the total potentially exposed group. Ethylene Oxide/Currier et al DO 095091 CONFTDPMr TA( 8. Eliminated from the list of unexposed were 127 females, 75 individuals with less than 6 months employment within the chemical complex, and 87 individuals judged to have potential exposure to benzene (benzene is sus pected of having hematopoietic effects' and thus might be considered a competing exposure). Another 93 individuals were eliminated because they did not participate in the 1980 health Inventory. The remaining 939 indiv iduals eligible for control matching represented 91.1 percent of the original unexposed pool. A single unexposed control was then matched to each member of the. po tentially exposed group. For each employee in the potentially exposed group, a search of the entire group of controls was initiated to determine matching candidates. If a control candidate was previously selected as a control match, he was excluded from the process for any subsequent matching. The matching criteria were, in order, as follows: age (within 5 year intervals), hire date (+ 5 years), race, smoking history (non-smoker, current cigarette smoker, ex-smokers and cigar or pipe smokers combined), and alcohol Intake (light, moderate, heavy). The entire list of control candidates for a partic ular potentially exposed individual was thus compiled. As a final matching step, the date of the 1980 health inventory exam was utilized to control for common seasonal occurrences such as influenza, common colds, etc., which might impact on the outcome variables to be studied. In two cases, appropriate matches could not be found. In one instance, the date of hire criterio'n was relaxed from + 5 to + 6 years for the control, and in another instance, the age group criterion was relaxed to match a 19 year old in the 15-19 year age group with a 20 year old in the 20-24 year age group. Ethylene Oxide/Currier et al no 095^? CONF 9. Analytical variables. The analytical outcome variables were extracted from records of results of the 1980 health inventory examinations. The laboratory test variables, their reference ranges, and the suspected direc tions of the change in value if an effect is present, are listed in Table 1. These tests were considered those most likely to be influenced if EO'Vas, in fact, causing health problems based on a review of the EO related litera ture. Table 2 lists other non-laboratory test variables, available from the questionnaire portion of the health inventory, which might be expected to be influenced by EO. Unlike the laboratory analyses, where the results are at least applicable to the day on which the samples were collected, and thus temporally associated with the potential exposure, it was not pos sible to determine whether these historical data had any temporal relevance at all to the potential exposure. Because of the historical nature of these data and their quality (based on subjective recall), no rigorous statistical analyses were applied to them. The exposure variable, or independent variable, was current Involvement in the production of EO, ethylene glycol, the DOWANOLS* or the ethanolamines as described above. The matching criteria described above made it possible to control for the potential confounding variables age, hire date, race, smoking and alcohol consumption history, and date of health inventory. Statistical analysis. Because using matched controls made the overall study population appear as ''pre-dose" (unexposed) and "post-dose" (poten tially exposed) populations, and because normal distribution was not neces- Ethylene Oxide/Currier et al D0 095093 COMF1 OF >slTT 10 sarily assured, the non-parametric Wilcoxon Signed Rank Test^"* was utilized as the analytic statistic for the continuous data. The value for the un exposed subject was subtracted from the value for the potentially exposed subject and both the positive rank sum .statistic (T+) and the large sample approximation (T*) to the normal distribution were calculated. A p-value was then calculated based on T* to test the hypothesis of a zero median difference between the two populations. The continuous variables analyzed in this manner were: Hbg, Hct, RBC, WBC, percent lymphocytes, total bibirubin, alkaline phosphatase, LDH, SGPT, SGOT, GGTP, BUN and serum , creatinine. In addition, because of the clinical implications of values outside of a given reference range, these data were also analyzed as dichotomous or multichotomous variables. For those variables where the im plications were important for values either above or below the reference range, (i.e. Hgb, Hct, RBC, blood chemistries and kidney functions), each value was categorized as being normal (within the reference range) or abnormal. A p-value was then calculated, based on the number of discordant pairs ("potentially exposed" versus "unexposed") which indicated more abnormals among the potentially exposed, to test the hypothesis that the values were the same in both groups (Exact Probability Test 15 ). Urine protein, urine WBC, and urine RBC, were also analyzed in this manner. Because the implications of values both below and above the reference range for WBC and percent lymphocytes were thought to be important, the Maxwell-Stuart extension of McNemar's test`d was utilized to produce a chi-square statistic from which a p-value could be derived pertinent to a multl-chotomous stratification. Ethylene Oxide/Currier et al DO 095094 CONFIDENTIAL 11. Results A. Cohort Identification for Future Follow-up. Table 3 gives the frequency distribution by race, sex, and age for-both the group of employees ever in volved in the production of EO, ethylene glycol, the DOWANOLS* and the. ethanol amines, and the group never involved in such production. For the purpose of reference for future studies, the first group can be considered as those potentially exposed to EO, and the second group as those probably unexposed. These descriptive statistics are important in determining the proper method ological approach in future studies regarding EO exposure. Both the lists of potentially exposed and unexposed were maintained for future access as necess ary. B. Cross-sectional Health Inventory Study. As Tables A and 5 indicate, the process of matching the selected potentially exposed and unexposed individ uals was very efficient. There were no striking differences between the potentially exposed and unexposed groups chosen for analysis, for either the matched variables or selected unmatched variables. In Tables 6, 7 and 8 are the results of the hematology, kidney function and blood chemistry analyses. For interpretive purposes, it should be noted that the higher values for the large sample approximation in the Wilcoxon test (T) indicate that the median value of the potentially exposed group was larger than that in the unexposed. For Hgb, Hct and RBC, the critical test area was at the lower end of the probability curve; therefore, the lower values were of Interest. For WBC and % lymphocytes, both low and high values were of interest; therefore, two-tailed significance tests were applied. For all other variables where the Wilcoxon test was utilized, a one-tailed test considering the higher values as critical was applied. Ethylene Oxide/Currier et al 'D0 095095 CONFIDENTIAL Also, the Wilcoxon's test applied to the continuous data was interpreted as indicative of physiologically significant effects, while the exact probability and the Maxwell-Stuart tests, applied to the categorical data, were interpreted as indicative of clinically significant effects. Table 6 shows that none of the differences in hematology between the potentially exposed and unexposed groups approaches statistical significance. Similarly, the results of the blood chemistry analyses (Table 8) show no statistically significant differences between the potentially exposed and unexposed, from either a physiological or a clinical perspective. In Table 7, however, it can be seen that a statistically significant excess of individuals within the potentially exposed group had abnormally high urine protein levels. Of eight in stances where the values for a potentially exposed individual and his matched control were discordant, seven had the abnormal value reported for the potentially exposed individual. Approaching statistical signif icance was the excess of individuals within the potentially exposed group with urine WBC above the reference range. Both urine protein and urinary WBC were reported and analyzed as nonparametrlc data; therefore, nothing can be said regarding the physiological significance of these findings. The clinical implications, however, will be addressed in the next section. Cursory review of the responses to the reproductive and medical history portions of the health inventory questionnaire, from a clinical perspective, suggested that these data were noncontributory. Ethylene Oxide/Currier et al CO 095096 CONFIDENTIAL Discussion Cross-sectional or prevalence studies are traditionally utilized as hypothesis generating instruments. Because disease incidence is not the outcome parameter, (i.e., not possible to determine whether exposure preceeded effect) the cross-sectional study is not appropriate for defini tively investigating cause and effect. This type of study is, however, generally believed to be capable of suggesting the presence of profound cause-effect relationships, given a sound study design. In this investigation, the methodological rigor was enhanced by the incorporation of the one-for-one matching of potentially exposed with un exposed individuals. However, with the inclusion of the date of health inventory as the last matching step, an additional potential bias was encountered. Because entire plants within the chemical complex were generally scheduled for the examinations each month, the matched control group could have potentially included a disproportionately large number of individuals from a particular plant or plants. If that were the case, the comparison studied might be of potential exposure to ethylene oxide and potential exposure to another chemical. To address this possible problem, the distribution of plant-specific job titles for the entire chemical complex was ascertained, and compared to the similar distribution among the matched controls. The only discrepancy between the distributions for the two groups was a disproportionably large number of workers from a cellulose production area among the matched controls (17.8 percent versus 6.6 percent of the totals). Ethylene Oxide/Currier et al 00 O95097 CONFtDFNTTAi However, the chemicals utilized in the cellulose production area were not considered to be of significance; thus, this slight discrepancy was judged to be unimportant in terms of affecting the results of the in vestigation. The only positive findings of this inquiry were the statistically significant excess in the number of potentially exposed with urine pro tein values above the reference range, and the borderline of significance excess of potentially exposed with high urine WBC levels. As noted above, it was not possible to assess the physiological significance of these excesses. However, neither an increase in urine protein or urine WBC is considered to be indicative of specific clinical conditions. Excess of urine protein and urine WBC might be considered of medical importance when found in conjunction with other abnormal laboratory or clinical find ings; however, within the context of this study, they were judged to be of little importance. The finding of clinical proteinuria on routine examina tion could be the result of many conditions including medically inconsequen tial sperm in the urine, acute febrile disorders, central nervous system lesions and genito-urinary infections. In this study, two of the potentially exposed individuals had sperm in their urine while one other had acute urethritis. The presence of excessive numbers of white blood cells in the urine is commonly associated with inflammatory processes and infections, and is also not considered alone to be a discriminator of specific disease. Ethylene Oxide/Currier et al 00 095098 CONFTDFNTIAL 15. This study, then, has found no major differences in the prevalence of various conditions between individuals potentially exposed to EO and unexposed individuals. This finding is consistent with earlier work reported by Joyner^ who found the frequency of abnormal laboratory find ings to be the same among 37 EO operators and matched controls. Within the limitations of this cross-sectional study design, there were no in dications of significant problem areas and no specific future studies were suggested. Ethylene Oxide/Currier et al 00 095099 CONFTDFNTTAl 16. References 1. Anon (1979a) Chemical products synopsis ethylene oxide. Report issued by Mannsville Chemical Products, Box 232, Cortland, N.Y. 13045, October, 1979. 2. Proctor NH, Hughes JP: Chemical Hazards of the Workplace. Phila: J.P. Lippincott Co., 1978, pp 262-263. 3. Hine CH, Rowe VK: Toxicology: Industrial Hygiene and Toxicology. New York: John Wiley and Sons, 1963, Epoxy compounds: Chapter 37 in Patty F.A., Vol. Il-2nd edition. 4. Joyner RE: Chronic toxicity of ethylene oxide: Arch. Environ. Health. 8:700-710, 1964. 5. Garry VF, Hozier J, Jacobs D, Wade, RL, Gray, DG: Ethylene oxide: evidence of human chromosomal effects. Environmental Mutagenesis 1:375-382, 1979. 6. Gross JA, Haas ML, Swift TR: Ethylene oxide neurotoxicity: Report of four cases and review of the literature. Neurology 29:978-983, 1979. 7. Snellings WM, Weil CS, Maronpot RR: Final report ethylene oxide two-year inhalation study on rats. Pittsburgh: Bushy Run Research Center Report 44-20 1981. 8. Ehrenberg L, Hiesche KD, Osterman-Golkar S, Wennberg I: Evaluation of genetic risks of alkylating agents: Tissue doses in the mouse from air contaminated with ethylene oxide. Mutation Res. 24:83-103, 1974. o CO 9. Watson WAF: An essential difference between the genetical effects of mono - and bi - functional agents. Mutation Res. 3:455-457, 1966. 10. Ehrenberg L, Gustafsson A: On the mutagenic action of ethylene, oxide and diepoxybutane in barley. Hereditas 43:595-602, 1957. 11. Embree JW, Hine, CH: Mutagenicity of ethylene oxide toxicology and applied pharmacology 33:172-173, 1975. 12. Hogstedt C, Malmqvist M, Wadman B: Leukemia in workers exposed to ethylene oxide. JOM 241(11'): 1132-1133, 1975. 13. Hogstedt C, Rohlen 0, Bemdtsson BS, Axelson 0, Ehrenberg L: A cohort study of mortality and cancer incidence in ethylene oxide production workers. Brit. J. Indust. Med. 36:276-280, 1979b. 14. Morgan RW, Claxton KW, Divine BJ, Kaplan SD, Harris VB: Mortality among Ethylene Oxide workers. JOM 23 (11): 767-770, 1981. 15. Remington RD, Schork MA: Statistics with Applications to the Biolog- ' ical and Health Sciences. New Jersey: Prentice-Hall, Inc., 1970, pp 1-418. 16. Fleiss JL, Statistical Methods for Rates and Proportions, New York: John Wiley & Sons, 1973, pp 1-172. 17. Netter FH, The CIBA Collection of Medical Illustrations: Kidneys, Ureters, and Urinary Bladder, Vol. 6. New Jersey: CIBA Pharmaceutical Company, Division of CIBA-GEIGY Corporation, 1973, pp 1-271. 18. Conn HG, Conn RB Jr., editors: Current Diagnosis 3. Philadelphia: W.B. Saunders Co., 1971. D0 09510^ CONFTDFNTTAL Figure 1 Identification of Potentially Exposed and Unexposed Populations DO 095107 CONFTDFNTT Al Table 1 Laboratory Test Outcome Variables, Reference Ranges and Suspected Direction of Value Change If An Effect is Present Group Hematology* Test Reference Range Direction of Suspected Change if Effect Present Hgb Hct EBC WBC Percent Lymphocytes 14-18 Gm% 40-54Z , 4.6-6.2 million/mni . 4.5-11.0 thousand /mm-3 25-33Z low low low low or high low or high Kidney** BUN Serum Creatinine Urine Protein (1+ or greater) RBC WBC 8-26 mg/dl 0.9 - 1.4 mg/dl negative 2 0-3 high high high high high Liver** Total Bilirubin <1.6 mg/dl Alkaline Phosphatase 35-148 units LDH 63-155 IU/L SGPT 12-53 units SGOT 15-55 units GGTP <38 U/L high high high high high high reference range based on Conn reference range according to Bio-Science Laboratory, Van Nuys, California Do 0951 03 CONFTDFNTTAl. Table 2 Non-Laboratory Test Outcome Variables Available Through The Health Inventory Questionnaire Questionnaire Item Has your family had any of the following: Birth Abnormality Miscarriages Stillbirths Death under 6 months Type of Response Yes or No for self; actual number among children Actual Number Actual Number Actual Number Do you have or have you had any of the. following: Liver Trouble Kidney Trouble Anemia or Blood Problems History of Blood dyscraslas Abnormal function of Blood Elements Liver or renal dysfunction Yes or No Yes or No Yes or No Yes or No Yes i Yes or No DO 095104 OONFTDFNTTAl Table 3 Frequency Distribution by Race, Sex and Age for Employees Classified as Potentially Exposed to Ethylene Oxide and Unexposed Race White White Non white Non white Potentially Exposed Sex__________ Age__________ N______% of Total Males <20 20-29 30-39 40-49 50-59 60+ All ages 1 38 28 9 0 '0 76 1.0 39.6 29.2 9.4 0.0 0.0 79.2 Females <20 20-29 30-39 40-49 50-59 60+ All ages 0 3 1 0 0 0 4 0.0 3.1 1.0 0.0 0.0 0.0 4.2 Males <20 20-29 30-39 40-49 50-59 60+ All ages 0 8 6 0 0 0 14 0.0 8.3 6.3 0.0 0.0 0.0 14.6 Females <20 20-29 30-39 40-49 50-59 60+ All ages 0 1 1 0 0 0 2 0.0 1.0 1.0 0.0 0.0 0.0 2.0 Overall Total 96 100.0 Unexposed N______% of Total 13 1.0 333 24.8 312 23.2 294 21.9 22 1.6 8 1.0 982 73.1 2 0.0 67 5.0 24 1.8 10 1.0 1 0.0 0 0.0 104 7.7 2 0.0 137 10.2 81 6.0 13 1.0 1 0.0 0 0.0 234 17.4 15 1.1 8 1.0 0 0.0 0 0.0 0 0.0 0 0.0 23 1.7 1343 100.0 00 COMpr 9Sl 05 TT$i Table 4 Results of Matching Individuals Potentially Exposed to EO and Individuals Unexposed, Absolute Frequencies in Age and Race Groups Age <20 20-24 25-29 30-34 35-39 40 + All Ages-- Potentially Exposed___________ Race White Non-white 10 18 2 19 4 12 4 14 2 * 80 72 12 Unexposed Race White Non-white 00 19 2 19 4 12 4 14 2 80 72 12 Median Age (in years) Mean Age (in years) 29.0 30.0 29.0 30.2 00 095106 CONFTDFNTTAL Table 5 Results of Matching Individuals Potentially Exposed to EO and Individuals Unexposed, Selected Matched and Un-raatched variables Extracted from Health Inventory Questionnaire Variable Potentially Exposed Mean length of service Total non-smokers Total former smokers Total current smokers Family histories of leukemia Family histories of cancer Family histories of blood disorders History of diabetes History of asthma History of arthritis 7.52 30 17 37 .5 33 1 0 3 4 Unexposed 7.48 30 17 37 3 28 2 3 10 5 00 095107 CONFIDENTIAL Table 6 Results of Hematology Analyses, Individuals Potentially Exposed to Ethylene Oxide versus Individuals Unexposed Outcome Variable Hgb Hct KBC WBC* % lymphocytes* Wilcoxon's Signed Rank T** p 1.02 >.500 1.18 >.500 -0.56 .288 -0.69 .490 1.06 .390 Exact Probability DEA/TD*** p 3/7 >.500 2/3 .500 5/9 .500 -__ Maxwell-Stuart Extension Test X2 P -- -- -- 1.14 >.500 2.06 .357 * two-tailed test ** large sample approximation to normal distribution *** ratio of the number of discordant pairs where potentially exposed was classified as abnormal to the total number of discordant pairs. 00 095108 C0NFrDFNTTAl Table 7 Results of Kidney Function Analyses, Individuals Potentially Exposed to Ethylene Oxide versus Individuals Unexposed Outcome Variable BUN Serum Creatinine Urine Protein RBC WBC Test Statistic Wilcoxon's Signed Rank Tf___________________________________________ Exact Probability p* * -2.02 -0.40 _ >.500 >.500 -- 0/0 0/0 7/8 4/8 8/10 -- .035 >.500 .055 * large sample approximation ** ratio of the number of discordant pairs where potentially exposed was classified as abnormal to the total number of discordant pairs no 095109 CONFIDFNT TAt. Table 8 ' Results of Blood Chemistry Analyses, Individuals Potentially Exposed to Ethylene Oxide versus Individuals Unexposed Outcome Variable Total bibirubin alkaline phosphase LDH SGPT SGOT GGTP Test Statistic Wilcoxon's Signed Rank x* _ ......... ... p Exact Probability **DEA/TD** p -0.32 0.90 -0.45 0.52 -0.82 0.42 >.500 .184 >.500 .302 >.500 .337 2/2 0/1 0/1 8/14 3/9 5/11 .250 .395 >.500 >.500 * large sample approximation to normal distribution ** ratio of the number of discordant pairs where potentially exposed was classified as abnormal to the total number of discordant pairs. 00 095110 CONFIDENTIAL i--JSsteaL_L so'__ U.j*O0/- ~Z2T *tA/zy&e. /A/ S&fA s*as2> SjC.f- --</?#> -AS__ /*/ *// \Ypeye. JZMHS. a?*mzus. ~J--Z%s3..ys3y`.. .^vs//,.^#. SSV&7&__ a*t&03T&A?S ^JA/S> -^V//y72' Atfi<U& <0AL<OA40tZ--j&/&U?L 1 JJ^ZZ/A&SL___ 2___^5SJ5S___ *4y6^is^ZM&S__^&*5r__ 70- \ /*ttvZs&arrj 7Zs<tf7~ &>&sx>SA/g <4_______ i___^/7x^--/sVi/<&s?z^y___ 2_^S^a^_vS5fi*2JJ5Z34&L. i____ /^y&fcL.-_____________________ ... & J&Ur^^tye. vSyedr^+fj//^ `&52j*________ z&T'^Z-Z4Pt&5L0-+USSZZ. SVfSS/*^ - ,, 755^ vf5S^ <HSH^ M#ry-_zz'' &V&C/9 y&SL /4Cl> ^V&__ ^QT,...*tg<g4 DO 09511] CONFTDFMTIAI w ^ JZM SWs<?XZU/sX/<? _. 'f22*9E2**5`__ stf &wz*t.a/. _^y__ yZ>se_^v&/. 'i W JSr AS02S3&04&' A&7~ ^AK X/XtHT ^2fWs<*U*S j&jiuus^Z-f-Aejk fe6 jzl,~a. t^SKSAstSLSU^S&Z^^ ^?e<f<ZU4Q&S s<WXLs2247l2*'-__ *J j*2e y^r j&6*su>yee* ><**& sito f^yei/so^*? _ #e0S*/se_s ^ 4jamM9^>; Ssktje&T^jpX ) ___ 7lr rZ*G7&r**4ye' ^._. >* ^v*ZtfZ&yL- - I ^Sc' s?trs> ' .; *. - ., . 44VZ>/2>*?jr^SS. jSZJLat AZZs JSVU1&iAsZ>j!Z>jfrsr a& ____ dfy&-ZV.,,<&&rt&QLc^2a*3z*i4ssa s/77#';___ZMS ^U'/^ J^/?'frag'^42A'jS_ _>& -.,,<*%&'*&- ^lo^ass S3 * ^ fajTTf 7Z&^aSSaar^ 00 095112 conpxdfntiai r s&s . &0V. \'A>ahz x?>& ^S&A/7r~ A7e&*t, _ ------- <5725*? <24--*SACA .77/'t&/ Z2>/r7> s&e. ____V StoOTC*/ __ a*S &a<2- ,A* ___ 0^j!&j&&^^a&s7&es< -ZMC-.. ___j&s+'Z&'A t54vZ>sXdf?r',, ^^j04S*s&e2>&& ^4//2>~,. rX'J/&SL--------------------------------------------f)-^7^^s&0<5r^As2>^&<%! X, --^s^zoaa, xs^r s* 7z> ,4er ac rate? _____ ________________ #A<22a<t^>Z*_^5-~J^ZZ/SAS^ZZ^_ ^Al^y^kLei TXT L*,,J&Zf . ___ &2>*2biA>L^SSL---------------- ____ ., s*2C5$Z>-7Z> .7MZ'--- AS.Z&5. _____ SJ-7i&Z^ -Z& 4&2>}lr2^^2Cr_ *, -VK^- ____ ..7&SL^^&eaAt-- A20M&yW* &- ____ j4-7&5Z--j- 7014## &-&?- /jgL ?___ 7/SZ Z'jy=bs2i_ 75W ;ggr~ /& rtr&iAr&ZA/A, fV? J&fijVT&Z,- &9*S&sJ&7&s3^A7Q4&&^ >8 &L _s%&K7T^ Atia&r&SjZj DO 095113 CONFTDFNTTAL '2 A'P'r#' -' ^ ^ >gat/^ //t^ /^-2><faP sf/JQ^/Sg ?&*&>____ - jSAM&r /sf e T&yir <5psszf,0j^ ________ JZ4'/44&e__ ^y ^g1 72&7^- J0 Ae/JZOA? z*> jsstzer ^sPZOTUS s4a/&&A4MS*' j&rfOA&A/-. . &OB . ..&, __ ^234/,,^23^^-^5^S0USl S<tfC.__ S^7ZW--J?ASL <3<zv:.____ _, les/zx&^o ^s. ^e/tss^--^L--ym' s%&*u&s*^-^5^jz6St -frMsvc?- /ter b?S4&^i^/yjty ,.. DO 095114 CONFIDENTIAL Zt&y&^ZKUi&l&^&y-*&&zd____ ^kZT&ASZrdjC. &WZ&ZO. . s*497ZU6-J&>e ZZfe' <&&&&>___ _ , sM/> ZMf t&Z&ZPjS&r ZZ^sS *&&344&> _SZie ^X^O^--yQ^ _______ davz&Px. _ s&s&7Z*</& ^ &e__jzM? jy&s<s> /A/Z>-y ZMZ'i/- >&-. z*s ^ /<gc, GjLJ&MtSiL___ <4 tj&taZjtSfZ&L 'ZL^m^COUSSL </4& . s&s&k 4*s2>-^77&f ZsST*/2f 4&& ,,/ksw&<2&&>,LJZkS.___ ZD 234l ZZ4%U*___ ... "___ &szzz/z& ZVJ&Z: -j&mu. TPS?* 43 Z&&& 2ZL 'y..j<A4t//t__ '.-A&zes- a>47z:_A/sm'A/. ^g&ggSfr &9^4JiL. Ke^L^SS-jSiA ^2G#r .At-JSSL ?Z*~J&4Z&L+______ _____ AjSA? S&tS}&*' ZZ*4a/ VU&-- '*M'&S&`PZ' SZ><?DrS _ _________ TKP /0S4&e&%>'4 aavz/zzize___________ Oo 0Aif- 9S 1 3 5 M/ ,j s s s y^ \ _ _ _ _ _ ^e^<aaiH4-^/" y^ s4*<z>,-ak _ _ _ _ _ _ _ _ _ _ _ _ 3*^ s&xy' &ZZ/0y - sfln >*2CJ0 StS&r /& &x5K*isA/%>s . . . . . . . . . . . &&?&&-'--^t _ _ _ _ _ _ _ _ _ &*?/&u? *r<r<r >-S/V/ZZ4&fr2>- s*S ^*4 . 7A/*^ ^<a/ ^gesg_ _ _ _ _ ll, f ! V t '" s. . . . ' " ' .............. : /Z&.y&!ZC> f ^/ZS^UA&CJS') .... S*s__ ________________ &sSL~s&4, */ JS*/jcaa0-fa$ *t&epo#&_ _______________ *&sa/^- J__... %/. 4/kz&2&&&&se&^ ay:. a 'ipzasezi -/&f&/y>~/-XA.____ _____ ___ _.jrz>________________ tz>z#^ ^s csA'jffjuefyJ _________ ___ /4$& ; yx &gA&^y^&&tt0k/jz&A f*orgM jff&ty'aMg zaa&siace_______ /_ VjCdZ- . AOX/YfysezMAs, j**f<2s&v&y j?/67~. /syf aj&v&t- _._/ szp&a+'rj <0&e.*4*' &<&:>C&ee0^&*AzCL^tg&SSScs&Ae?A ______ ______ __:_ S&wr/sr : If V&e, <ut6e*'z~ jtt&eyy_________ r_r___^. yYS ^sxsMses 7P . __ exS77VA& 0AS-.TV4YA' MST-tG^ Oq CoMpr Tn^l 18 rop*rrft, .J J .J ^ J O U vj ' J O' --J&&&0EL1_________ ______________ ^__ &s*'/*2&&'&*zas ./? ^saaaaBfiHg 4&&&j0*SX2SL ____ s4 '*) j#?sf>*<a><:______ jts&&z> ss%&xtic---^o2ir- s*qjsa* -^ewv' stfJsae4AS * " ** fs*24<*y- . - ^gStg---Af-vUjj ^C^ ____________ --SLt*j4j> -- ^r - g* s> *uuv& C`4A0toWei9J - / - f0tf&irz>&F2> S&e2&&9/BZ.y: te&A/sJ5Q4.__ sGAffsepraAt&s _____ V ^au _. ^igy graggggegga^ DO 0951]9 CONFIO^NTTA! ^ WT *>*t&UAS` &*& S**&& 4<S^5W^ ^ >*** S>At&a& StlS&MS&T&A/ X<e2> J5V <^4p^r ^?*S>*/. &S&e/^17*&ASS sAfAsOATZ s//aAAS77@c/ ^r>es // ^ &&ct& 4&aoe& c^r ^juk/ ^ /v /v^stf^f, 5^ s&e*7- aZf^ast- &c*a*z<kss&i'a&& &/S S7ztS& sx&ees> Ataa&ebs* 70 ASS, A&SS, ASJc. S/aA# SSA AS XSddAS/l/Sd&' A/AbdJ? S&A 77/S &7&&&S zwS a&z&a?/-' aASAamas/ a*sjxaat>&& 77/S <S0c aa/7> 7QOSS AASA/rZ&ee*is S&A Ssosaa ASS d347/&aAjr aasasz>, sy&eZMsa 0*77/ &*s ass &**&&<*& <sS auass asas tasks a? <ysAe&, 7 sahtsa/ 77*s Avt&soa*/ /ear** xK^r c> 70 SA&> dSA M4*rz/ ^Sf <&4>e TS&AA&S. jgs sr/A/d jb9 S/c * '&& ss/oass* /e&c&e7~ aa/saoatss &<& a&&70 ao^mat as& as? 7## stt/sua/us ax*s>s st&s& S7&>*a> ,,as 7&ews as &saas 7m m?es> Aes&er- sa/Asasz&s rm sass^^o^ aasZt r*e o^/^/scattca Az&Gt&us&if as rm amcs ssapas 7&s***s#z> *&e 77/g /2&<iSS*r AAA&S s>a^y &*&&>. 72# /yS7^7> STm/Gd/S T7AS <SSSIAS ^4Ki*9z6C5#b AsS SAAT/SAA/asS J&ST&S' s&vsv /tvs '2&&zk&ea***/d: saza^ , sa/2> TP'S A'd/ams* fsses0/SA/e&) 07 77/7 &&CX0S&&5` (STTddTJfJ /&**&> 2*#&*ei 77*7^- SX&AXUocAbe. ^c/SXfy7. J7/4dt/>&b AVs/AS 77/7 S7esSd/S//d/ dS MaS sS 77/S SddAid/AS7td4S 7&&&A<SAA*ft 7&& ASA^TT/ ms&er SsATSaJ/ SA/&/d*?/&S X*< AT#**/ SSA/S&/ aS SSA&7SS /&AS XicZ. 7m sT&MfrS&GS */Ss?SS. 77AS 07Uty &ues> AA> 7&&&S- a/ 70/s s/stt/ swz> s/asau, agsa&j&t- AA'dAssa&s s OMAea&AysK&ff ,a*ui*z> *s /**>- *7^zv* iS/WXSJtS &/e 77/S SW& &&U/3. SA#*4i&t> xS* A &e&9AfS#u>A/ <rr ypjf SAtaSWP A/SS7&&S& A7d> 727#*- A///A0SS& A<W~ &*/*&&$j 7Z>Wd~> /7/M7SSS /ZSs&S S0?%t&eSj AW> 7277k, ///MSS# S/H&SSW S>A0/ese& {Ss&b as/ /9S0 /sastt/ ///v&s/t&asy *4*n). ^ ^esAs^vu/A/ as ?2/ A/Saz/ad, S&S7&A& /S st*s& A/a/t/sSh si/a// as A///s a/Aoa4^ Z>/Z/S/A>eSj S&&0SS&7& AdS&vao &e4Y/t&e&j A*& ASAl/*S /QjtSGASJLt #&jwt&es * * AASAAW&tos s/r s&mt*/ x5* s*/aU70/A/0 /0//SAssjS#5e s*?s%&&&& s// szrsiasdb Ss&nu/o zt/Or- MaZ> S7&79/4& /&S70A/SS as J8aC7& /TtHAAd-5/ A/sSA^PsSS as ss*//fSszsAj s&sp/sjs M&7&/esss as A^aas> &*s0s&&e>s, aa/Z> A&md*& ss&Tpsyss' of S4*A3&e% Adsa, aurs'sss s&&$//As to ai&eteas sm&AOAtes &*&> a/^A*sc> ssssrSb sssidd&svs ^7t/- AvXSsAS SxAOSdAeSS 70 A/S**A7~A*AS, a*?**7~, dZd&s, sa*ss&77&/2>s, A/y/WS7USS /&rS`/AA'S//&eS`* snSTAz, Ad&frv&es, S&drfSA&/A/a ASSa/73 AAP Sa&Af/dAAS z/$s2> s/s Aa&r* Sd/rsA&/A'S AS/W/sTTSS. A*/2> S/a/A^c 7077US /USAS S//Add/2A& 2a/ %/&Sd?2> SA/SkSj/SSS /// ZMS S?*dAS/ 7frS7 tf/ss&sis%> S0s/77<//z# 70 A/Ass^e/ss as &/std2ersS/ Aszv/n/J, aa/2) Arsj?/7ys, 00 0951?0 CONFTDFNTTAI '* ?5->4 * *?V ` ' i 9- ' 0 **-&* 0' '-. ,.. ft,V ': 0 '* tF ' ?.' *'i.c.*' * feiteW'*4m ;>-:? -.1 ,12 %.; >.>- r6-V.c *;.; ? : ** *i)*1* 45^1 ` *;V'.:*.^S^-(rtJ-&?jVs,,,&.J,' V&* W- "*fr>J*,.i*.-."-V3<116*'*, i ' ''*,L. ,*,*,^`''_"_" __1i, >':4-~~'- .-1,i*' ->^.?.*\--.\'1`_2>O2f';*,f*s**."'>.>\'<-.v*x :^fio>-fl^l-..r*'#/:/1trisI-.fv.'-'*^.i*ViW^a^':*V-S.:*//V:./>1-s'?%Fte.,'v1v"J,jVf'oo{tr^-^$.v-v<>'>v'u'%;4J--O'4i/j'rV'i*^*.&tW''^t/5JWir)'-:V.1.,-;..-*' 1 '''--..' TY,'--.Vi,'<'/,.r:*. \-'4r* .<;'/-v: 0r- *.-> Q < ivg .//, g*^ ,? -, > Q^~* .^s^rsws g^AgggauTORT''l**'f3^^TT^gJ&Jaifc'.Tg T*1 T "` .r*-'--1 ~v r J!*"1'^'1 ^ .--TV, ,`Tj'-_ -'I. u. 'j sr^'flBsr J>i,.*tgi,';.v' .vgu.i '*. r ` jk/f a. r` ^ '--,*. -"n:^ -r ijrai'y.-r jfi n^-i^iir Jw N *-'A`: C^V o.wta-o'VO'^ .TV1. ` ' : '. ,?3Vv ` L r~~ . .`v :^!.> ^ scv . !a;'--*J. s>>V-S v,-. 09B121 -ONFID^TIOL Vf `'ftV I:- - ' *;V ' ' ' '-'OF' fVV' :-*S; ' . '* .' -- ` * ''' ' * v <'.\' %'-'' ..' -';; '' '* ''*'. 1'\'. .--fi !?' ; >jD 7Tf ,* if V . ,l*,1^`. , j^ v' fi * 'r'>'? -?-?:? .,/> ' :, . t *v m |p?v;S'^V *>' '' ' .' i.Kr*&>*'i*Str-.V- '' : P |-f,f */' `r^'. ' * V- f?;' , '. .' ', * " , ,1` . < '' ' ' ^ vr ' ' <*-$' ' 4 ' - 'iTM, ' *, ' v ,* ',^,+k.VvO' ;::s5 :V^v\' ** .. ' --^Vr-&* , .4 _ * v. 1. TvTr!T~::jy. ,'-r'-,sV5-^b4'-l&'*.'M. s,s.'*''.$i'-L'-<*`&??>fo*f`-`.t`",' vN1-' >? : 'A- **vi. t -' r-: .. .*'. 1rLV..T 'i -t'TXf' 'v ' '' V .=. 4, i , * . ` .; V :;4A-.r:- ; ........ __________K*. : '<Sr" ^? A-:yr ;:,.,.:. " f'?V %-'-.j. . .; :V ir <*; -' rP i" ujjjIh .".V^ iP 'r|iL... ^ i-'>i. > ...... Vi *.-\ :\ _, . ,jSJSiJfrL*--. 1^_ EL1 w'n '. M-jx. - i-itri* ' "> '<vr^v'X :/- V. -' v `1':> *' v.v-' ;...... 2 ^ ' !-\ . T 'J" *' - ' ' : *% ". ;-Ts.'V .. " ^ ;, **' -J -,r ^w'. ..o6 - , ,VW` .\. w.`jv;.. -','*'-'.'. ' ' ''vV 'i .' ' -% ,'i." : ` 1 --/ ~-r ?7^ . v* "V-*?* //..;,* *,&: /> :'4 &* >v-- &4>. / - * v**ffg)v-. ,.f*VL Jv* -J;,V,,y.- * , .V_ m^ .^ . * *,; \ *,-*',\;,j.mi . m ' - ^r.*., " ., i .; t*1 . - 'i*i. , '' ; *; . ,'"*,**, ,...... 1 t * *- '*.*<.** '*' '> l- V JJ'^! `,.^/iv ,-*;+* A ' " . 'Vf.' : v V '.^r, ^ `! 1 *j,'. ' ',' j n--ii.iif T^ . r. I . r. j-- I it Ifm--*, ! '.pi a I " <!'* III % 1 * * ni IT 1 f,| i fWl jgf.i.n ' rn 'ii-iy W^ .--f tij,ii I lf M.I ..IM|W iL l|i|r "' '.* ~ J-* m im n.m j -- n ' iJ p 1|" j <*-/,V;-' J ^ W-'- * 3>a ^'"; /?;' >:;,'i;y';1 -'iy-, ^T^rfeNdp^O^RS - * ..> . ,,,.. . ,,.vt, ., . ; ,v^ TOT ftli fcilft R^T * SHOK E R S - 37 T '' -** * ' ~ ;V .*, -., .^. , ; ,,* L.Unr'^K 15UN Ut l Uiti 1 1 NUUUb ULKt-NULNI V AK 1 AbLLb 11 ( 1 VARIABLE - wbc <rr:o-3> KADL 1 j 1 2 PAIR l 4 PAIR 2 * 6 PAIR 3 PAIR <* , PAIR 5 1 0 PAIR 6 PAIR 7 , PAIR 8 i PAIR 9 PAIR 10 PAIR 11 PAIR 12 PAIR 13 PAIR 14 PAIR 15 PAIR 16 PAIR 17 PAIR ifl PAIR 19 1 PAIR 20 PAIR 21 PAIR 22 PAIR 23 PAIR 24 PAIR 25 PAIR 26 PAIR 27 PAIR 28 0 PAIR 29 30 PAIR 31 PAIR 32 PAIR 33 PAIR 34 PAIR 35 PAIR 36 PAIR 37 PAIR 38 ! PAIR 39 PAIR 40 , PAIR 41 PAIR 42 l PAIR 43 ) PAIR 44 1 1 PAIR 45 PAIR 46 t , PAIR 47 ' * 1 A* b , *'> > ) 1 1 J 4 s t PAIR 48 \> AI'TVi-IB* 1 4'sj9-.'V* i V < NORMAL - ABNORMAL) EXP CNT 4 EXP CNT 44 *4 44 44 44 -4- 4 +4 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 - 44 44 44 44 44 44 44 44 44 44 44 EXP CNT \ EXP CNT *4 . . -- *4 - - 4 -4 -- '4 -4 DO 095176 CONFIDENTIAL * 2 /-MAY-81 CQMPARIiUN Or CUNT i R'UOUii ULKLNUtM VAhlAbLLi * VARIABLE - URINE PROTEIN (RRINEG) r A O . i \ PAIR i 4 PAIR 2 S PAIR 3 6 PAIR 6 PAIR 5 ! PAIR 6 PAIR 7 PAIR 8 PAIR 9 4 PAIR 10 A PAIR 11 7 PAIR 12 V PAIR 13 0 PAIR 14 > PAIR 15 2 PAIR 15 4 PAIR 17 S PAIR 18 A PAIR 19 ,fc PAIR 20 A PAIR 21 0 PAIR 22 3 PAIR 23 3 PAIR 24 4 PAIR 25 A PAIR 26 7 PAIR 27 PAIR 28 PAIR 29 '1 pxnr 30 3 PAIR 31 PAIR 32 It PAIR 33 PAIR 34 1 PAIR 35 PAIR "56" 0 PAIR 37 j3 PAIR 38 3 PAIR 39 4 PAIR 40 t PAIR 41 ,7 PAIR L PAIR' 43 PAIR 44 f PAIR 45 >2 PAIR 46 ,4 <$ PpAAIIRR 47 48 >6 PAIR 49 .7 II -.m.X n s 7ft 72 n < NORMAL - ABNORMAL) EXP CNT 44 EXP CNT 4 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 - 44 44 44 44 44 44 44 44 44 44 EXP CNT EXP CNT --4 ... . * -4 --4 00 0951?8 confidential ^ /-iiar-t'.i iUrt FAHiiUW Ut LUftl llVUUU^ Ut-KLNULft 1 V Ar. i nOLL .'j .VARIABLE - URINE 0 v PAIR 50 PAIR 51 t PAIR 5? PAIR 53 PAIR 54 PAIR" 3S PAIR 56 PAIR 57 PX1(T~W PAIR 59 PAIR 60 pxnr"61 PAIR 62 PAIR 63 PATFT 64 PAIR 65 PAIR 66 ~ FXIR 67 PAIR 68 PAIR 69 PAIR 70 ) PAIR 71 PAIR 72 PAIR 73 sO OP * . U PAIR 74 PAIR 75 PAIR 76 O PAIR 77 PAIR 78 --------------PXIR T9----------- o PAIR 80 PAIR 81 PAI ft 82 o PAIR 83 PAIR 84 PROTEIN (RRtNEG) t NORMAL - ABNORMAL) "EXP cnt 4 rxp "Cnt 44 44 4 4- 4 44 44 44 44 *4 4 % 4 4 4 4 4 4 4 4 4 4 4 4 44 44 44 44 44 44 44 44 44 44 44 44 44 44 EXP CNT 4- ' f'ADL EKP^CXr -4 -- -- -*- + -4 -4 <L TOTAL 75 PERCENTAGE 89.2% TOTAL POPULATION SIZE - 8*. 1 1.12 1 1.12 7 8.'Xi Z7 " DO 095129 CONFTDFNTTAL J 27-MAY-81 COMPARISON OF CONTINUOUS DEPENDENT variables 1 VARIABLE - RBC (RRI0-2) PAGE 1 /V. 3 PAIR 1 < NORMAL - APNORMAL) v-, E XP CNT EXP CNT 1 4 EXP CNT 4 PAIR 2 1 PAIR 3 6 PAIR A 44 44 4- 4 a PAIR 5 PAIR & 10 PAIR 7 44 44 4 iifi PAIR 6 PAIR o 14 PAIR 10 44 44 44 1A PAIR u 17 11 'PAIR 12 PAIR 13 44 4- 4 44 10 PAIR 14 44 31 PAIR 15 33 PAIR 16 44 34 PAIR 17 38 34 PAIR IS PAIR 19 44 44 44 i*9f 30 PAIR 20 PAIR 21 PAIR 22 4* 44 44 33 PAIR 23 33 PAIR 24 34 PAIR 25 44 44 44 36 PAIR 26 57 PAIR 27 9 PAIR 2e 4 44 4- 40 PAIR 29 41 PAIR 30 43 PAIR 31 44 44 4-- SA4i PAIR 32 PAIR 33 44 44 PAIR 34 4 41 PAIR 35 49 r PAIR 36 4 4-- PAIR 37 44 ;; PAIR 38 * PAIR 39 a ss PAIR 40 54 PAIR 41 57 pxnr 42 i PAIR 43 44 44 44 44 44 Jf PAIR 44 li PAIR 45 4J AIR 46 44 44 44 44 PAIR 47 65 PAIR 48 46 PAIR 49 67 *8 1? 44 44 70 71 73 73 -----7---74 75 EXP CNT -4 --4 - 4 DO 09S>130 confidential \ ' ^ - M fl T - 1 LUhPAH i !>UN Uf LkiMlftUUUb L't HL IMUL iM I VAhlABLt^ i) % 'VARIABLE - SBC (rr:o-2> J) .s c 'I 3 v I * 10 13 13 > W 13 1* t? / 11 vf>t > 33 34 35 U tI,f 30 ) 33 33 --: 34 ) 33 T A* 1 ) 31 40 41 43 ; 44 is 44 ) 4f 4t 30 SI R ) 34 ss s* 3? S3 j 4* 40 1 43 ) 43 44 45 44 ) 47 . V- PAIR 50 PAIR 51 PAIR 5? PAIR 53 PAIR 54 PaHTTS PAIR 56 PAIR 57 ^AITT-50 PAIR 59 PAIR 60 "PTIim" PAIR 62 PAIR 63 PAIR 64 PAIR 65 PAIR 66 PAIR 67 , PAIR 68 PAIR 69 PAIR 70 PAIR 71 PAIP 72 pair-ts-' PAIR 74 PAIR 75 PAIR 76 PAIR 77 PAIR 78 PAIR 775 PAIR 80 PAIR 81 PAIR 82 PAIR 83 PAIR 84 ( + NORMAL - ABNORMAL) DTP CNT 4 4 EXP CNT 44 44 44 44 44 44 44 44 44 44 44 4 44 * -* 4 4 ^4 .4 4 4 44 44 44 44 44 44 44 44 44 44 44 44 44 44 total 76 PERCE NT AGE 90.8X TOTAL POPULATION SIZE - 84 0 0.0% - J EXP CNT r'ADl eTp cTTt `n O^V \-UAMW------ fcfc----------------- // /& ^7 3o ..A 7 3X </</ Y7 'TL, _ r? fo O c/ (. </ 7/t? 77 "- .^ ^735-e /<? 4^/ 4)^ <9>=* X? Z3' /4^fi^<r /369/ar/iv; &Cl*tS 7ZA A?S<2, 4v5> Ut&Z X*^ Xp?S x> az\ t/rf^jues s??az**&e 7&4a/ x' *,t/*??ges<s. i&zj/s-, 7&y <u>u*j>sYa7- && ^ee>a^$ssz> ^ &w77A'*/e>us 7v/e 2*t7?? gvr M4J> 7Z> 2& ssrTO &rit4>/e/A*2j_ Prfyp^ 2X47# <U>s??/messv* ?z/ Z&t# fitaeus^ 70 stt /e&z^se&A/at' (stSSTT&O *rsZZ' 7&& S6444J*? 2S , #3Js XtK? 0JCS7& Ate^TZKlOZ., /f^US^xU/Tf' ^ &*?,,*'&.*'* ay * <52*^ xn<2> 4350KRU^ *T !*&&' ftoe 7*v *asb*a<H. ^ ^y-srs xva *- 0* - **%%**'** BB%j&a2> S4S STB** S', St's77? ?f 7&7ZU. ytF *a*? jcssrtlo >r r <vi> ____ j z*&u& s^a.totogo *??w 'tt/=` ?7 &*su?7i&> 70 f&ry&s s y>-s?a**e 7&e yutc?/ zht* .^2?^ ^ ^ 2X&K4 J^x7XS^7S. /Avz> tg&3 , *Cs&73 S&/C *<'& */> 7&S DO 095133 CONFIDENTIAL 7*70 S*Z*Oy 70 ^ 07/7*000 *377/2^7^X4705' Oa*5o\ k t <s/z&>y*3&&7> M*2> 30&f 0&</*/dbf /&z> 72*0 00*0700*, A64%> y?&y s&y*fa***&2>, 77*0 /F' 7^772*7*^7/0 o>x&7)A 70 *2/0 ^ssr ^ deo/tTO -~ b7on&it, sg&/&jr*>*7aa*' *"& /&ey&z>, *b*ye& 70 ^ &Z&JS A&T!/? 7*7/M> 7770 OmSTPHOjC. d7XY2>/2>M7MS' tus/eo /^r 7* <3*70/0)'S/Or/3 yTe&Os&e**? &/? A*/?7tZ*/7/0. So7a4&&f 372*/>$/ -<H*2> &a'Zs&' &77/2<7j/o&. o/yw o^yysj/oor w 7740 7z*i{t/ S=tyy*/727**/j t> sov/sot* ofi tmo ot/ttoo* &/*7*c&/5 *2**x*z>/2>Zd <&e&u/=> aJ4S /7//7?'075> 7Z7 2>0?&e/?*///0' /2/2/Z///MO &U&>/&lZg3. JF 7* *>//3&?o O9#07&47& **& ^CO**70*S7*/ &SZ0*Z7o> /7>yC M Os^TTGou^ **7*75//, MO */* *M/Z*/I>& *70Q? 7MO 77*/?72>7'/*tO *e03SS /Zxe M*o* oMOso&iOA/'r //*7r///M*of M*3/eo& O'x//0//t7*7*<p 7770 ?&ssAa/**y OF Tty** 7SMP S7Z/>*/ 0*7*7*>0t003 AOZ&/I//A*#. J*>**M 7M<0 7*7*0 ^SSv^6W dTM/^sZxdTO. 7740 /*%07` e/Z&&M Oo2> 777 77*0 T7*7Z3**7/4 9e0OS& **4S 7*00. AtS&o &>/ aor &&&es 7z> ys&s&s&e m*4t/ ?? &vms **&eo 0*7*07* s/y7& Z&O 0700^703 /o&ee ea<e7S/ts 0A3M - 07". O * <7/7 iMG&Rs- J C77 M. J&> Jt&S 7T4&7 sto/tes, o#r. 3 * zsjte*?* 7**eu ** &*SjJB*-g*isr tuo 77*0 <s&u>& &7*3*0yOO 7XUtW77270)4MQ, 04077 &>77Z&>7, *5uM2xxJ2 7070T&O f47^ 7*70 77000 ^ 7*70 M&A277/ /7/1707/7ZDG5 OTtO/t?J /v/OS O470**77*7O2> 7*MZ> OOTOO/iO/ZOb . -ZM &/0 /00O *24700*^3/03 T07?7Ti7*o!t>f 72*0 /0AXSM/M* ^FOOOS 25*s*7//**o2> M>/2 z&*o 00(700 Okyr&eajL. t A47*Z>/27470'^ JM A&Oj M* A/OM7' /5C7'2V*2X07O' /#70O 50*0*3TOh. 7t> 77*0 7*7*70/777/7. /Ze&ooos <so&e7z> o*o. &/0 A/OXO 7*77*70*7777/0 d/703<4 /tOS 74/30 &>/*70 . O/t/OO 0407/ 57*700 O0/OX/&OO M*/2> 7770 *3**70Z>/, &*40*Z>A7O 0/7*0 >7V*Z*rssoS> 7M52> *Z> 774/0 TO/eO &U*POO Tt'77*/777 O 507100 <D/? Oa/O 07/0)77*0,0. 7MO /OMO &07S& OO 4S4U7 00/30 O&*O70/> 79T/^ /f&,/Z/*0e/20$> <20&//0&07W*t? O/TO&s /t+'O /&/&00/07&37/0 M&30 707*0)0 /& 00700/V/A/O /7r 7/70 00/7007/* 7/743 7/40/. /O OFj 7/70 ^Ofadl35S'^/& /$3/s/7Z7//ASio> /O 7/77-f A TTOa* 4&V7&P0 04t/Z>/7u*70 U*?0 OOXOZ*72&>. 7WZ> 7*VO ' 7407077/A/0 Mseooo&o/eo *000 /eo-////Z7*7oh), 0/*00 MrV/> OO/C 7740// OS0O 7**/*77i7*0/> &0/ZOO07/ O/VO/O&OO /*/*> 7770 /kW/X&0 /3*0Z/Z*37O 0/7*0 *e/>303302>. 77*00 7OO70OUOS- /77>> 007/ OX<5UASO>> 77/ 77/0 /&OV/XA773*7/ &00///s7Z?a/ 7?e/&03St, SOM **4& #0 AZeod/Os** o&e t&atz!*a/t/o, a? *&ao **& OMM.vexs/eZb /&&** 7tto <2OX**&AOaO&/0 707*000 A7*2> a OF3G /t*/)Z>0. J&* /&*00 /Z4077*^> , 77/0 707*7*^/77< MZeo&OSS 0*3^77*/*/^j Jg/r- //*7^. 7VO /ZM77Z-//7S0 A44& yer~S*79^70/> /V/77* A7&*J> 507/7*^0 07/Z>/2>d**0. S/04777/0 7t7*z> AU/uwac, *JS3 ^***?&, */&eo a*omt /*/ Xxe /*777?J////0- /9Q*>OtOX/eOS. S>/00/t'/7/<OA M&/e ZZ/O 72*/y O70/M4S/&O 7*40 0*^3007/7' 30**Z&P/ 47^7^*)70 0/7*4 ?&>a03sg<5 A*000 X?f34d/0^ A*/Gf>70 77*0 *&7*/*/s70e/007> m*4'/047, <2>0>Z* 0/703, 71770 07*aM/70h) 7//M& 7377* - S/T/F/tf&Oj OiiO/TPO/O <S0*/>7fiO/<, At*Z* <2ee/3/30A/7` <7*//t'& . 3 #&JL, /MO 7UA774H0 7^07210/03 MOOO OM*75aa70^ 75^ *2)47/1 0*703 /**/& 470*k)*OZ0> 77<**> 7707*7'A 7*/*i0Ose07S, 04 7*7* t/& ....... 7*.________________ :' 00 09513A ------------- - CONFIDENTIAL .i /USCH/az. //SSSS K JS 077/ 77/S Sosa/eT/s* SaSOa^u //&& #?&&&& asa/tz!*/s> ss7zossa/ <^r S7u&t/ a*&sa.t&c/ss sa/Z> ttvs C/a&&*7~ &W9&JL. 04/*>/Zi<l72' / ZMS /&S77tA//A/a S&OOSSS /S *'>7', XVf S&>a2>// &// /?K4V e/SsAS^ S /Z&C> StA/ZS&A^ 0 ^ 7SS 0///7PBa^ /2S//Z>sZ>S7g /9&%3SSS2> ^7' /gUS 7SS S$ea&S2w*e Z*2ry&e*4, A/S suOS A&&/2> 7Z> TVS "\SS07~A7/S/0(a ~S/a*z&?/L, 4*'J>s2i4?&S`. SA/2S 77//S jC/37~ as *as/a/-S/A/A '* &//ZS>/Z*l7S s'/f^ Shv?/=>js7 SaS zss snisy &&eysnyss Sa//a /%&&&*> &> <? , 77/S A7ZZ7S& /Z/37~aS OAA/ZGe^ &4*'Z>/Z>*&&S` A*fe> ges*/ >$*/srs& s&e was sv/Sy j t ta/s Seaa&w /*&*? A? asoseA/ as zta/S Saa/ou* " s/aS Sa/Z> S7&Qt&/A/sb W Z>/?t sS 77/S Aff& A&/U77/ s*/i/SAs?z*& 70/#, gass2> aas S 07eAa7'~ZX&-/&/ S&/2#?sS~. 77Z4S >072 arts WSa/ d&a/S&yes% 7& 7MS &azs as 77/s /%&? /kmw /s/rs'Ays&Gy sxas? s&e t&s S7az>y saipaassS > sa/z> 77/s su>A//zeaa, /2<mz*2**z& aa/s 77/s s/r sxstoz //S/ieass 7Z> TZ/sOS" as 77/s srus>y sa/saas/ss a#& assesa7^> ss 7//s s#A7Z!a/. Sa/A/a 7Z& a/s/?a2S7~ mz sswo ZbO/s oss/ea/te//' a&eot> g/vt/ze# ~7& &/s??/ssfS/-f Sasr/xoA/ sSa/soa/^s O&U/A&frvCSS a/r ///SU/SSS/f, dOUZ>S, S72. y //// J70&S O/A 77/s SZSS/S7& as 0S7X /&#&/.7s aazv/A//fe> seas# 77/s a77/ay sa/snasss ys. 77/s /U>//7&>0 0?A/7&*Sm ///as /7 A/a>zU/ s&s Z*/s 7##*/ Szvs&pj/ss saa sssts g&zs/A/Jfr, 77/a szf/xr&s/ ajos /^aas7> /v 77/s ybSasaoTT&A/ S/'/S Size yzsfW&e sesxzss&o/s /a/ ws >t&usass />/Go5er szz^y^ o 7U/x/ s/osassss atszQ/a/s&y ^yZ> a A/Sio a&b&SM as 77/e dev/Z&s?/^ SA>Asz/7>a7 S/7S a/as a//aZ7a77^>, J5S a/o 07072// 2aa/Z> s&r az#/A/e>> S&a? ttas S7us& skPSxa&sEr //// /^eaassss7> J &sS A&&/77A=/2/27-?0A/ A&s? 77/47- 7Z/>& *77A7>/S A/ATS StesA/T&Zbs Abi/D /9eaSSP^//GSSff &S2.5&esgSZ> j&OA&a /O&GS /S//77/777SZ7 OS7&AAS A? OOASseaa, 4WZV. A=iacasa/s\/a /a o as&7- a>sza/*/aZ> asqosp? ws /z?A&as//7aa S7sS. //* AA/aasa/ss 7?zs 7/7aAAr7Ss<sa7^A/ ao2> s%stG'g/7&s as 77/s S7Z/A& AWZ> 2&VZX&*-' SZ/SaOJ/SSS A/SSfe /V 77/S Ssz/7y% afi/ce 77/e su>/77SajC 07//7&//A/& S/eoa^tmee /#/& sssaa ^otsast^zp, AS S/OS >7&&02/AaZ> Z7/A>7~ 7&S& S7UZ>/ &SgA&S(/&gs &/Z> A/OS AA4i& ^az/S^aa o/otsa/ss c&z*//Aa&>. 77//a a/ss Aa^e azo/^ #d fz\Sa&a/j - /VM/7& 07ZS7& fr&as) , />//> S&A/e 7a C7XZAUAZZ<3S> - AUA/ATS Sr^oy S^as/Sj . Zas 77/s o&z/a/s >/ss, A& 0/07ZU/ da&AiO ^Ss /&&//& sa/ a&uc/*4*s AAA 77/s jtotovKfar- ass a*/&t,A>y ,,**.??- yfSySZaSSSZ Af&**eo*e,,7- (W ^ ^ *e* W Z'*T TM" * *** 77/s y ifeae ^a//ssossa/S. 7Z> asaasA/ s s/?/?>2// MAM Ua 66n>,ZZ#w Z*Tj 'M>/t DO 0951 35 ** ^ S 07///&/ST, ' C0NFT0FNTTAI ................. ............ -. ---.......... ...... .----------- 20-APR-fll PG 1 SUMMARY OF DENOGRAPHIC/U0RK HISTORY DATA FOR STUDY/CONTPOL GROUPS WHITE STUDY/CONTROL GROUP --------------------------- STUDY------------------------------------ ------------------------------ CONTROL---------------------------------- EMPLOYEE S R AGE HIRE EXAM SM AL --------- ----------------------------------- DJLIE___ O-AJE--ST _ST EMPLOYEE S R AGE HIRE EXAM SM AL _ DAXE______________________DATE ST SX ------- aoiAjuuxi__ ixi-_*a xi .29.1x4 03x0.-621X3X6 -C_J____ L'LRUSSCLL- SR LCHUGHES JR M U 3A 680102 800319 F L RWWITTMANN M W 3A 680527 800515 F L RLHrMHINGER M U 72 79051 A 800306 N L RWHOPE M W 22 790102 800110 N L -- G WB-AXO-WXN--M--V _3 650-6Q2- RQ0 7JL X .M MJeROUILLETTE-J_B.-U.-3S fcfl flXl 5-&J.0.8 21_ C. M. JR ROSS M W .?8 780213 800305 F H MEKELLY M W 29 75\1'D0 6 800A15 F H OBOENHAM M W 20 790618 800328 N L TMBENNETT M W 20 790219 800627 N L M TSFALKE M U 20 780911 800314 c M DRDENOVA M W 21 771010 808124 C M CGMOUCH JR M u A3 590320 800728 F M pur.oo* M u 38 AA0A21 FH CPSIMONEAUX M w 42 630606 800808 F M nFftjnr KMiiu M u 39 ft30 5 ? ft ftflnqift F H JAWINTZ HI M u 35 661010 800701 N H MJTORRES JR M U 35 650429 800508 N H WT AUSTIN M U 29 7A0501 800326 F L rrfiSAVFS .ir M u 29 73n220 RO 0 70 ft N | PGBULLIARO M u 28 770822 800307 F L GAPFRR Y M u 29 7 A 0 ft 2 ft ftOOftl1 N L RLSPILLMAN JR M w 2A 770110 800708 N L UCDUCK M w 24 76052A 801023 N L BLHARRIS M w 2 A 790226 80071A C M mai A wn ry M u 20 7JR 0RP-8 800M i) N I JCCREEL M V 24 740909 800708 C M M.lp AR B IF R M U 2fl 7ft0ft05 ft 0 0709 N 1 TASOLDHON M u A3 670605 800327 F M HLHEBER T M w A3 680 4 0 8 800501 F M RLGOBERT M u 30 7A0805 8C0708 N L OTBONIN M U 30 690818 800929 N L rnwrRr ri? v, u 29 7A0A21 8 00 70 2 N | F rtmOnsdn t T t M w 29 740521 ft o o ft n r N I TJPELLOWSKI M u 33 72091A 800326 C M BJALEXIUS M w 31 700831 800428 C M JWMASEY M U 27 751103 800321 F M ART IN M u 25 7 A 081 2 _ft$0 70 9 r L hmachtermeyer M u 27 7A0826 800509 F M V GLAVlAwn jr M u 25 74191ft R0051? r i DL ARMAND M u 25 7A0909 800318 N L UJBOURGEOIS JR M w 25 760105 800502 N L SCSMITH M w 25 780626 800306 N M .JHrllTRFR M u 47 6A 0714 BOO.371 N I JKPHILLIPS M u 26 741202 800111 N M JJFRYDMY M u 47 ftfi 0407 ftOOft1l W 1 ALWHITTINGTON H U A 2 600801 80071A C M DJTERRITO M u A3 610501 800930 C M UUSCOTT M U 38 680520 800730 t L F.ipniiRftrn jc M u 36 A50517 ftfin*04 r I CUCHEMIN M U 38 641217 600812 C L rral MfiNri M U 3ft ftftn303 ftnnft03 r i CJVILLIBER M w 33 690818 800310 N L JRGOODWIN M U 33 690722 800206 N L FTHACKLER JR M u 36 67061 A 800703 C L TTTIINfi M w 34 73010? ft 0 D 71 o N I RCSIBLEY M u 36 660608 800915 C L CUHODD .JR M w 34 ftqn 7? ft 800505 N 1 RJPEPITONE M U 33 69012 A 800319 N L CASPEARS M w 33 730416 800624 N L RGCOOPER M u 33 690210 800701 N L V.ITTMPI FT M u , 25 76^ 11 n ftnfl3? 1 r I WAMILLER M w 33 670411 801024 N L rnnuvAi M u 26 750 ^ ? n An n ftir r i WEHORSFALL M U 27 7A0909 800716 c L KUMARTIN M u 27 740729 800910 C L JWGRISSOM JR M w 21 770705 800313 N H UDN OR DF R .IR H u 2 n 790514 fin 0 31 3 r H KAFRUGE M w 21 780116 800307 N H RGKFNNFDY M U 21 770919 ft 0 0 A 21 r M_ DAPATE M w 23 771205 800702 F L KJDENOVA M U 22 760517 800716 N M n.n fki iif M u 1 9 790529 ftp0 31 9 F | RECUTRER M w 23 781127 800618 F L . 6PCHUSTZ M u 22 790924 801015 N M , R.IARMAAin M V ?f) 7Qnft?5 Annqn4 F 1 9 WHALEY M w A 8 590212 800911 C M JpK1NCHEN M V AS 561203 800826 C M KPBROU H w A3 630705 800716 F L CLWIGGINS M w A3 661213 800415 F L .ifI nvF M u A 5 A.3nM a SO 0707 r I r.lT h T ponFAJX .IR M VI 45 qqm n? nm m 7 r l ... HE MURRAY M U 38 660228 800701 N L JLOEASON M w 38 640217 800725 N L ; .. DJBOECKMAN M' u 25 760202 800327. C M XPi FMOtWF ;M u ?fl 7R0424 ft 0 0783 J DFEERGERON H u ?5 780313 800403, C M X CFRARRFR M V 21 790521 800915 WATURNER M u 3 2 751215 800703 N L / GWRACHAL M w 32 740107 801029 , >*- ... 4 '......................................... 4 . * <- ^ >' ' ^ ' >.* > <=__________ * V' != - , * . " bo 095136 3 ^rtkirTnCNlTI Al- PG OF DEMOGRAPHIC/WORK HISTORY DATA FOR STUCY/CONTROL GROUPS ^&W^(stUb.y>CONTROL GROUP STUDY--------------------------------- -S R AGE HIRE EXAM SH AL ______ :____da IE-- n-A.tr sj_-st_ EKPLOYEE CONTROL------------------------S R AGE HIRE EXAM ______ ____ -DAT-E _ fiATF 7J-1Z05--SHD.Xfl-2--C--l --RJP.UJOLJL 750908 800718 N L CRWHIT AKER M W 28 740805 800509 790416 80070A C L MAARMSTRONG M 21 770822 800403 6 3-L210-89-070 9__C-.H_ ._RAWBB-JR-M_ 36_68711J3_808JIQ2 LCD AY 660324 800703 C L AJDENOVA M 39 660602 800904 DEPANEPINTD 6A0309 800702 F L HTKING M 36 680123 800402 _RD LENOIR 8812.02 8-00 701 C L. _____ JMWQO.D-_M_W_3-5_7.00 427 80 04 0 9 TJUATKINS 760614 800703 C L MJDIDIER M W 25 741118 800723 TDSMOTHERS 690818 800707 F L RJHOTARD M U 33 670110 800902 i .11, rjnojWF .m feftipin Ann*?* r i HRLAMES--JfLM__W__33_690 7 21 80 0612 BUABADIE M W 25 751020 800312 N H RCGUSTIN V 25 740722 800912 , HRVIATOR M W 28 721003 800305 C L PURAMAGOS W 28 741202 800219 JARQt OMOM H-W 31 730608 .8-0-0-40 3 H L _____J6EARK.UL W 31 740819 800407 .OTJUILLIAHSON M U 26 740107 800714 N L RDTHORNHILL U 26 751229 800815 ' ____ MLHIMEL M W 24 740729 800714 C L --W-24- 7MSJ_L_ 800326 LL WCGRANT U 24 740826 800723 - T.V78 MP. L EJ H U- 2 4__ 770926 fiOmil ` rkbonanno M W 23 770926 800320 C L PEGRACE H U 23 771219 800303 RLAGUILLARD M W 23 760426 800317 N H UJLEGER JR H U 23 780417 800416 IDGfiXSSBM- ALU--26- 76959-7- 800325 _E--L WLJACKSON JR M U 26 760802 800321 C H iU1BA_I.LE7--M W 26 -7606-21 80011B JLSTOUTE M W 26 790219 800508 > `JLCAZES M W 37 680409 800710 N L KRTRICE M U 37 640130 800723 ASGFftlRE I JS--M--3L 7-30 31-2- 60 0 716 B--M___________ .JLBRANC8L M_W--31 --74 0 3.04--8 0 0 4-16. JRALFORD M U 46 69021,0 800708 N M T BREKKEN SR M U 47 630603 800813 JLSHELT ON H W 35 680617 800312 F L STRICHARDSON M U 35 680130 800505 JESTEARR H-W 39 681 02-2 80 C-326--E-L _____- QADAI6L L_ M--W 38 680513 800605 \` * f* 'al , ; * \w "j 2finABR*i8i `- - A PG ^V^MsUMMARY OF DCMOGRAPHIC/UORK HISTORY DATA FOR STUDY/CONTROL GROUPS 3LACK. sfUJay^CONTROL GROUP pCrs, - * ,-X'v'Jk-v ^,juqY ^'-"HMp' -s' R- AGE HIRE EXAM SM AL date____ o^xl.st st EMPLOYEE -CONTROL-------- - S R AGE HI RE EXAM SM AL V`-*%&.^ ^-ssi-tAft* v, R 3P 74nqnq ftnrwnn c M :/iiKRtE'UEUNEy-M vB 23 770829 80 C 32 A F M U ROBINS0%-M B 38 721115 e00303 C H T P ATT.pbsnK .TfilrH; R OS 751013^800324 N L %dAVpUj\LARD H"B 31 73 0 612 80 0 701 C M 20 790016 800711 N L .-" -'.&7mif.r-HR 31 78 0 ftp ft 800318 C L_ cy'E 'PRDV^STyyR:;MiB 30 781106 80 0 328 'N L AN V^ b;.,2 7 7 5 q 10 fa SCO 311 r h :*%K> M vfr.&& -fi9-0-72?J,: -Bfl131fl_C_L^ :/.;. L;bBRAtfFSRDi V ^ 28 770A18 800707 C (M 27 740611 800707 CL "^.'TJT'JIHJl'i `W'XI `AlAJS1 L V T H OJ-J A S. _fLJ3 30 .7310.03 00604 C M MCRUFFIN M B 23 780109 000401 F M If JAR R 1NGTON M 8 38 690811 R00'806 C H HJJAYLO.R_M B 25 74040,1 800403 N L_ DRCLARK P 13 30 76011*9 800918 C M RUPARKER M B 20 790521 -800903 N L R AU1L I. TAMS M a _3.3_ 7Afl91lS_aflflB17_.Jll._L . -U HARRIS JR M B.31 J605.03 800430 N L HIMELLIEON M B,2jt- 750324 800617 C H E JF R E DLfilClLS Ji^B- 38 ^9JE71.4 ,.800222 C L ` - J. JT1LLMAN M B 29.1790521 800709 C M CATUCKER -M Q 2.7 740805 800826 C L ... PARTICIPATION DATA GROUP NUMBER I. Population Universe. A. All employees eligible for Periodic Health Examinations (PHE) in 1980 (Total in plant except certain management^' B.' Total for whom 1980 PHE data available. C. Female employees, .Vj* D. ^Population universe used as source for in population " II. Exposed population, selected from source population($Xi , A. No. males in G-lycol II at time of 1980 exam. V B. `No. males who worked in Glycol^H on or before 31 Dec. 1980. C. No. males in Dowanols at time of 1980 exam. D. No. males in Dowanols on or before 31 Dec. 1980. E. Subtotal in exposed population - 1IA + lie ir III Control population (aJU A. Potential control population, selected from source population (ID ) minus v ___ exposed population Jtic +3L \l < I ^ K 4 M ^**4 CkjU*rX**d f T*' >"A- *PARTICIPATION RATIO '5i'lMj (fSubtotal in control population *1^5^ ^aiyWwiil Ol --i ^Participation Ratio No available divided by no. PHE. ployees in the group for whom a lovees in the group eligible for 198 /' * bQ ,v CHARACTERISTICS OF EXPOSED AND MATCHED (1:1) CONTROL POPULATIONS CHARACTERISTIC* * I. MATCHED CHARACTERISTICS. Number Mean Age Median Age No. non whites *------V Yrs\in Plant , Current Smokers Non Smokers VALUE EXPOSED CONTROL r, > . v* r K r * ' -/> %? /# v* :1 '.v.-tmmm ~ >. h 4^ | '< / ' *-> I 5-.v\`.1,'' - I -s/V. .' ^ II w'. *,,>"'v<?-C+ a' : j > < ,H --. I f*'''- *' * ^ ' '__-- I \f 'V'"1 ,> " ' ;* l^r^.jiwl^*.J,l hm\jk * *~-r i*- nv. R,- x v,v,v%;:\w `-t-A '}. * i" - f ', J -.fi V-V-':/ __ J~ /VC :' - /i fi ],; --?- - `"j-^.- .'.v. .-.L--. .. - '..- '--^--Jj-.~v. twiv>t.'--jcA,.,.*w?*r*. '''iXat.v:TM^-V*'*^'':?nl':,ll-'' -r*'' - :V " .' ' * ' :,i .*-' w- ..:v '-' . ______ .. ;.:...rTf^fevr--- : -.v ' rw,/jre~s: ' >> IH^v: - *. - I yr f' 'v :r ^ (."' IIIv%i,<- ''~'"-'^_.~'.~.V:-ir<g`n.<y----":.'.fi'^.. 'j,g/,- T-J;y.r--: . .- ---, -f *,f DO 095141 iSS CONFTDFNTTAl SW; / "\ , 'T ' ^ ^ 1 -h, /' ,' ' * .* -f \ ~ ,* y ,V' -/ .*!-* ,'j 5*'S' ^5^%. f. ^ S;if*'$' ':} 'fi^J*'* "Sv '*. ;I*TW''{' ' '* ; V '<, : '" ' <W. rb (Ik* ------- :-- ,- '1,^ - 'x 't L^'GS'i Vn -aga IfK/^/j U* ?'-: s. */r -X ' ',;; .V.- y.'i , ; F J '- ' * %. t tFx ;'-;.^;:^'i^"b|;V-: " 7-; S&v>''- >'.> :^ ':X;y :;>llr $ :-r X' v ;? j**' ^'t.v * .-----.^..m ,.s.gtg,^'; tv'* f:- ,.y^^ :s . - - .1......... .. - ,, *, ; ^ ^ . ^V ->'-' ` ''-" .''*? - ' V ` - H- J-A' - * ,. '*.1 r ............ - -- 1- ^ ' *X - > ^~ '*-' - ^'' * \ f ' - ]r _ .>' '.jl\; v . :-, V,---A" '< - -S-41 V,:.- ' r. ' ' . v".vi J.,--1'/: ------w-r-*^r* 1----------'--- -- ' -'*-*-*---,r'" >- : ^ ^ - _ >l _______.t -~--------------- - ------- - v-^ygsT . ^ '- K1 *V- ^ *&'*!'* --s--*. '-'1 atii^ J. :i -m VWTJ -' :' 1 W-. <, I'v'. /' Xr f'f;-5';'| 'i''7'>-'>'t' _ '- " \*T>.- '] 00 09514,1 CONFTDENTTAI VAR 1AHLE UR INF P ROT FIN <f?R:NFO> PA t P f^O PAIR 51 PAI P 52 PAIR 53 PAIR 54 PAIR 55 PAIR 56 PAIR 57 PAIP 56 PAIR 59 PAIP 68 PAIR 61 PAIR 62 PAIR 63 PAIS 64 PAIR 65 PAIR 66 PAIR 67 PAIR 68 PAIR 69 PAIR 70 PAIP 71 PAIR 72 ~ ' PAIR 73 PAIR 74 PAIR 75 PATS- 76 PAIR 77 PAIR 7ft --------------PAIR 79 PAIR 80 PAIR 81 pxnr"E2~~ PAIR 83 PAIR 64 '~ (* NORMAL - fPNOPPALI EXP CNT FXP CNT 44 44 44 4 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 4* 44 4+ -- EXP C NT 4 ToTJTl----------------------------- T5--------------------- i PERCENTAGE 89.2X 1.1% TOTAL POPULATION SIZE - flA 1 1*1% FXP CNT - 7` 8.31! VARIABLE whc t: :n-3) *> t* NOR HAL - ARN0PHAL1 , CXP CNT FXP CM 9 PAIR 1 PAIR 2 PAIR 3 9 PAIR 4 PAIR 5 *4 44 44 - 4 PAIR 6 44 9' PAIR 7 PAIR P . PAIR "9 9 > PAIR ia PAIR li --_ -- PAIR' 12------------ 44 44 44 44 "4 ~ 4 9 PAIR 13 PAIR 14 44 44 ,, PAIR 15 44 9\" ---------- PAIR PAIR PAIR PAIR 16 17 IB 19 44 44' 44 :: PAIR 20 44 ~prmr*21 '* 4 PAIR 22 44 ~4M -- PAIR - PAIR 23 24--------- " 4 ' - -4_. 4 - " ' ~------- ' 9 _ PAIR 25 PAIR 26 pa nr 2 f 44 44 44 9 PAIR 2 8 44 PAIR 29 44 -- - PAIR' ~3tr------------------ ------- --- - ' " PAIR 31 44 PAIR 32 44 41 PAIR 33 44 9 PAIR 34 41 PAIR 35 -- -- PAIR" 36----------------- "4 4 4 ----- 9 k1 ,, PAIR 37 PAIR 38 PAIR^9 *4 44 PAIR AO 44 PAIR 41 4+ i r PAIR *2 PAIR 43 44 PAIR 44 ,, "PAIR 45 4 9 * PAIR PA IR -- "PAIR" 46 47 4B------------------------- 4 4 ~4 4 4 4 9 PAIR 49 *4 fXP CNT , 'j. I'sM-. DO 095145 C0NFTD5NTTAI VARIADtr - mf f - p : r - ^ > PAIR 5n PAIR 51 PAIP 52 PAIP 53 PAIR 64 PAIR 55 ^ AI P 56 , PAIR 57 PAIR 58 I FAn 59 PAIR 60 r pair 61 1 PAIR 62 ! PAIR 63 , PAIR 6* PAIR 65 1 PAIR 66 j------------- PAIR 67 PAIR 68 PAIR 69 , Pair ?r ') PAIR 71 PAIR 72 ,r--- PAIR 73 PAIR 74 PAIR 75 , pair 76 PAIR 77 ; PAIR 76 I------------ -p>TR 79 ' PAIR SO PAIR 81 , PAIR 62 i PAIR 83 PAIR BA 1 4 50` y AL - CXP C NT rxr cm 44 44 *4 *4 *4 44 44 44 44 4- +4 *4 + 4- 4 44 4 4* 4 4 4 44 *4 44 44 +4 4 4 44 +4 44 "TtTTfil PERCENT AGE 86* 9% TOTAL POPULATE SIZE - 64 fXP CM 095\A6 oo TOfHTT.Al. OOH? I r B`l , b, ]< v i . Cl j,i t V aj % ,e " "! te' i ,F ..! | VARIABLE RB C PAIR *50 PAIR 51 PAT R 52 PA J R 53 PAIR 54 PAIR 55 PAIR 56 PAIR 57 PAIR 58 PAIR 59 PAIP 60 PAIR 61 PAIR 62 PAIR 63 PAIR 64 PAIR 65 PAIR 66 PUR 67 PAIR 68 PAIR 69 paTr 70 PAIR 71 PAIR 72 PUP 73 PAIR 74 PAIR 75 ' PAIR 76 PAIR 77 PAIR 78 " PAIR 79 PAIR no PAIR 81 PAT R 8P PAIR fl3 PAIR 84 -------- -- _ '4 NORMAL - rRNORKAL* IXP CNT EXP CNT 44 44 44 44 44 44 44 4 44 44 44 _ --- 44 44 4 4 ' 4' 4 4 4 ~4 4 4 4 4 * ~ 4 4 4 4 4 4 4 - 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 EXP CNT TOTAL " ` 76 PERCENT AG T 90.AX TOTAL POPULATION SIZE - BA 8"` 0.0X " ' ft A.7X rxp CNT a 4.71 00 095148 CONFTDETNTTAl DO 095149 CONFTDFNTIAl '\ ; ! tz> ^ ssS 774r <;z&z^ - &z^a*vu. *S7&&>y or 4MSJPS 4/0/&C&Z0&&* %>&<z>^<c^e's ^uar ^7-<400*540*7*4 Z>S\S/5*'4>*7 7Z& /^UP/SfM^ ^V0 - -5/' S"^/) ^>=- *S7 *0UASs*U7j *Zi/r*&**A/ ^ a%* TTr&S-S ^TZAPyS *>S &Sp/;!0y&&S` <7*77/ &4&ZA7i60; 7C7^S*/*e^S 7& >*SJ<4asG TTCtoe s4r ^ust&sAAMet t&si/!*s^<a*s {AteZe&Ae>&&, s^&c&e<$&Z> 70 <& " Zrylzas? 4>Jt73S ^7ZKS>j* \), dteOL&O'TT&AS <2* **9&&e7tTOsfy 4SM> ^4^ s&*<Z> /Ot0!^5' <fl^ 14&C&&S <&?zr >Wi^ -3M5e /^O^S' STteoUJi/Y -w iSJST <y/- 4#frG0#& &Z&S^S*?7tULs s&7Z?if'<2>S, &edst&<Z&n &&/ &*&*> *SS*7* *S^7~ S&G0a&*6e` 7$>J&Z0*i>Z*V7& 77 s7~ *t*SZ> /?5z/775 4S1 MgXsC- 7SS7~~ Wfre&JZs 7s&4x4 *4*iZ> && &eu&nr>$ <45 >677>*sr- *?*&>*z, 777d7^ 7C^e7re4A^9/& 7G/>, % \ DO 095151 CONFIOF.NTTAI- yr$y J, *^/v/r/<ay' Ay pyy aKpSA *sxye>y S=&*k4&4>y&. /WS' <3^eT* /99^a?-o /^p^Ar^v^b <?/V &j&?/S4Ss>y. s& ^*4?/ /l?/Ji^>7(av y&y #SF SAS yp&yy*/y OXS##' &z**>y. ZVy &^xEV# <** &4Z*y&e prygr y^y&yg&ye? y&yts y >& yvy y^yyjcssypesgJzZi #?<&?/(' y/psp&egp p?pmZ> tz&t&s&G&pA&e. /^^5" /?as^/^ <3^ p//y <2>&94&zfc&< /&/yp/stf'aep*^<s&si& y&y*&r y: *&Sy?-ss fz>s$syo^ ypz&py&pyz A&p/y&ys'). y/&yys- y'y>y>&&appp& yo&yy yzoe*#4z>y/ y/y yysS7&y, y yysr~/#yS& y&v&e<&ie&5 $y sdjty* ^4*/&ssa*M y^s&vA/i> zv#y yy&e /&?&&&> s*s y*v? 4#xvsas ^y^sy^yy ?&#r &&ac2> *#h& yyr&yz?^ 7z//k&i/y y&Cs&d? &(/%>'/&'* 7i*4&: s&ppy/y*' a>& a>//y//M%> ^z> ^f* ^%r &to&#*&o3 sdofryr' sOt#? 7?y ytypy/z jz yyywr. &/& we <Uw%& y/sy u*&& A&\&ex?y&2>s ^ /&&>&?<*&> s&A&dTze*/ 4F $&?/*&&?& *yy$ ya&p&^r ^&sy/A &*/<$^s'l&sspee&t&ifr'> Ta/tfr 77/C &&S&0J&& <g/r a? yucc TW/e? &gk? <&&/&&'` ~ --- v, _ c- _jr ,.--.-^.vV^ -^.-* J"-_~ - , ppy 77*"^ /=&sySy> azpk/> y?e#'&szy24&0a&vr~ yspy? /^fl58wiK6 f ses&sttys/> 7z> s&te y&GA&s* * yy$s *M4as 6 p&*yy&AS' y&rsose 7z> y/c /<?j>o /#&&&& yvi&AsTzxtf $pcasp?j x<as> y/or- &&*/ &e>zk> cuyr- X&u/2&&. yeSsy'/z PT/C /y?ysyyypap?ysvus y?y 7p/- ^ppycA<&yy$ /?/$7xz*v2> yss yyy y&e$y y yfra&eM*r? auas &&&&>&& &/yty /jpypy7?yy<y2> yypppyy^$ yeyppp p^y y*e$r- ys&y, y$ a&xy. & y/& -Csppyxt&yy *sy&yzt$ /2&csi<ms(., yp^ papas' /?4y7- <#^x?ssy*y& appos AA&A^y4X?sAsey sXAffPPspyp pPyZ^yyp, 2>a2PP? yyy& pz> jpyy^^i^y appAp py<?p> yy y<7<+y /&yyp-p&/x&p^iz> s*s 7&y spy# y/yyp-TTs /A'Myps7Zx& 7Z/xy yyyy^ y/&y #y ym? yc/p-z^yy A^y^yys^yssSMS yy ype0#2&Gr&. y? &sz?yyys<eAy ysey yy y>yy<*pyyry& uyyy y^&&Qy^ s*sz>pyp<?yy*sy. yy/y yzs&jyr* yyyyyyytyy yy %&&$& yv?yp0j/&<g- p^ipy sz&eyr&y&yypyiSL ypy y?y<y /#?& MfPpp^y syytyyxsm*& yeoAepp^, yea*? ypy *& y=p^y*yr&. s&jwsjz ?tey>aZxsxgr jyy& 7*&</ y+//?9/fr2& y> ysda&ypyAS /eyytz&Aps ypc yyppL/pyy yy T&ysy; &??yt4y&&& 7z> y&ey&syz&yg' ^ pppy yp&4^7ys yyyyy^^y^yy/pyteyjpy y*s yoyy <?y yy^psyyy^ &/i//&sy>Ay& yyp^yyyU, /?*/>/ypApp yva&yZxc/ye&s* 72 yypyycyps`^?mrys?yysy^sAypyp&^ *pp??p sypy&pv&Ay *y Z&y <$&/*oy ypypyyyypy /** <&%/act?s%> ys pzs/r yyppaappspy: yeovyaesy % DO 095152 CONFIDENTIAL ^ " m #- f\ -- ^ j, .j, ., ,^ , .<' 'Vv '**, \ **' )\ ; ``r'Vi y j '-yi vpifcAf. *kV -\ * f i j t m j a o I a l, _> r w r f a .* * c jlLU i iuj> , L miLL.Jt ,. i >. a u l . iuu I hOr'i, . ' , . ',< JY;-^-VCt.yr;.y>,i*<- - . - . , , ,tr > ' * * ' '`-` `i *'4 ^ /?V>);, , . '*' " * 'J..X J ^ ,, ` - ' - j - +,i'< -,V-^, Y.L.3.* - i a4^--3 . . -7'.4 ,;7/. - 5 - <if^ Ti V' 'X y `TB '' \ '- * / -ft " "` "'-" ' j-.iw-i"- , ,, y>'>' [W,1 ' ' rry-- ^fStTL^'NUM B E R ' F MP.LPVtfcS "IN^OR IGINAL'sVuDY SELECTI ON-!1 KwraS|W}jv$', i*vi`>4i(,,ll''" rare :#*>i-. ? *> 'Vr. 3*-' -- -1 number employees removed from ortginal selection for MISCELLANEOUS reasons -. NUMBER FEMALES IN ORIGINAL STUDY SELECTION - 6 NUMBFR FHPinYrFS IN ORIGINAL STUDY SELECT I ON HAVING PRE-EMPLOYMENT STATUS - `.1I, -.v,1,>;V' -Tr-,, ' - *w* v';.'.:,'.-*< ' -ft lH ,.i ., C .'i- .TOTAL NUMBTR EMPLOYEES ELIGIBLE FOR STUD_Y__GROUP - _2JL NUMBER ELIGIBLE EMPLOYEES PARTICIPATING IN 1EBC HEALTH INVENTORY - 84 NUMBER ELIGIBLE EMPLOYEES NOT PARTICIPATING IN 1980 HEALTH INVENTORY - PERCENTAGE PARTICIPATION OF ELIGIBLE EMPLOYEES TOTAL EMPLOYEES IN, STUDY GROUP - 84 33,3 _ "STW-T" T-T! V.W- tQ) 1 _______ ^_____________ f*, V j -iN I - `, 'AM . . > o| 5SL! 's]-S ?s?,, , t/ "fe. ji v> | inttultik^tyttu^Kr^-, 31 i -; ,,____-- ^ t ..rw^u . . -_______,*c_; DO 095153 CONFIDENT T Al aets&d 4/O&&&0&&' %$&&$& 7Z> <7^>&>V'` >X42e' d&w&iAASA 3>/\fs&<?sss 72& ^/<r ^ --2>/=^v' V/&aJ ^ ^r Z?Q&&*4>e!&` ^5!f2> /af^QV/vS^ /-V' 45^i^x^; *z?2C>y' tpS &SP7*^yeses' &sZ*s 2P*ir TZf&SS -SS&TxcnsJU Xf3&&&e&& sz> <&C<>e S0r *<0t*&sAA*tt <&S%/*Sl*>-<gS trj&e. >eS-C5SC^> -*? -AS 75^ ~ szz'yz&ve <0jcs& &zz/^>y' v, a* *&4A& s6vz> g4/57X*S^tAssee ^k2> /OVJ^^S' ^ S30*7 ^**e tttK&STS -W 4sf <2V~ SZ&ZtsZXUU* S0f<&z*w>s, 72e4d&&&S'. ^/S?r' /^ift^4eS' a^^s#7& 7? sr s4tvz> ^S- /<r>66 77S7^ M&&4uzs 7Z&A&4 **& 4C /egfr&tTG ^7f&e&nrh> dS &U77>ZS7- <**?&**, Z^dT" TS/>. \ 00 095154 OONFTDFNTTAI S&ey&e A&e&eAAC* s&rs&v 4^ W# WAsSfe* &&A*7~ s*X&n*20*-, #;&& ^&v*ur*>*b> &ZX*?/ASssis. ^ &7ZS>y ^ ^sr ^ ;zr ewyx&ve AXs&e axet*y. Z&e **&& ae &4z# ^e z#& 4&e4k&&6ee a&*s >& & we *A9ws*&es&j2> sssw/t' x/sssz*ey SPMZ> a&^4PG*V*#e. W& WSs&sWA <W JA/f /Hostile. <&&*t0W? /&/> //-& j&SsA/iCr ; /^-// ^5?s*sS45^2> *?*/*&/*&}. al&ss*'*- s^A&Aea/#w W**~ /usze* AUse*42>y' W %*S7&va J ,4 JSST- ^5* ^9ie<^eb -^6 W/&~ Wt& *& /3K^>5' ^r &,iSssmAs T&ftr <3e*t42> stwse s&pwzzsU. ew/u&tfe <&K/2>e &(s%>&vee. 7B4S W4'*'** && &#*/**&> >Z> <&# ZW &W4A&CS /Q&&vr~ w&> 7?* &us&x* -2Z ww/r* av/ze we A/7~ #4e yf 4^* ^ae4^r^ &9ZRU&*g& sStr At SSU*J* P&fte &aed &&/**&& ^ --- ,r . .. .../ ,--, -*sW/ W ` -----;' - - . s07~JVSs:'? S=&sy/?r`> /o^ ,w /%?eaM5i4S6 ^was# /<?SS^<flVi#^ '^S^S' - jteSiS' W4AS & ^/t^T ^^ S/e#A7W /YMZASt&Zj &x*s??J s*/0> /Ye?- J7~ &*/ &*&>&> *&- W*Z> ex&u/z&s*. /&/SV& W# ^ ^SW**tfS> /*/ 7Me w&rr- /ssr-, /? *&*> &ea*Mr? ume &&/>*>* W#?" >^<M^ /^SSJ 7SW >^7- ,4^ ^ ^5^ *S*'^W /c^,^ 7ZPS? ^4#eSM*A6^ AtSAOS A<4S&/&At&<?SASS7~ &tS&>AS* J9&S&&^ 2X&Z? /^Bf /?#I>iSfi>ir ^44? ^ i4%s/" ^ 75^ AXZW-W /Xsts2rss72i*& SZ9AS&&63*?. ^?4<y* WsysVU, ^vr <5^ &^/406&ZS* <4^f5* /&S?SsU'*W*/J&> &O/^0SiSS&QA4if A^Pe AZAP^tt aWWA>asa^^a/ W WefeeZxSGZ*. /? &**a*Ws#as sA'&r- or &#si<a>/&2?'& a&Ge <&ue}&e>&& aas&aa^h^*- //SS?rX*/ or W06& ^W-9^MSfiS- ye^p?&s**tp^s^ yAS XXW /f?e> w#*w sssv&sx^9 ^com^ s*ea*? we eo fte&te&t/*& mx& 7#&a/ w/X>jareb m ^sz^^wa/ &<*&>*& f&e jz&s^/ce ae T&ese e^fUAyees & /werxsssz&z' ^ n## SASt'&v'T&^y/*&aAe&Atf ^s amasS/aas,# &>/i//&s#A/& /MT&e^/AU, /e*/H'77*i ete&*s5xc/,&&. ^ Jkj/w <^^37^/ /}r 7ssr ^ *M2c0Sf2> /AS TZSA* etHUAf&j/AStZ &e*+'T0c*7r'. 00 0951 .SB oonftdfnttai i pzzz' y?syjL&A/& <z>x/Z> s724z>y J/4& &aAX? >*'/>&*'' 0PCs& <4(?a**7&9Aj s&<AaS *?M E^/f?/V'/7?bv' ^ X*as2> &4*&S&J&*/7' 4oaap&.<j yzhAZu*^?z>iPAS. 7& /*zead&>>*Ss6* jT&*as j*s??s 4Sjm*vw# XP ASS~7~ #- XUX s&**e>*uj?2> s8y p&e s&G&sdxu* _________ U>A/S<2*/ &AOS/ ZScaasT" *t/J?6 S*x&ZXJx?2>. S9&%*? 75**aS &SAs7~, A> S2?A>A/tyxX- tUS&tZ*: AJ*G y^AH> av zwr xysp" pv> S&S&4& OtVTX&PA' #*f2*2>4TPS' ^X>ypp TTSd? 727ZU-> /?isAK'-*^ycr. 2*zx?asp-s /&**#?' Aj&te &ye&9tv%> AjGdm* d&&x>Z> ^/* /3^ f&r&Afy*#** a&toQs2*tX- &&e# ZZ/zr Afrs&x^&e/ r ^w/> /Ok2> 2fc*^ J*//V^* &<>XPnsA/jQ xfo*/ ^ *& &&S& ^se^ux^s d&ws&r s*0 M4aJz>*^l /**&/ z&*r szogsexs?- '3MJ/ swv&t&*spb*& S5eae&t??s', /tus# zpxxpsax&z #&*? && xx&r- &&&* z#- *sAVr- A^&e&34se&e*G SZtAtoPS, **/ &&*S&6g Xr&zvr? J&*'0*s' &>*XZ> XKUS< -*> > PD /UAl&T - - XJ.C awdg zaa x&*ute- /^5r* xssp~ #*4 AS7P&?> SAA& A dkAZAXST&esttb &&?<&&&?? &4/as\2) 7Z> A&&SA/ tU&G&AS#- ZH/y&AAZ X?f*0 /*/ Zfr&SiS' Z2&e<U#44. yOMs?Sm M<2# ?&&&&' Zc^/^/xrzrses&rb A^o-oe"- a/*spd^*/ &ZXS ^>vz- e&s- **#>& *Sz#?&azJz* xae. *afcw XD >?>&e0?Avsff ^ XAte&A3Z> 7Z> ^PP^CAS^ A<4Z> aSOZTAS tc&ex' X&sxsyay #tk*? sas t2*a6#w,<i>xs s <zy&?^. zz^ ## <u&0u~ 2? _r ca^-r *&a^/S x 4<*^y yZogzzze zsS> pe> AvSttez* mz> & SXsOetSAcegj ^XASae: &AGSZ&s Ak&tsr </Z> XAS2T /A/ #X/t**X. ^Z **ts ****** * -4* ^Sffx^MBc.^. -*^ JSmAzweisr sua* jss^^ Afrexu**. 7PM44& t> PCX*y*S>, MP s*e& ZPCdUzKS^ AZGn*? ps*# xvear- <&y*4?ssre **? A>z>2>2> 7*> P&zr #ar. Ovae za^s zssr jk*s 4azx*>ar, , ' 4e*'2>y&JPr ewu^zt aav+'aw s^ezr-z^x*'xtsvtesvr* <Sp&xzs<s utee# &&uas>&. zm? 2/&T- #** 7A'fr#M*Z>A&2> X?XXU#<S7- Z***' A0. 2>sk*SPZ>*s *&Z>sO<lC XSZX sAape ssxM&vrzK*/ /v ^ AteAxjms <4 Assr* ^ 77a*s ^/^^a5' /ve>s~ /2&e27'dlA**9>70r& AV ^ /PSZ> A&ZfxXh' AAPy^ZPO^^ 4^5 Z*eSAS>&> i XZ3e X?A*V6!<Zs A%/Z>SPrZ*/< s&*/Z> X&XttoX -</>, x**> 7ZA2if 3^r ^7". Zte*S- Stt#PPlt2SS&.>V> f /(tV p&aP#0AS44& /fi> >5^ ^ -W* Z&AAZGaC' /t?AsZ>s2>AP~S &X& <3^?2?5i sVWP&A/s'ASA /96A^)^ &&*/aS. 7Z/' ^*!*S'^ ^a^/W7><JS g&VPPQex. &axtZ>s2>jdZZ& gf&uxakv&. MS/PZP S&cZ*7Z<0a/' ff~ yxg DO 095157 OONFTDFNTIAl 2 Ly SUMMARY STATISTICS FOR FINAL SELECTION OF ETHYLENE OXIDE CONTROL PO( /1-- ------------ ............................... ....-------------------------------------------- -- - - I TOTAL NUMBER EMPLOYEES IN ORIGINAL CONTROL SELECTION - 13A3 V NUMBER FEMALES IN ORIGINAL CONTROL SELECTION - 127 NUMBER EMPLOYEES IN ORIGINAL CONTROL SELECTION HAVING PREVIOUS EXPOSURE - 87 NUMBER EMPLOYEES IN ORIGINAL CONTROL SELECTION HAVING pre-employment STATUS - T OT L NUMBER EMPLOYEES ELIGIBLE FOR CONTROL GROUP - 1054 ii i DO 095158 CONFTOFNTTAi TITLE: CROSS-SECTIONAL STUDY OF EMPLOYEES WITH WORKPLACE EXPOSURES ' TO ETHYLENE OXIDE AT LOUISIANA DIVISION PROTOCOL NUM3ZR: epi_dr-9999_oo99_2081 DAii: january 2$, 19S1 PROPOSED STARTING BATE:^rebruar>T 19BI SIGNATURES: : Prirary Investigarcr: tkX U~X________ P. P. D, VanPeenen, M.D. Approvals: R. R. Cook, M.D. ( ic,Tk~ (nn (Daze) (Daze) R* j_* Dosn&l K. F. Currier, K.D. (Bite) z-zv-*/ (Daze) zh/tf (Daze) DO 095159 CONF TDFNT T Al epidemiology HEALTH AKD ENVIRONMENTAL SCIENCES DOW CHEMICAL U.S.A. PROTOCOL STATUS: PRELIMINARY-DRAFT PIKAL 12 DRAFT TITLE: CROSS-SECTIONAL STUDY OF EMPLOYEES WITH WORKPLACE EXPOSURES *" TO ETHYLENE OXIDE AT LOUISIANA DIVISION PROTOCOL NUM3ER: EPI-DR-0114-1975-2081 DATE: February 3, 1981 PROPOSED STARTING DATE: February 1, 1981 ESTIMATED DATE OF FINAL REPORT: june 1, 1981 SIGNATURES: Distribution: R. L. Dostal M. F. Currier, M.D. (Date) (Date) OO 0951^ gonftofntTAI EPIDEMIOLOGY HEALTH AND ENVIRONMENTAL SCIENCES DOW CHEMICAL D.S.A. PROTOCOL STATUS: PRELIMINARY-DRAFT T1NAL -Mth DRAFT aix CROSS-SECTIONAL STUDY OF EMPLOYEES WITH WORKPLACE EXPOSURES TO ETHYLENE OXIDE AT LOUISIANA DIVISION r PROTOCOL NUMBER: epi-dR-9999-9999-2181 DATE: janu^y. 2$, 1981 PROPOSED STARTING DATE: February 1, 1981 ESTIMATED DATE OF FINAL REPORT: june i> 3.9&I SIGNATURES: Primary Investigator: F. F. D. VanPeenen, K.D. Approvals: ______________________ _ R. R. Cook, K.D. \. (Date) tnn (Date) J. R. Venable, K.D. (Date) Fv. L. Dostal (Date) no 095161 CONFTDFNTT Al EPIDEMIOLOGY HEALTH AND ENVIRONMENTAL SCIENCES DOW CHEMICAL U.S.A. PROTOCOL STATUS: PRZLIMINARY-DRAFT PINAL 12 DRAPT TITLE: CROSS-SECTIONAL STUDY OF EMPLOYEES WITH WORKPLACE EXPOSURES \TO ETHYLENE OXIDE AT LOUISIANA DIVISION PROTOCOL NUMBER: EPI-DR-0114-1975-2181 DATE: February 3, 1981 PROPOSED STARTING DATE: February 1, 1981 ESTIMATED DATE OF FINAL REPORT: june 1, 1981 SIGNATURES: Investigator: ________Q\a Frirtary --'fU------------- ----------- -- ( V ^ j Distribution: J. R. Venable, M.D. R. L. Doscal M. F. Currier, M.D. (Date) (Date) (Date) DO 095102 OONFTDFNTTM 12th DRAFT February 3, 1981 TITLE: SHORT TITLE: CROSS-SECTIONAL STUDY OF EMPLOYEES WITH WORKPLACE EXPOSURES TO ETHYLENE OXIDE AT LOUISIANA DIVISION Ethylene Oxide Study PRIMARY INVESTIGATOR: P. F. D, VanPeenen, M.D. Epidemiologist Michigan Division Midland, Michigan Telephone: (517)636-4769 CO-INVESTIGATORS: J. D. Burek Veterinary Pathologist 1803 Building Midland, Michigan Telephone: (517)636-0737 R. R. Cook, M.D.Director of Epidemiology U.S. Area Medical 1803 Building Dow Chemical U.S.A. Midland, Michigan Telephone: (517)636-1383 M. F. Currier, M.D. Medical Director Plaquemine, Louisiana Telephone: (504)389-8292 William E. Ledford - Industrial Hygienist Plaquemine, Louisiana Telephone: (504)389-1659 Perry Poston Computer Specialist Plaquemine, Louisiana Telephone: (504)389-8310 DO 09Sl^3 ^FTDFNTTAI OBJECTIVES; To determine whether Dow employees who are exposed to ethylene oxide (EO) have more abnormalities of the hematopoietic, hepatic, renal and reproductive systems, as measured by data collected for periodic health examinations, than do employees without such ex posure. IMPORTANCE OF THE CHEMICAL: Ethylene oxide (EO) is an important industrial chemical, produced from feedstock ethylene, which is used * # primarily for making ethylene glycol or polyesters (Anon, 1979a). It is significant from a health viewpoint that EO is also used for cold sterilization of medical equipment, although this usage is minimal from a commercial viewpoint. U.S. production of EO increased from 1.3 billion pounds in 1960 to 5.3 billion pounds in 1979 (Anon, 1979a). Dow has produced EO since 1941 at the Texas Division and since 1959 at the Louisiana Division. At these locations, McClimans et al, (unpublished manuscript) estimated that at least 550 employees ex perienced some workplace exposures to EO in the past, while Moore et^ al, (1979) estimated that 300 employees had worked in EO production for one year or more. At the present time, there are 88 employees with potential exposure to EO at the Louisiana facility (44 each in Glycol II and Dowanols ). DRAFT PROTOCOL Ethylene Oxide Study/VanPeenen, et al 1. no 095154 conftdfntiw The American Conference of Governmental Industrial Hygienists (AGGIH) lists an 8 hour time-weighted-average concentration (TWA) of 10 ppm for E0 in its Notice of Intended Changes for 1980; an 8 hour 10 ppm TWA was also the Dow IHG, but this has very recently (1980) changed to 3 ppm with excursion limit of 50 ppm. The current OSHA permissible exposure limit (PEL) is 50 ppm TWA. REVIEW OF LITERATURE: Acute intoxication with E0 at high concentra tions has long been recognized as a cause of irritation of eyes, respiratory tract and skin, and central nervous system (CNS) depres sion. Delayed effects include headache, nausea, vomiting, inco ordination, and electrocardiograph abnormalities (Proctor & Hughes, 1978). . Peripheral neuropathy with decreased nerve conduction velocity has recently been reported in 4 men after overexposure to EO (Gross e_t jil, 1979); whether cause and effect can be assumed in these cases is debatable. Levels responsible for acute effects, have been esti mated at above 2000 ppm for very short single exposures and' above 200 ppm for continuous 8 hour exposures (Hine and Rowe, 1963). According to Dr. Ray Flake of the Texas Division, acute overexposure of humans to EO may cause an increase in serum SGOT and SGPT levels (Fishbeck, personal communication). DRAFT PROTOCOL Ethylene Oxide Study/VanPeenen, et^ al 2. DO 09S16S CONF TDFNTTA! The experimental toxicity of EO was reported by Dow toxicolo gists more than 20 years ago (Hollingsworth et al, 1956). There were no effects at 49 ppm, but histologic changes in liver, kidney, and testes were observed in some animal species at repeated seven- hour vapor exposures at levels of 204 and 357 ppm. Reversible impair ment of nervous function and muscular paralysis were observed in rats, rabbits and monkeys at the 204 ppm level dose. Interestingly, laboratory results, including blood urea nitrogen and blood counts, were normal in the species studied even at the higher (357 ppm) exposure levels (Hollingsworth ejt 1956). Recently, there has been renewed concern'about possible chronic toxicity of EO. Results of an industry-sponsored 24-month experimental inhalation study showed an increased, dose-related incidence of mono nuclear cell leukemia in male and female Fischer 344 rats (Snelling et al, 1981). Effects were voted at all level of exposures: 100, 33, and 10 ppm. Past literature in these regards included a 1964 report of tumorigenicity in female mice (Reyniers, 1964), alkylating agent pro perties (Ehrenberg et al, 1974), and demonstrations of mutagenicity in several test systems (Anon, 1979b), including Neurospora (Watson, 1966), plants (Ehrenberg and Gustafsson, 1957), and the Salmonella typhimurium strain TA 1535 bacterial assay (Embree, 1975). DRAFT PROTOCOL Ethylene Oxide Study/VanPeenen, et a^L 3. CoN rnrQ^e. 1 As regards human disease, there are two recent published reports from Sweden of malignancies purportedly linked to EO exposure. The first report was of a case each of acute mylegenous leukemia, chronic myelogenous leukemia, and Waldenstrom's macroglobulinemia diagnosed between 1972 and 1977 in employees of a small factory which used EO for sterilization (Hogstedt et al, 1979a). These cancers may, of course, be unrelated; but in a subsequent retrospective mortality study of employees of an EO production facility, the same senior author reported more observed deaths from leukemia (3) than would have been expected (less than 0.5) using Swedish national death rates (Hogstedt et al, 1979b), Two of the leukemia deaths were caused by chronic lymphatic leukemia, one by acute myeloid leukemia. It is difficult to interpret results from this paper because of the finding of an elevated overall standard mortality ratio (SMR) with 47 deaths vs. 27 expected. Such a finding would be extremely unusual in the U.S. where most U.S. employed populations reflect the ''healthy worker effect", and experience overall SMRs less than 100%. Another unusual finding was an increased SMR for diseases of the circulatory system (20 observed vs. 13 expected), which is reported only in office employees in the U.S. Finally, chronic lymphatic leukemia has not, to this author's knowledge, been linked to environmental factors. A retrospective cohort mortality study of employees with work place exposure to EO is planned by British colleagues at Imperial Chemical Industries Limited (ICI). Results of the ICI study will be helpful in interpretating the Hogstedt reports. DRAFT PROTOCOL Ethylene Oxide Study/VanPeenen, et_ l^ 4. .? The Russian literature includes two reports of a variety of health problems associated with workplace exposures to EO. Yabukova et_ al (1976) reported an increased gynecological sick rate, increased complications of pregnancy, and increased incidence of interruption^" of pregnancy among employees of an EO manufacturing plant compared to pregnancies in management personnel and from outside the factory. However, if I read a translation of the paper correctly, although spontaneous abortion occurred in 10.5% of pregnancies of the exposed group of instrument controllers compared to about 8% of laboratory workers (also exposed) and about 8% of management controls, there were more stillbirths among management employees (A/65) than in instrument controllers (1/57). According to Yabukova e al (1976) the "overall corresponding contamination of the-factory air did not exceed 1 mg/m3" (= about 0.6 ppm). Ostravskaya e_t a_l (1971) reported a variety of subjective com plaints from 76 individuals with 3-6 years work experience, presumably in EO manufacturing. In addition, they reported "bradycardia", "vegeta tive shifts" and "EKG changes" in some; the significance, if any, of the quoted abnormalities is difficult to assess from the translation. 3 Ostravskaya ej: aT (1971) reported a leukopenia (A00O-A400 WBC/mm ) in 9 of the 76 employees. DRAFT PROTOCOL Ethylene Oxide Study/VanPeenen, et al 5. DO 0951 OB CONF T DFNT 1 Al There was only one U,S. epidemiology study of EO prior to the recent interest in carcinogenesis. A cross-sectional study of 37 EO operators engaged in manufacturing operations and of age-matched controls (operators working with chemicals other than EO) was reported by Joyner in 1964. The highest air level of EO recorded was 127 ppm but average exposures were less than 10 ppm for the working day (Joyner, 1964). The EO group actually had fewer symptoms and less absenteeism than the controls; and there were no more "abnormal" laboratory findings in the study group than among the controls (Joyner, 1964). More recently, in a Dow Chemical U.S.A. response document (Moore e al, 1978), it is stated (page 7) that health surveillance data from 129 Dow employees "revealed no unusual medical effects, i.e., no effects were observed in employees potentially exposed to EO within the TLV". Tests apparently included blood counts and a battery of clinical labora tory tests. A cross-sectional study of individuals involved in cold sterilization using EO was recently published by Garry ej: al, 1979, who found various frequencies of symptoms (up to 9 with histories of sore throat) among 12 EO-exposed employees versus no symptoms of any kind in a control group. Similarly, cytogenetic studies by these authors reportedly showed more frequent sister chromatid exchanges (SCE) (average 8.65) among the 12 employees than among 11 controls (average 6.37). DRAFT PROTOCOL Ethylene Oxide Study/VanPeenen, et al 6. t A cross-sectional study was also sponsored by the American Hospital Supply Corporation (AHSC). As yet unpublished, the AHSC study was of employees of 9 U.S. manufacturing facilities which utilize EO as a sterilant gas. According to data kindly provided by Ronald H. Abrahams, Ph.D., director of Regulatory Compliance of -- * the AHSC, 6 of 75 exposed employees had slight anemia and one had a slight increase in lymphocytes (there were apparently no controls for this phase of the study). Sperm counts were obtained from some of the employees: 8 of 46 exposed and 1 of 9 controls had sperm counts less than 20 million/cc. The AHSC study also included cytogenetic data: except for stable forms (Cs), all modalities of chromosome aberration including number of breaks, unstable forms (CV), and SCEs, were statis tically increased in the exposed group when compared to control groups of office workers. RATIONALE FOR RESEARCH: There is a growing volume of experimental toxicological data indicating that EO, at chronic vapor exposure levels much higher than experienced in the Dow workplace, can cause detectable health effects. Whether measurable effects might also occur at lower levels of exposure is unknown, but, if so, they might include those listed in Tables I and II. A major concern about long-term effects of workplace exposures to EO is carcinogenesis, so * Industrial Hygiene data were not given, but vapor levels were stated to be less than the OSHA 8 hour TWA limit of 50 ppm. DRAFT PROTOCOL Ethylene Oxide Study/VanPeenen, e a_l 7. DO 095170 CONFTDFNTTAl it would seem reasonable that any epidemiological study be designed not only to detect the phenomena listed in Table I over the short term, but to detect excess cancers over the long term. In fact, a draft proposal for a retrospective cohort mortality study of EO pro duction personnel at the Texas Division, with leukemia as a primary concern, was prepared by McCliraans and colleagues in July 1979 (un published data). Very recently, Bond and colleagues have prepared the protocol for a case-control study of leukemias at Texas Division. Our proposal is for a cross-sectional study of Louisiana Division employees involved in jobs with potential EO exposure, to be followed by continuous surveillance of the same group of employees. Since cross-sectional studies are preliminary and do not usually yield definitive results, it will be necessary to plan for appropriate follow-on research depending on these preliminary findings. Alternatives for follow-on research are detailed on page 11. Abnormalities which might be expected if workplace exposures to EO were exerting a chronic toxic effect include hematological, hepatic, reproductive, renal, cytogenetic, and carcinogenic (Glaser, 1977). As proposed at the present time, however, our project will be limited to analysis of results from existing medical surveillance data. Hemato logical abnormalities reportedly associated with exposure to EO include anemia and absolute lymphocytosis (Ehrenberg and Halstrom, 1967), leucopenia (Ostravskaya et_ al, 1971) or, simply "changes in peripheral blood" (Spasovsky e_al, 1978). Because of the reported dose-related DRAFT PROTOCOL Ethylene Oxide Study/VanPeenen, et al 8. OO oonftdfnttai alkylation of hemoglobin by EO (Osterman-Golkar et al, 1976; Calleman et al, 1978) it would not be unreasonable to expect increases in' laboratory tests associated with hemolysis, including total serum biliru bin and lactic acid dehydrogenase (LDH), as well as anemia and reticulocytosis. Similarly, direct hepatic and renal tubular damage, causing abnormal results in simple measures of liver and kidney function, might be expected from any alkylating agent and, in fact, have been reported experimentally at high levels of EO vapor exposure (Hollingsworth e al, 1956). Questionnaire data from the periodic health examinations are available and can be used. We recognize the weaknesses and biases of using questionnaire data as evidence for disease or for adverse repro ductive function, but by judicious and a priori designation of which questions will be analyzed it should be possible to compare the pre valences of symptoms relating to diseases of the blood and of adverse pregnancy outcomes. The list of effects which can be studied using existing data is in Table I. Table II is a more comprehensive list, including tests which cannot, and, in certain cases, probably should not, be studied / at the present time. As more reliable techniques become available, however, it may become appropriate to examine for additional effects in the future. Table III lists data elements which are presently available for extraction and analysis. DRAFT PROTOCOL Ehtylene Oxide Study/VanPeenen, et ^1 9. 00 09517? CONFIDENT! Al APPROACH: Exposed Population. The exposed population will consist of active A Louisiana Division/vemployees who were assigned to Glycol II or Dowanols for at least 6 months prior to participating in the HealthInventory examination during 1980. Since intensity of exposure cate gories by job cannot be estimated because detailed historical industrial hygiene data are lacking, length of employment in the 2 plants will be used to divide the exposed group into subsets of longer and shorter dura tions of exposure. Control Population. There are 88 persons eligible; but since it is anticipated that-nen-whiteo-and females will be few, only data from whifre males will be included in the study. The control group will consist of an equal number of males matched by age (within 5 years or better), race, date of hire, and smoking and alcohol use history. The job category will be as of the date of health inventory and, of course, results of the 1980 health inventory examination must be available. Finally, to be eligible as a control, the employee must have no history of EO exposure. It is recommended that the initial pool for controls con sist of all Louisiana Division employees, and that eliminations be made by matching criteria, the best age match being the final decision step. DRAFT PROTOCOL Ethylene Oxide Study/VanPeenen, et^ al 10. Dependent Variables. Dependent variables are listed in Table HI, and consist of answers to questions regarding reproductive history and past medical history, complete blood count values, and results of the liver and kidney function tests. Test results will be considered abnormal if outside the reference ranges shown; a summary of some re cently published ranges is in Appendix i. Covariables. Covariables or potential confounding variables other than those used in matching include family history of illness, use of medications, exposure to chemicals other than EO, radiation exposure, and substance abuse. The effects of some of these covariables cannot be analysed, and others, such as known substance abuse or use of certain medications might best be handled by elimination from both control and exposed groups. Those with a family history of blood disorders need not be eliminated, but the frequencies of such histories in the 2 groups should be checked. Obviously, affirmative answers to questions such as "Did you have a birth abnormality?" should be considered if the abnormality were considered hereditary. Analysis of Data. Data will be collated and analysed using the com puter system available to the Louisiana Division Medical Department and capably programmed by Perry Poston. The SAS statistical software system will be used for analyses. The null hypothesis of no statis tically significant difference in values from the exposed and control samples will be by the jWilcoxon signed rank test for continuous data and McNemar's test in the case of dichotomous data, I Continuous data will also be analyzed .by McNemar's test using the'categories of .within or without the normal range as defined in Table III. CONFTOFNTTAl DRAFT PROTOCOL Ethylene Oxide Study/VanPeenen, et al 11. RESULTS, WEAKNESSES, AKp FOLLOW-ON STUDIES: This proposed cross- sectional study should indicate whether or not employees potentially exposed go have relatively depressed red cell indices or depressed or elevated white cell counts compared to their non-EO-exposed col leagues; whether results of their laboratory tests of renal and hepatic function are comparable; and whether their reported prevalences of adverse pregnancy outcomes are similar. Identification of potentially exposed groups will also permit continued future surveillance and such groups would also be available for testing as new techniques become available. Concern has been expressed about ability of this study to detect leukemia. By definition, leukemia would, in fact, be detected by the combination of anemia associated with increased number of white cells in peripheral blood, and the diagnosis is easily confirmed by microscopic examination of bone marrow. Cross-sfectional studies have many weaknessess which need not be detailed herE. The proposed study, however, additionally suffers from the following; DRAFT PROTOCOL Ethylene Oxide Study/VanPeenen, et al 12. HO 095175 CONFTDFNT TAl 1. Pregnancy outcome data from fathers are less reliable than from mothers. 2. The questionnaires permit bias towards "socially-acceptable" answers. 3. The "power" or ability to detect subtle differences between exposed and control populations may be low. 4. It is difficult to separate out age and work-ethic effects when exposure categories are defined by duration rather than intensity. For these reasons, we suggest that a negative result (i.e., no significant difference in dependent variables between exposed and con trol groups) not be considered the final word. Thus, it will be appro priate to continue to monitor employees with potential EO exposures. Populations to be included in continuing surveillance (pros pective morbidity) studies will be those identified for this crosssectional study, with addition of those entering or leaving jobs with potential EO exposures. Controls who enter such jobs will join the exposed group and will be replaced as controls using the original match ing criteria. Personnel new to EO exposure jobs, and staying there for 6 months, will be added to the study along with new matched controls. Personnel leaving EO exposure jobs will continue in the study, but, of course, in a different category. DRAFT PROTOCOL Ethylene Oxide Study/VanPeenen, e_t al_ 13. 0g5\76 DO TDFNTl Al Similarily a positive result (i.e., statistically significant difference in dependent variables between exposed and control groups) cannot be considered definitive either. In order to avoid decisions based on chance due to multiple comparisons, it is suggested that future studies at this time be stimulated only if a positive result is found for more than 2 laboratory tests (of the 15 to be analysed) or for more than 2 questionnaire responses (of the 11 analysed). In such an event, a retrospective morbidity study of a cohort of all EO-exposed employees at some point in the past (I'd recommend a period prior to the first Health Inventory) should be designed. Such a study might permit evalua tion of the question of "which came first, exposure or morbidity?" DRAFT PROTOCOL Ethylene Oxide Study/VanPeenen, et al 14. DO 09S177 00NFTDFNTTA1 People. Estimated personnel needs are as follows: Individual VanPeenen - Epidemiologist, Midland Currier - Medical Director, Louisiana Cook - Director of Epidemiology, U.S. Area Poston - Computer Specialist, Louisiana Ledford - Industrial Hygienist, Louisiana TOTAL Time (Hours) 120 80 20 60 40 320 Hours / Dollars: For personnel, 320 x $55.00/hr = For computer time, estimate about $17,600.00 4,000.00 The computer estimate is for analyses only and does not include gathering and storing health data; It is based on a one month's cost incurred in our department during a period of heavy computer usage. For travel and miscellaneous 3,000.00 TOTAL............ ............................................................ $24,600.00 TIME TO COMPLETION: Approximately 3 months from approval of protocol to final report. DRAFT PROTOCOL Ethylene Oxide Study/VanPeenen, et al 15. Oq COA/p- 09s 7 rO F7v TABLE I. EFFECTS WHICH CAN BE MEASURED USING EXISTING DATA \ \ \ \ Expected Effect (Dependent Variable) How Measured 1. Hematopoietic Anemia lymphocytosis or lymphocytopenia Red and white blood cell counts 2. Hepatotoxicity 3. Mutagenesis 4. Renal tubule toxicity Liver function tests, liver enzymes "Adverse pregnancy outcomes" according to questionnaire Kidney function tests Why Direct toxic effects on red and white cell pre cursors: stimulation of the lymphocytic system Direct toxic effect Alteration of germinal stem cells Direct toxic effect n0 o^7C5 OONFTDFNTI^ TABLE II &^ EFFECTS WHICH MIGHT OCCUR IF EO EXPOSURE WERE TO CAUSE ADVERSE LONG TERM EFFECTS ON HUMAN HEALTH, BUT WHICH CANNOT BE MEASURED USING EXISTING DATA AND WILL NOT BE ANALYSED IN THIS STUDY* * Expected Effect (Dependent Variable) 1. Hematopoietic Anemia 2. Cytogenetic 3. Spermatocidal Depression of spermatogenesis A. Immunotoxicity 5. Neurotoxicity How Measured Reticulocyte count (Chromosome aberrations, sister chromatid ex changes) Sperm counts FSH levels Counts of T&B cells Mitogen stimulation Depressed antibody formation Measures of nerve conduction Tests for tremor Why Hemolysis * Direct effect on DNA in dividing cells Toxicity to germinal tissue DNA effect; direct toxic effect on bone marrow and blood Reports of neurological effects Abbreviations: DNA = deoxynucleic acid; * FSH = follicle stimulating hormone; T&B cells = thymus and bursa derived cells. DO 0951 BO CONFTDFNTT Al TABLE III DATA ELEMENTS WHICH CAN BE EXTRACTED FROM HEALTH INVENTORY EXAMS LOUISIANA DIVISION A. Dependent Variables. 1. Prevalence of Positive Answers on Questionnaire Items. a. "Has your family had any of the following: (1) Birth Abnormality. "Analysis will be by those who checked child only as well as by those who checked child and'self. (2) "Any Miscarriages?" Analyze by "Yes" or "No" and by numbers per 100 respondents. (3) "Any children with abnormality at birth?" As above. (4) "Any children stillborn?" As above. (5) "Any death under 6 months?" As above. b. "Do you have or have you had any of the following?" Analyses will be by prevalences of "Yes" answers. (1) "Liver trouble." (2) "Kidney trouble." (3) "Anemia or blood problems." (4) "History of blood dyscrasias." (5) "Abnormal function of blood elements." (6) "Liver or renaldysfunction." Analyzed with and without an additional yes answer to "Exposure to drugs or chemicals that injure the liver"; "Exposure to benzene or other bone marrow depressants" and/or "Outside exposure to possible marrow toxins." - O 9 S1 ^ 1 no nftdfniia, TABLE III (continued) i* 2. Results of Laboratory Tests - Hematology a. Hgb Reference range (RR) : 14-18 Gm% b. Hct c RBC d. WBC RR : 40-54% 3 RR :4.6-6.2 million/ram RR :4.5-11,000/mm3 e Percent lymphocytes RR : 25-33% Results of Laboratory Tests - Kidney a. Blood Urea Nitrogen RR : 8-26 mg/dl b. Serum Creatinine RR :0.9-1-4 mg/dl c. Urine Protein 1+ or greater RR :neg d. RBC; WBC RR :2; 0-3 Results of Laboratory Tests - Liver a. Total Bilirubin RR ; Less than 1.6 mg/dl b. Alkaline Phosphatase RR :35-148 units c. LDH RR :63-155 IU/L d. SGPT RR :12-53 units e. SGOT RR :15-55 units f. GGTP RR :under 38 U/L Reference ranges for blood tests according to Conn (1971), vhich are the most conservative of those reviewed - See Appendix i. A Normal ranges for liver and kidney tests according to Bio-Science Laboratories, 7600 Tyrone Ave., Van Nuys, Calif. 00 095187 rONFTDFNTTAl TABLE III (continued) Characteristics (Potential confounding variables). Prevalence of routine answers on questionnaire items. a. Past history of birth abnormality, leukemia, blood disorder or cancer in any family member. b. "Yes" on "Exposures other than at work" c. Diabetes d. Asthma e. Arthritis Physical Exam a. Height b. Weight c. Diastolic blood pressure LITERATURE CITED: 1. Anon (1979a) Chemical products synopsis ethylene oxide. Report issued by Mannsville Chemical Products, Box 232, Cortland, N.Y. 13045, dated October, 1979. 2. Anon (1979b) Has any research shown ethylene or ethylene oxide to be carcinogenic? If so, in what concentrations? Questions and answers in Brit. Med. J., issue of 5 May, 1979. 3. Browning J.B. (1979) Letter to Document Control officer. Environmental Protection Agency dated 13 Spetember, 1979. Browning is Corporate Director for Health, Safety and Environmental Affairs, Union Carbide Corporation, 270 Park Avenue, New York City, N.Y. 10017. 4. Calleman C.J., Ehrenberg L., Jansson B., Osterman-Golkar S., Segerback D., Svensson K. and Uachtmeister C.A. (1978) Monitoring and risk assessment by means of alkyl groups in hemoglobin in persons occupationally exposed to ethylene oxide. J, Environ. Path. & Tox. 2^:427-442. 4a. Conn, H.F. and Conn, R. B., Jr., editors (1971) Current Diagnosis _3. Philadelphia, W.B. Saunders Co. 5. Ehrenberg L. and Gustafsson A. (1957) On the mutagenic action of ethylene oxide and diepoxybutane in barley. Hereditas 43:595-602. . 6. Ehrenberg L. and Hallstrom T. (1967) Work by these authors reported by Railings L.O. in a discussion section, pages 327-334, of the International Atomic Energy Agency Proceedings Series Radiosterilization of Medical Products. Vienna, 1967. OO 095184 OONFIDFNTTAl LITERATURE CITED (Pg. 2) 7. Ehrenberg L., Hiesche K.D., Osterman-Golkar S., and Wennberg I. (1974) Evaluation of genetic risks of alkylating agents: Tissue doses in the mouse from air contaminated with ethylene oxide. Mutation Res. 24:83-103. 8. Embree J.W. and Hine C.H. (1975) Mutagenicity of ethylene oxide. Toxicology and Applied Pharmacology 33:172-173. 9. Garry V. F., Hozier J., Jacobs D., Wade R. L., and Gray D. G. (1979) Ethylene oxide: evidence of human chromosomal effects. Environmental Mutagenesis _1:375-382. 10. Glaser Z. R. (1977) Special occupational hazard review with control recommendations. Use of ethylene oxide as a sterilant in medical facilities. Washington D.C.: Government Printing Office, DHEW (NIOSH) Pub. No. 77-200. 11. Gross J.A., Haas M.L., and Swift T.R. (1979) Ethylene oxide neurotoxicity: Report of four cases and review .of the literature. Neurology 29:978-983. 12. Hine C.H. and Rowe V.K. (1963) Epoxy compounds. Chapter 37 in Patty F.A., editor, Industrial Hygiene and Toxicology. 2nd edition. Vol. II. Toxicology. N.Y. .* John Wiley and Sons. 13. Hogstedt C. , Malmqvist M., and Wadman B. (1979a) .Leukemia in workers exposed to ethylene oxide. J. Am. Med. Assoc. 241(11): 1132-1133. DO COMp LITERATURE CITED (Pg. 3) 14. Hogstedt C. , Rohlen 0., Berndtsson B.S., Axelson 0., and Ehrenberg L. (1979b) A cohort study of mortality and cancer incidence in ethylene oxide production workers. Brit. J. Indust. Med. 36:276-280. 15. Hollingsworth R.L., Rowe V.K., Oyen F., McCollister D.D. \ and Spencer H.C. (1956) Toxicity of ethylene oxide determined on experimental animals. A.M.A. Arch. Ind. Health 13:217-227. 16. Jacobson K. H. , Hackley E. B., and Feinsilver L. (1956) The toxicity of inhaled ethylene oxide and propylene oxide vapors. A.M.A. Arch. Indust. Health 13:237. 17. Joyner R. E. (1964) Chronic toxicity of ethylene oxide. Arch. Environ. Health _8:700-710. 18. McClimans C., Daniel R. , and Halphen C. (1979) Ethylene oxide protocol draft. Unpublished manuscript dated July 29, 1979. 19. (Restricted) Moore W. M., Dalman D. A. and Fraser E. B. (1978) Dow Chemical U.S.A. Response to the recommended Toxicology testing for alkyl expoxides by the interagency testing committee -- ethylene oxide. CRI report 78-2092 dated March 20, 1978. DO fOFNTTAl CONF * LITERATURE CITED (Pg. A) \ \ 20. Osterman-Golkar S., Ehrenberg L., Segerback D. and Hallstrora I. (1976). Evaluation of genetic risks of alkylating agents. II. Haemoglobin as a dose monitor. Mutation Research 34;1-10. 21. Ostravskaya, R.S., Nabieva, G.V. and Dushina, N.N. (1971) Health of Workers in the Manufacture of Ethylene Oxide by a Method of Catalytic Synthesis. Tr. Ufimskii Navchno Isledovatelki Institute Gigiena Prof. Zabolevaniia 6^89-93. (Dow Translation 80-11-61) 22. Proctor N.H., and Hughes J.P. (1978) Chemical Hazards of the Workplace. Phila: J.P. Lippincott Co., Pages 262-263. 23. Reyniers J.A., Sacksteder M.R., and Ashburn L.L. (1964) Multiple tumors in female germfree inbTed albino mice exposed to bedding treated with ethylene oxide. J. Natl. Cancer Inst. 32:1045. 23a. Snellings W.M., Weil C.S., and Maronpot R.R. (1981). Final report ethylene oxide two-year inhalation study on rats. Bushy Run Research Center Report 44-20 issued January 28, 1981, Pittsburgh, Pa. 24. Spasovsky M., Hristeva V. , Kirkoff V., Pernoff K., Panova Z. Dryanovska T., Bobeff G., Stamova N., Kapurdoff V., Gintcheva N., Ivanova S., and Ivanova I. (1978) Epidemiological study on ethylene and ehtylene-oxide production. XIX International'Congress on Occupational Health. September 25-30, 1978. DO 095187 CONFTDFNTTAl LITURATURE CITED (Pg~. 5) 25. Watson W.A.F. (1966) An essential difference between the * genetical effects of mono - and bi -- functional agents. Mutation Res. 3^455-457. 26. Yabukova Z.N., Shamova H.A., Muftakhova F.A., and Shilova L.F., (In Russian) (1976) Gynecological sick-rate of working women occupied in the production of ethylene oxide. Kazanskii Meditsinskii Zhurnal 57(1):558-560. DO 095188 CONFTDFNTTAl 7' A DISTRIBUTION-FREE SIGNED RANK TEST 27 Paired Replicates' Analyses by Way of Signed Ranks data. We obtain 2/t observations, two observations on each of n subjects (blocks, patients, etc.)'. Subject i A'. Y. lyses in ulation. portant. pairs of ncerned second ' from a plicates ion-free cd rank ntcrval. . data, ierlying ^ lored in ^cussed lion 4, ; related ires are of the he sign 'ample isymp- Assumptions Al. We let X{ and lake as our model Z, = 0 + e,, / = 1,... ,/i, (1) where the e's are unobservable random variables and the parameter of interest 0 is the unknown "treatment" effect. A2. The e's are mutually independent. A3. Each e comes from a continuous population (not necessarily the same one) that is symmetric about zero. 1. A DISTRIBUTION-FREE SIGNED RANK TEST (WILCOXON) Procedure. To test //,,:0 = 0, (2) 1. Form the absolute differences |Z,|, ... , |Z,,|. Let R, denote the rank of |Z,| in the joint ranking from least to greatest of \Zt\,. . . , |Z,,|. 2. Define indicator variables w i = 1where fl Vi = [o if Zi> 0, if z,. < 0. 3. Form the n products Rt, Rnyn, and set t-i (3) The product R,yt, is known as the positive signed rank ofZ,. It takes on the value zero if Z, is negative and is equal to the rank of |Z,| when Z, is positive. The statistic T+ is the sum of the positive signed ranks. DO 095189 confidential! 28 THE ONE-SAMPLE LOCATION PROBLEM 4. For a one-sided test of H0 (2) versus the alternative 0 > 0, at the a level of significance, reject //,, if T+ ^ /(a, ri), accept H0 if T+ < /(a, n), (4) where the constant /(a, n), which satisfies the equation P0{T+ > /(a, )} = a, is obtained from Table A.4. (Also see Comment 8.) For a one-sided test of H0 (2) versus the alternative 0 < 0, at the a level of significance, rej.ec.t H,,a accept H0 if ,,T+ <,, --(-'---+----1-)----t(,a,, n), i rf T ^>/n-I-I(-n/-I-----II------1I-)-I--- t(a, ).,, (5) For a two-sided test of H,, (2) versus the alternative 0 ^ 0, at the a level of significance, reject H0 accept //0 n(n +1) if T ^ /(ot2, n) or T <, - f(i, n). if ~L?.? _ (#li n) < r+ < l(a2, n), (6) where a = a! -f- a2. Large Sample Approximation. When Ha is true, the statistic has an asymptotic (n tending to infinity) 7V(0, 1) distribution. The normal theory approximation to procedure (4) is reject //,, accept 7/0 if T*^zM, if T* < zM. (8) Tics. If there are zero values among the Z's, discard the zero values and redefine n to be the number of nonzero Z's. If there are ties among the (nonzero) |Z|'s, use average ranks to compute T+, and carry on as in (4) to (6). For the large sample approximation, compute 7'+ using average ranks, and replace var0(7,J) in (7) by var0 (T+) = (24) n(n + 1X2/1 + 1) 2i=i lX'i + 1) where g = number of tied groups and /, is the size of tied group j.* (9) * In (9) an untied observation is considered to be a lied group or size 1. Hence, if there are no ties at all,g = it, tj <= 1 fory = 1......... it, and the right-hand side of (9) reduces to (24)-1[( + l)(2n + 1)]. Exai data, b a dout was th< the X the Y' been d In th to low> that tl 45-3 We t DO 095190 CONFIDENT! Al. t the a (4) ')} = . ; a level (5) c a level (6) (7) | e normal (8) alues and mong the as in (4) lgc ranks, i) , (9) * ncc, if there >) reduces to A DISTRIBUTION-FREE SIGNED RANK TEST Table 1. Hamilton depression scale factor IV values Patient i 1 2 3 4 5 6 7 8 9 Xi 1.83 0.50 1.62 2.48 1.68 1.88 1.55 3.06 1.30 Yi 0.878 0.647 0.598 2.05 1.06 1.29 1.06 3.14 1.29 Source. D. S. Salsburg (1970). 29 Example 1. The data in Table 1 were obtained by Salsburg (1970). The data, based on nine patients who received tranquilizer T, were taken from a double-blind clinical trial involving two tranquilizers. The measure used was the Hamilton (1960) depression scale factor IV (the "suicidal" factor;) the X value was taken at the first patient visit after initiation of therapy, the Y value at the second visit after initiation of therapy. The patients had been diagnosed as having mixed anxiety and depression. In this example involving tranquilizers, an improvement would correspond to lower factor values. We will perform test (5). From Table A.4 we find that the critical value at the a = .049 level is (9(10)/2) -- t(.049, 9) = 45 -- 37 = 8. We illustrate the computations leading to (3) in tabular form. / ?t \z<\ R, V. R,v 1 -.952 .952 8 0 0 2 .147 .147 3 1 3 3 -1.022 1.022 9 0 0 4 -.430 .430 4 0 0 5 -.620 .620 7 0 0 6 -.590 .590 60 0 7 -.490 .490 50 0 8 .080 .080 2 1 2 9 -.010 .010 1 0 0 We then have 0 t+ = 2 R,vt = 5, 30 THE ONE-SAMPLE LOCATION PROBLEM and, since T+ = 5 < (9(10)/2) -- /(.049, 9) = 8, we reject //0 in favor of 0 < 0 at the = .049 level. Note that the lowest level at which we could reject H0 using this one-sided test is .020, since (9(10)/2) -- /(.020, 9) *= 45 -- 40 = 5. For the large sample approximation we find, using (7), 5 -- (9(10)/4) -17.5 T* =-------- --= ---------= -2.07. [9(10)(19)/24]/a 8.44 Thus we see from Table A.l that the lowest level at which we reject N0 withthe normal approximation is .0192. Hence both the exact test and the large sample approximation indicate that there is strong evidence that the tran quilizer does lead to an improvement as measured by the Hamilton scale factor IV values. Comments 1. Assumption A1 does not require that Xt and Yt be independent, and indeed in most applications they are dependent. 2. For paired replicates data, the symmetry part of Assumption A3 is often inherently satisfied. In particular, if X{ and Y( arise from populations differing only in location (i.e., the only treatment "effect" is a change in location), then e, = Zi -- 0 comes from a population that is symmetric about zero. This is also true under more general conditions. 3. To test the hypothesis 0, = 0o, where 0o is a specified nonzero number, obtain the modified observations Z[ = Z, -- 0O, for / = 1,... , n, and compute T+ using the Z"s (instead of the Z's). Procedures (4) to (6) may then be applied as described. 4. When Ha (2) is true, the distribution of 7*+ (3) is symmetric about its mean n(n + l)/4; this implies that Pb{T+ >*} = f>o{:r+ < "(" - zj, for x = 0,... , (n(n + l)/2). From this it follows that the lower a percentile for the null distribution of T+ is [n( + l)/2] -- l(tx,n). This explains the use of [n( + l)/2] -- /(a, n) as the critical value in procedure (5). 5. Whereas in order to perform the sign test (Section 4) for paired replicates we need only the signs of the Z, differences, in the case of the signed rank test Ave need to know both the signs of the Z's and the ranks of the |Z|'s. 6. The statistic T+ (3) is the sum of the ranks (of the absolute values) corresponding to the positive Z observations. Define T~ to be the sum of ranks (of the absolute values) corresponding to the negative Z observations. Note that T~ can be defined by (3) with y>, replaced by I -- y>,, for i = 1, . . . , n. Test procedures (4) to (6) could equivalently be based on T~, since T~ = [n(n + l)/2] - T+. (See Problem 3.) 7. Let the rank tend to 1 observat 0 > 0 f< SLi',.' 8. The 2<V,< possible (J)" Fo B 0 1 1 1 2 2 2 3 Thus, for (i), (Cf. ' The de lying e p and symr distributi constant; satisfied, subject to depend o 9. Repl comes fro independt where Zx independt alternativi //0 in favor of vliich wc could - /(.020, 9) = reject H0 with t and the large that the tranlamilton scale ependent, and imption A3 is in populations is a change in is symmetric izero number, and 4) to (6) may ctric about its or a. percentile s explains the (5). ired replicates e signed rank >f the |Z|'s. solute values) j sum of ranks vations. Note i-- 1 , since T~ = A distribution-free signed rank test 31 7. Let B be the number of positive Z's and let /y < r2 < < rn denote the ranks of the absolute values of these positive Z's. For 0 > 0, these ranks tend to be larger than the corresponding ranks of the absolute values of the observations that are negative. This suggests rejecting H0 (2) in favor of 0 > 0 for large values of r,. As mentioned in Comment 6, T+ = ri This is partial motivation for the procedure defined by (4). 8. The null distribution of T+ can be obtained using the equation T+ = ri (see Comment 7), recognizing that, under H,, (2), each of the 2" possible outcomes for the configuration (/y,... , rB) occurs with probability (J)". For example, in the case n = 3, the 2s = 8 possible outcomes for (i-j, . . . , rB) and associated values of T+ are given in the following table, a B (/-!, r2, . . . , rB) /'0{(/-1, r2, , . . , rB)} T+ = r. 0 1 rl *= 1 1^ = 2 1 r, = 3 2 rt = 1, r2 = 2 2 n = 1, r2 = 3 2 r2 = 2, rt = 3 3 rl = 1, r2 = 2, r3 -- 3 * i l i 4 \ 2 3 3 4 5 6 Thus, for instance, P0{T+ ^ 5} = P0{T+ = 5} + P0{T+ = 6} = (J) + (*) = (J). (Cf. Table A,4.) The derivation of the null distribution of T+ does not depend on the under lying e populations beyond the point of requiring them to be continuous and symmetric about zero. This is why the procedures based on T+ are called distribution-free procedures. For example, when we choose the critical constant t(a, n) of procedure (4) so that the equation /,0[7'+ ^ /(a, n)] = a is satisfied, we control the probability of (falsely) rejecting //,, when H0 is true; subject to Assumptions A1 to A3, moreover, this error probability does not depend on the underlying populations of the e's. 9. Replace Assumptions A1 to A3 by the Assumptions AI': each Z comes from the same continuous population and A2': the nZ's are mutually independent. Then letting 0* = P{Z, + Z2 > 0) - <i>, where Z, and Z2 are random members from the Z population and Zu Zz are independent, the test procedures (4), (5), and (6) are consistent against alternatives for which 0* >, <, and ^0, respectively [see Pratt (1959)]. THE ONE-SAMPLE LOCATION PROBLEM For the special situation covered by Assumptions A1 and A2 and the stronger version of A3 used in Property 1, this consistency statement implies the consistency statement given in Property 1. Properties 1. Consistency: For our consistency statement we strengthen Assumption A3 so that each e comes from the same continuous population that is sym metric about 0. Then the tests defined by (4), (5), and (6) will be consistent against the alternatives 0 >, <, and ^0, respectively. See Pitman (1948), van Eeden and Benard (1957b), Pratt (1959), and Comment 9. 2. Efficiency: See Pitman (1948), Hodges and Lehmann (1956), and Section 10. references. The T+ statistic was introduced by Wilcoxon (1945). Fraser (1957b) proposed a normal scores analog of the signed rank test that has desirable efficiency properties, Pratt (1959) presented a detailed discussion of methods for handling zero observations and tied ranks in the signed rank procedure [see also Cureton (1967)]. Hollander (1971) considered a procedure for testing the absence of a treatment effect under a more general model than that of (1) (see Section 10.3). Other competitors of the T+ test include a binary test due to Savage (1959), a generalization of the binary test considered by Doksum and Thompson (1971), and various classes of tests considered by Hemelrijk (1950a, 1950b), van Eeden and Benard (1957a, 1957b, 1957c), Govindarajulu (1960), Gross (1966), and Doksum and Thompson (1971). See also References of Section 4. PROBLEMS The data in Table 2 are a subset of the data obtained by Kaneto, Kosaka, and Nakao Table 2. Data on blood levels of immunoreactive insulin (\iUjmI) Dog/ xt y. 1 350 480 2 200 130 3 240 250 4 290 310 5 90 280 6 370 1450 7 240 280 Source. A. Kaneto, K, Kosaka, and K. Nakao (1967). (1967) ai stimulati weights, just befo stimulatii stimulati' '2. Cha observati about the in which rejection 3. Let or illustra 2. AN Procct 1. For 2. The (see Con Let W if M is O' If M is (A graf ExampI values of -.821, -.590, -.4375, -.2365, .0685, .08 k = 22. T CHAPTER 3 The One-Sample Location Problem The procedures of this chapter are designed for statistical analyses in which primary interest is centered on the location (median) of a population. We encounter two types of data for which such analyses are important. The first of these, referred to as paired replicates data, represents pairs of "pretreatment" and "posttreatment" observations; here we are concerned with a shift in location due to the application of the "treatment." The second type of data, referred to as one-sample data, consists of observations from a single population about whose location we wish to make inferences. In Sections 1 to 3, procedures are considered for analyzing paired replicates data using signed ranks. In particular. Section 1 presents a distribution-free signed rank test, Section 2 a point estimator associated with the signed rank statistic, and Section 3 a related distribution-free confidence interval. In Section 7 these procedures are applied to some one-sample data. An asymptotically distribution-free test for symmetry of the underlying population (one of the assumptions in Sections 1 to 3 and 1) is considered in Section . Procedures for analyzing paired replicates data using signs are discussed in Sections 4 to 6. A distribution-free sign test is considered in Section 4, a point estimator associated with the sign statistic in Section 5, and a related distribution-free confidence interval in Section 6. These sign procedures are applied to some one-sample data in Section 8. The asymptotic relative efficiencies for translation alternatives of the procedures based on the signed rank statistic and those based on the sign statistic with respect to their normal theory counterparts based on the sample mean are discussed in Section 10. In this section we also deal with the asymp totic relative efficiency of the test for symmetry covered in Section 9. DATA. (blocks, Assurr Al. W where th 0 is the i A2. T A3. & one) tha 1. A D Procei 1 1 in Of X/ l: '/ I / 3. Fo The pre value zt The sta Author Coulter Todd & Sanford Merck Conn Jandl SKF NIOSH Bio Science APPENDIX i SUMMARY OF REFERENCE VALUES FOR ADULT MALES BLOOD COUNTS Hemoglobin (Gm %) 14-18 13.5-18 13-16 14-18 13.5-17.5 14-18 13-18 00 Hematocrit (%) 42-52 40-54 42-50 40-54 41-49 38-54 40-54 Erythrocytes (Million/mm ) 4.7-6.1 4.6-6.2 4.2-5.9 4.6-6.2 4.4-6.0 4.5-6.0 4.4-6.3 4.5-6.0 Leukocytes (Thousand/mm 4.8-10.8 4.5-10 4.8-10.8 4.5-11 5-10 4.2-10 N/A EFFECTS WHICH CAN BE MEASURED USING EXISTING DATA Expected Effect (Dependent Variable) How Measured 1. Hematopoietic Anemia lymphocytosis or lymphocytopenia Red and white blood cell counts 2. Hepatotoxicity 3. Mutagenesis A. Renal tubule toxicity Liver function tests, liver enzymes "Adverse pregnancy outcomes" according to questionnaire Kidney function tests Why Direct toxic effects on red and white cell pre cursors: stimulation of the lymphocytic system Direct toxic effect Alteration of germinal stem cells Direct toxic effect TABLE II EFFECTS WHICH MIGHT OCCUR IF EO EXPOSURE WERE TO CAUSE ADVERSE LONG TERM EFFECTS ON HUMAN HEALTH, BUT WHICH CANNOT BE MEASURED USING EXISTING DATA AND WILL NOT BE ANALYSED IN THIS STUDY Expected Effect (Dependent Variable) 1. Hematopoietic Anemia 2. Cytogenetic 3. Spermatocidal Depression of spermatogenesis A. Immunotoxicity 5. Neurotoxicity How Measured Reticulocyte count (Chromosome aberrations, sister chromatid ex changes) Sperm counts FSH levels Counts of T&B cells Mitogen stimulation Depressed antibody formation Measures of nerve conduction Tests for tremor Why Hemolysis A Direct effect on DNA in dividing cells Toxicity to germinal tissue DNA effect; direct toxic effect on bone marrow and blood Reports of neurological effects Abbreviations: DNA = deoxynucleic acid; FSH = follicle stimulating hormone; T&B cells = thymus and bursa derived cells. DO 095198 CONFTDFNTTAI X i TABLE I. PARTICIPATION DATA GROUP NUMBER I. Population Universe A. All employees eligible for Periodic Health Examinations (PHE) in 1980 (Total in plant except administrative and some management). B. Total for whom 1980 PHE data available.* C. Female employees. D. Population universe used as source for exposed and control populations (= IB minus IC). II. Exposed population, selected from source population (all males). A. No. males in Glycol II at time of , 1980 exam. B. No. males who worked in Glycol I & II on or before 31 Dec. 198O. C. No. males in Dowanol/Ethanolamine at time of 1980 exam. D. No. males in Dowanol/Ethanolamine on or before 31 Dec. 1980. E. Subtotal In expo.sed population = IIA + IIC PARTICIPATION RATIO DVP:dll V21/81 TABLE 1 (Con't) Page 2 GROUP NUMBER ^PARTICIPATION RATIO III. Control population (all males). A. Source population (ID) minus exposed population (IIA + IIB + IIC + IID). B. Employees with potential exposure to benzene in 1980. C. Employeees with unknown chemical exposures due to assignment in Research and Development or in Instrumentation. t D. Subtotal from whom matched controls selected = III A minus IIIB and IIIC. 976 ^Participation Ratio = No. employees in the group for whom a 1980 PHE available divided by no. employees in the group eligible for 1980 PHE. DVP:dll 4/21/81 D0 CONFTDFNVtAl TABLE II. CHARACTERISTICS OF POPULATIONS FROM LOUISIANA DIVISION: EO EXPOSED, EO MATCHED~CONTROLS (1TTTT AND ALL PHE PARTICIPANTS. CHARACTERISTIC I. MATCHING CHARACTERISTICS. ^.Number Mean Age Median Age Number non whites (/`Yrs. in Plant (mean) % Current Smokers % Non Smokers yS % Ex Smokers & Pipe Smokers yS % Non Drinkers Exposed II.OTHER CHARACTERISTICS. . with Hx. of diabetes * % with Hx. of asthma Mean Height Mean Weight Mean diastolic blood pressure #From whom a 1980 PHE available DVP:dll *1/21/81 VALUE Control ALL n0 noK/p TnfrNTTA| TABLE I. PARTICIPATION DATA GROUP NUMBER I. Population Universe A. All employees eligible for Periodic Health Examinations (PHE) in 1980 (Total in plant except administrative and some management). B. Total for whom 1980 PHE data available. C. Female employees. D. Population universe used as source for exposed and control populations (= IB minus IC). II. Exposed population, selected from source population (all males). A. No. males in Glycol II at time of . 1980 exam. B. No. males who worked in Glycol I & II on or before 31 Dec. 1980. C. No. males in Dowanol/Ethanolamine at time of 1980 exam. D. No. males in Dowanol/Ethanolamine on or before 31 Dec. 1980. E. Subtotal in exposed population = IIA + IIC. 85 *PARTICIPATION RATIO DVP:dll 4/21/81 DO 095?0? CONFTDFNTTAI TABLE 1 (Don't) Page 2 GROUP NUMBER PARTICIPATION RATIO III. Control population (all males). A. Source population (ID) minus exposed population (IIA + IIB + IIC + IID). B. Employees with potential exposure to benzene in 1980. C. Employeees with unknown chemical exposures due to assignment in Research and Development or in Instrumentation. D. Subtotal from whom matched controls selected = III A minus IIIB and IIIC. 976 Participation Ratio = No. employees in the group for whom a 1980 PHE available divided by no. employees in the group eligible for 1980 PHE. DVP:dll V21/81 00 00Nf 095?03 TOFi^i Tai TABLE II. CHARACTERISTICS OF POPULATIONS FROM LOUISIANA DIVISION: EO "EXPOSED, EO MATCHED CONTROLS~TiTTTT AND ALL PHE PARTICIPANTS. ' CHARACTERISTIC I. MATCHING CHARACTERISTICS. /^.Number y^Mean Age r Median Age v* Number non whites */'Yrs. in Plant (mean) % Current Smokers % Non Smokers ,/ % Ex Smokers & Pipe Smokers % Non Drinkers Exposed II.OTHER CHARACTERISTICS. Vy% with Hx. of diabetes * % with Hx. of asthma Mean Height Mean Weight Mean diastolic blood pressure #From whom a 1980 PHE available DVP:dll V21/81 VALUE Control ALL DO 095P04 OONFTDFNTTAI People. Estimated personnel needs are as follows: Individual______________________________________ Time (Hours) VanPeenen - Epidemiologist, Midland 120 Currier - Medical Director, Louisiana 80 Cook - Director of Epidemiology, U.S. Area 20 Poston - Computer Specialist, Louisiana 60 Ledford - Industrial Hygienist, Louisiana 40 TOTAL 320Hours Dollars: For personnel, 320 x $55.00/hr = For computer time, estimate about $17,600.00 4,000.00 The computer estimate is for analyses only and does not include gathering and storing health data; It is based on a one month's cost incurred in our department during a period of heavy computer usage. For travel and miscellaneous 3,000.00 TOTAL.................................................................... $24,600.00 TIME TO COMPLETION: Approximately 3 months from approval of protocol to final report. DRAFT PROTOCOL Ethylene Oxide Study/VanPeenen, et al EPIDEMIOLOGY HEALTH AND environmental SCIENCES DOW CHEMICAL U.'S.A. PROTOCOL STATUSi preliminary-draft 'INAL DRAFT TITLE: CROSS-SECTIONAL STUDY OF EMPLOYEES WITH WORKPLACE EXPOSURES TO ETHYLENE OXIDE AT LOUISIANA DIVISION PROTOCOL NUMBER: epi-dr-9999-9999-2181 DA.TE: jg-iUary - 2, 1981 PROPOSED STARTING DATE: February 1, 1981 ESTIMATED DA.TE OF FINAL REPORT: june lj 1981 SIGNATURES: Prisary Investigator: u~4vu- P. F. D. VanPeenen, M.D. Approvals: R. R. Cook, M.D, (Dare) iin (Dare) J. R. Venable, M.D. (Date) R. L. Dostal (Date) 00 095?06 t0FNTT Al CONP * J"v*^*jS' /^ ' * EPIDEMIOLOGY HEALTH AND ENVIRONMENTAL SCIENCES DOW CHEMICAL U.S.A. r '^ PROTOCOL STATUS: Sr**--' PRELIMINARY-DRAFT FINAL 12 DRAFT TITLE: CROSS-SECTIONAL STUDY OF EMPLOYEES WITH WORKPLACE EXPOSURES TO ETHYLENE OXIDE AT LOUISIANA DIVISION PROTOCOL NUMBER: EPI-DR-0114-1975-2181 DATE: February 3, 1981 PROPOSED STARTING DATE: February 1, 1981 ESTIMATED DATE OF FINAL REPORT: june i 1981 SIGNATURES: Primary Investigator: Approvals: Distribution: / J* R. Venable, M.D. R. L. Dostal M. F. Currier, M.D. (Date) (Date) (Date) HO 095707 CONFTDFNTTA! 1,2th DRAFT February 3, 1981 TITLE: SHORT TITLE: CROSS-SECTIONAL STUDY OF EMPLOYEES WITH WORKPLACE EXPOSURES TO ETHYLENE OXIDE AT LOUISIANA DIVISION Ethylene Oxide Study PRIMARY INVESTIGATOR: P. F. D. VanPeenen, M.D. Epidemiologist Michigan Division Midland, Michigan Telephone: (517)636-4769 CO-INVESTIGATORS: J. D. Burek Veterinary Pathologist 1803 Building Midland, Michigan Telephone: (517)636-0737 R. R. Cook, M.D. Director of Epidemiology U.S. Area Medical 1803 Building Dow Chemical U.S.A. Midland, Michigan Telephone: (517)636-1383 M. F, Currier, M.D. Medical Director Plaquemine, Louisiana Telephone: (504)389-8292 William E. Ledford Industrial Hygienist Plaquemine, Louisiana Telephone: (504)389-1659 Perry Poston _ Computer Specialist Plaquemine, Louisiana Telephone: (504)389-8310 DO 09570ft CONFTDFNT TAI OBJECTIVES: To determine whether Dow employees who are exposed to ethylene oxide (EO) have more abnormalities of the hematopoietic, hepatic, renal and reproductive systems, as measured by data collected for periodic health examinations, than do employees without-such ex posure. IMPORTANCE OF THE CHEMICAL: Ethylene oxide (EO) is an important industrial chemical, produced from feedstock ethylene, which is used primarily for making ethylene glycol or polyesters (Anon, 1979a). It is significant from a health viewpoint that EO is also used for cold sterilization of medical equipment, although this usage is minimal from a commercial viewpoint. U.S. production of EO increased from 1.3 billion pounds in 1960 to 5.3 billion pounds in 1979 (Anon, 1979a). Dow has produced EO since 1941 at the Texas Division and since 1959 at the Louisiana Division. At these locations, McClimans et al, (unpublished manuscript) estimated that at least 550 employees ex perienced some workplace exposures to EO in the past, while Moore et al, (1979) estimated that 300 employees had worked in EO production for one year or more. At the present time, there are 88 employees with potential exposure to EO at the Louisiana facility (44 each in Glycol II and Dowanols ). DRAFT PROTOCOL Ethylene Oxide Study/VanPeenen, e_t a_l 1 DO 09B709 CONF'TDFNTI Al The American Conference of Governmental Industrial Hygienists (AGGIH) lists an 8 hour time-weighted-average concentration (TWA) of 10 ppm for E0 in its Notice of Intended Changes for 1980; an 8 hour 10 ppm TWA was also the Dow IHG, but this has very recently (1980) changed to 3 ppm with excursion limit of 50 ppm. The current OSHA permissible exposure limit (PEL) is 50 ppm TWA. REVIEW OF LITERATURE: Acute intoxication with EO at high concentra tions has long been recognized as a cause of Irritation of eyes, respiratory tract and skin, and central nervous system (CNS) depres sion. Delayed effects include headache, nausea, vomiting, inco ordination, and electrocardiograph abnormalities (Proctor & Hughes, 1978). Peripheral neuropathy with decreased nerve conduction velocity has recently been reported in 4 men after overexposure to EO (Gross et al, 1979); whether cause and effect can be assumed in these cases is debatable. Levels responsible for acute effects have been esti mated at above 2000 ppm for very short single exposures and above 200 ppm for continuous 8 hour exposures (Hine and Rowe, 1963). According to Dr. Ray Flake of the Texas Division, acute overexposure of humans to EO may cause an increase in serum SGOT and SGPT levels (Fishbeck, personal communication). DRAFT PROTOCOL Ethylene Oxide Study/VanPeenen, e al - 2. DO 095?10 OONFTDFNTTAI The experimental toxicity of EO was reported by Dow toxicolo gists more than 20 years ago (Hollingsworth nt jal, 1956). There were no effects at 49 ppm, but histologic changes in liver, kidney, and testes were observed in some animal species at repeated sevenhour vapor exposures at levels of 204 and 357 ppm. Reversible impair ment of nervous function and muscular paralysis were observed in rats, rabbits and monkeys at the 204 ppm level dose. Interestingly, laboratory results, including blood urea nitrogen and blood counts, were normal in the species studied even at the higher (357 ppm) exposure levels (Hollingsworth et_ al, 1956) . Recently, there has been renewed concern about possible chronic toxicity of EO. Results of an industry-sponsored 24-month experimental inhalation study showed an increased, dose-related incidence of mono nuclear cell leukemia in male and female Fischer 344 rats (Snelling e_t al, 1981). Effects were voted at all level of exposures: 100, 33, and 10 ppm. Past literature in these regards included a 1964 report of tumorigenicity in female mice'(Reyniers, 1964), alkylating agent pro perties (Ehrenberg et al, 1974), and demonstrations of mutagenicity in several test systems (Anon, 1979b), including Neurospora (Watson, 1966), plants (Ehrenberg and Gustafsson, 1957), and the Salmonella typhimurium strain TA 1535 bacterial assay (Embree, 1975). DRAFT PROTOCOL Ethylene Oxide Study/VanPeenen, e_t al 3. 00 7 7DfNT 7A( As regards human disease, there are two recent published reports from Sweden of malignancies purportedly linked to EO exposure. The first report was of a case each of acute mylegenous leukemia, chronic myelogenous leukemia, and Waldenstrom's macroglobulinemia diagnosed between 1972 and 1977 in employees of a small factory which'used EO for sterilization (Hogstedt at jQ, 1979a). These cancers may, of course, be unrelated; but in a subsequent retrospective mortality study of employees of an EO production facility, the same senior author reported more observed deaths from leukemia (3) than would have been expected (less than 0.5) using Swedish national death rates (Hogstedt jet al, 1979b). Two of the leukemia deaths were caused by chronic lymphatic leukemia, one by acute myeloid leukemia. It is difficult to interpret results from this paper because of the finding of an elevated overall standard mortality ratio (SMR) with 47 deaths vs. 27 expected. Such a finding would be extremely unusual in the U.S. where most U.S. employed populations reflect the "healthy worker effect", and experience overall SMRs less than 100%. Another unusual finding was an increased SMR for diseases of the circulatory system (20 observed vs. 13 expected), which is reported only in office employees in the U.S. Finally, chronic lymphatic leukemia has not, to this author's knowledge, been linked to environmental factors. A retrospective cohort mortality study of employees with work place exposure to EO is planned by British colleagues at Imperial Chemical Industries Limited (ICI). Results of the 1CI study will be helpful in interpretating the Hogstedt reports. DRAFT PROTOCOL Ethylene Oxide Study/VanPeenen, jet sQ 4. DO 09571? CONFTDFNTTAI The Russian literature includes two reports of a variety of health problems associated with workplace exposures to EO. Yabukova et al (1976) reported an increased gynecological sick rate, increased complications of pregnancy, and increased incidence of interruptions of pregnancy among employees of an EO manufacturing plant compared to pregnancies in management personnel and from outside the factory. However, if I read a translation of the paper correctly, although spontaneous abortion occurred in 10.5% of pregnancies of the exposed group of instrument controllers compared to about 8% of laboratory workers (also exposed) and about 8% of management controls, there were more stillbirths among management employees (4/65) than in instrument controllers (1/57). According to Yabukova et al (1976) the "overall corresponding contamination of the factory air did not exceed 1 mg/m 3" (= about 0.6 ppm). Ostravskaya ^t ^1 (1971) reported a variety of subjective com plaints from 76 individuals with 3-6 years work experience, presumably in EO manufacturing. In addition, they reported "bradycardia", "vegeta tive shifts" and "EKG changes" in some; the significance, if any, of the quoted abnormalities is difficult to assess from the translation. 3 Ostravskaya ^t 1^ (1971) reported a leukopenia (4000-4400 WBC/mm ) in 9 of the 76 employees. DRAFT PROTOCOL Ethylene Oxide Study/VanPeenen, et al 5. DO 095313 OONFTDFNTTAl There was only one U.S. epidemiology study of EO prior to the recent interest in carcinogenesis. A cross-sectional study of 37 EO operators engaged in manufacturing operations and of age-matched controls (operators working with chemicals other than EO) was reported by Joyner in 1964. The highest air level of EO recorded was 127 ppm but average exposures were less than 10 ppm for the working day (Joyner, 1964). The EO group actually had fewer symptoms and less absenteeism than the controls; and there were no more "abnormal" laboratory findings in the study group than among the controls (Joyner, 1964). More recently, in a Dow Chemical U.S.A. response document (Moore et al, 1978), it is stated (page 7) that health surveillance data from 129 Dow employees "revealed no unusual medical effects, i.e., no effects were observed in employees potentially exposed to EO within the TLV". Tests apparently included blood counts and a battery of clinical labora tory tests. A cross-sectional study of individuals involved in cold sterilization using EO was recently published by Garry et al, 1979, who found various frequencies of symptoms (up to 9 with histories of sore throat) among 12 EO-exposed employees versus no symptoms of any kind in a control group. Similarly, cytogenetic studies by these authors reportedly showed more frequent sister chromatid exchanges (SCE) (average 8.65) among the 12 employees than among 11 controls (average 6.37). DRAFT PROTOCOL Ethylene Oxide Study/VanPeenen, et al 6. DO 095714 OONFTDFNT TAl A cross-sectional study was also sponsored by the American Hospital Supply Corporation (AHSC). As yet unpublished, the AHSC study was of employees of 9 U.S. manufacturing facilities which utilize EO as a sterilant gas. According to data kindly provided by Ronald H. Abrahams, Ph.D., director of Regulatory Compliance-of * the AHSC, 6 of 75 exposed employees had slight anemia and one had a slight increase in lymphocytes (there were apparently no controls for this phase of the study). Sperm counts were obtained from some of the employees: 8, of 46 exposed and 1 of 9 controls had sperm counts less than 20 million/cc. The AHSC study also included cytogenetic data: except for stable forms (Cs), all modalities of chromosome aberration including number of breaks, unstable forms (CV), and SCEs, were statis tically increased in the exposed group when compared to control groups of office workers. RATIONALE FOR RESEARCH: There is a growing volume of experimental toxicological data indicating that EO, at chronic vapor exposure levels much higher than experienced in the Dow workplace, can cause detectable health effects. Whether measurable effects might also occur at lower levels of exposure is unknown, but, if so, they might include those listed in Tables I and II. A major concern about long-term effects of workplace exposures to EO is carcinogenesis, so & Industrial Hygiene data were not given, but vapor levels were stated to be less than the OSHA 8 hour TWA limit of 50 ppm. DRAFT PROTOCOL Ethylene Oxide Study/VanPeenen, et al 7. DO 095710 CONFTDFNTTAI it would seem reasonable that any epidemiological study be designed not only to detect the phenomena listed in Table I over the short term, but to detect excess cancers over the long term. In fact, a draft proposal for a retrospective cohort mortality study of EO pro duction personnel at the Texas Division, with leukemia as a primary concern, was prepared by McClimans and colleagues in July 1979 (un published data). Very recently, Bond and colleagues have prepared the protocol for a case-control study of leukemias at Texas Division. Our proposal is for a cross-sectional study of Louisiana Division employees involved in jobs with potential EO exposure, to be followed by continuous surveillance of the same group of employees. Since cross-sectional studies are preliminary and do not usually yield definitive results, it will be necessary to plan for appropriate follow-on research depending on these preliminary findings. Alternatives for follow-on research are detailed on page 11. Abnormalities which might be expected if workplace exposures to EO were exerting a chronic toxic effect include hematological, hepatic, reproductive, renal, cytogenetic, and carcinogenic (Glaser, 1977). As proposed at the present time, however, our project will be limited to analysis of results from existing medical surveillance data. Hemato logical abnormalities reportedly associated with exposure to EO include anemia and absolute lymphocytosis (Ehrenberg and Halstrom, 1967), leucopenia (Ostravskaya t al, 1971) or, simply "changes in peripheral blood" (Spasovsky e_t al, 1978). Because of the reported dose-related DRAFT PROTOCOL Ethylene Oxide Study/VanPeenen, e jil 8. DO 095715 CONFTDFNTTAl alkylation of hemoglobin by EO (Osterman-Golkar e a]., 1976; Calleman et al, 1978) it would not be unreasonable to expect increases in laboratory tests associated with hemolysis, including total serum biliru bin and lactic acid dehydrogenase (LDH), as well as anemia and reticulocytosis. Similarly, direct hepatic and renal tubular damage, causing abnormal results in simple measures of liver and kidney function, might be expected from any alkylating agent and, in fact, have been reported experimentally at high levels of EO vapor exposure (Hollingsworth et al, 1956). Questionnaire data from the periodic health examinations are available and can be used. We recognize the weaknesses and biases of using questionnaire data as evidence for disease or for adverse repro ductive function, but by judicious and ji priori designation of which questions will be analyzed it should be possible to compare the pre valences of symptoms relating to diseases of the blood and of adverse pregnancy outcomes. The list of effects which can be studied using existing data is in Table I. Table II is a more comprehensive list, including tests which cannot, and, in certain cases, probably should not, be studied at the present time. As more reliable techniques become available, however, it may become appropriate to examine for additional effects in the future. Table III lists data elements which are presently available for extraction and analysis. DRAFT PROTOCOL Ehtylene Oxide Study/VanPeenen, et: al 9. DO DO NF 95P77 TFNT TAl APPROACH: Exposed Population. The exposed population vill consist of active Louisiana Division employees who were assigned to Glycol II or Dowanols for at least 6 months prior to participating in the Health Inventory examination during 1980. Since intensity of exposure cate gories by job cannot be estimated because detailed historical industrial hygiene data are lacking, length of employment in the 2 plants will be used to divide the exposed group into subsets of longer and shorter dura tions of exposure. Control Population. There are 88 persons eligible; but since it is anticipated that non-wh,ites and females will be few, only data from white males will be included in the study. The control group will consist of an equal number of males matched by age (within 5 years or better), race, date of hire, and smoking and alcohol use history. The job category will be as of the date of health inventory and, of course, results of the 1980 health inventory examination must be available. Finally, to be eligible as a control, the employee must have no history of EO exposure. It is recommended that the initial pool for controls con sist of all Louisiana Division employees, and that eliminations be made by matching criteria, the best age match being the final decision step. DRAFT PROTOCOL Ethylene Oxide Study/VanPeenen, e_t al 10. DO 095?18 CONFTDFNTTA npppnripnt Variables. Dependent variables are listed in Table III, and consist of answers to questions regarding reproductive history and past medical history, complete blood count values, and results of the liver and kidney function tests. Test results will be considered abnormal if outside the reference ranges shown; a summary of some re cently published ranges is in Appendix i. Covariables. Covariables or potential confounding variables other than those used in matching include family history of illness, use of medications, exposure to chemicals other than EO, radiation exposure, and substance abuse. The effects of some of these covariables cannot be analysed, and others, such as known substance abuse or use of certain medications might best be handled by elimination from both control and exposed groups. Those with a family history of blood disorders need not. be eliminated, but the frequencies of such histories in the 2 groups should be checked. Obviously, affirmative answers to questions such as "Did you have a birth abnormality?" should be considered if the abnormality were considered hereditary. Analysis of Data. Data will be collated and analysed using the com puter system available to the Louisiana Division Medical Department and capably programmed by Perry Poston. The SAS statistical software system will be used for analyses. The null hypothesis of no statis tically significant difference in values from the exposed and control samples will be by the Wilcoxon signed rank test for continuous data and McNemar's test in the case of dichotomous data. Continuous data will also be analyzed-by McNemar's test using the categories of within or without the normal range as defined in Table III. DRAFT PROTOCOL Ethylene Oxide Study/VanPeenen, et al 11. RESULTS. WEAKNESSES, AND FOLLOW-ON STUDIES: This proposed crosssectional study should indicate whether or not employees potentially exposed to EO have relatively depressed red cell indices or depressed or elevated white cell counts compared to their non-EO-exposed col leagues; whether results of their laboratory tests of renal and hepatic function are comparable; and whether their reported prevalences of adverse pregnancy outcomes are similar. Identification of potentially exposed groups will also permit continued future surveillance and such groups would also be available for testing as new techniques become available. Concern has been expressed about ability of this study to detect leukemia. By definition, leukemia would, in fact, be detected by the combination of anemia associated with increased number of white cells in peripheral blood, and the diagnosis is easily confirmed by microscopic examination of bone marrow. Cross-sectional studies have many weaknessess which need not be detailed here. The proposed study, however, additionally suffers from the following: DRAFT PROTOCOL Ethylene Oxide Study/VanPeenen, et al 12. D0 0^7?0 1. Pregnancy outcome data from fathers are less reliable than from mothers. 2. The questionnaires permit bias towards "socially-acceptable" answers. 3. The "power" or ability to detect subtle differences between exposed and control populations may be low. 4. It is difficult to separate out age and work-ethic effects when exposure categories are defined by duration rather than intensity. For these reasons, we suggest that a negative result (i.e., no significant difference in dependent variables between exposed and con trol groups) not be considered the final word. Thus, it will be appro priate to continue to monitor employees with potential EO exposures. Populations to be included in continuing surveillance (pros pective morbidity) studies will be those identified for this crosssectional study, with addition of those entering or leaving jobs with potential EO exposures. Controls who enter such jobs vTill join the exposed group and will be replaced as controls using the original match ing criteria. Personnel new to EO exposure jobs, and staying there for 6 months, will be added to the study along with new matched controls. Personnel leaving EO exposure jobs will continue in the study, but, of course, in a different category. DRAFT PROTOCOL Ethylene Oxide Study/VanPeenen, et al 13. DO 095??1 OONFTOFNTTAl Similarily a positive result (i.e., statistically significant difference in dependent variables between exposed and control groups) cannot be considered definitive either. In order to avoid decisions based on chance due to multiple comparisons, it is suggested that future studies at this time be stimulated only if a positive result is found for more than 2 laboratory tests (of the 15 to be analysed) or for more than 2 questionnaire responses (of the 11 analysed). In such an event, a retrospective morbidity study of a cohort of all EO-exposed employees at some point in the past (I'd recommend a period prior to the first Health Inventory) should be designed. Such a study might permit evalua tion of the question of "which came first, exposure or morbidity?" DRAFT PROTOCOL Ethylene Oxide Study/VanPeenen, et al 14. Do CO/\/p- r0Fr^TrAi *T AAfc /V0&XA4U, &SiZ&uz/77K2A/ /<*<? *ma. S&&OSMXto*s'7a<> &?as Z^ata/t^ Afr*Z> d*4&&a<2x&&<s7ya& r/etv/A/siZi, 7&& /use/' ^/<5if w 4>Af77`S*'4Sa4A$' 2*772? ^ ^/&***& ^w A&S9?*?7BA04,/ {MZ3 r 7S7~, TZ<27 4*SC f. #777? /3S^S^r X-4//%A>/A<P&*A7&S{77>XZZUl. 7?7*<A?a**/& AZAtasAT/ATZSTT, *2*?, SiG^r, SO&T, Aas2> ste/vzw^c?^ 7Vs& &#za &&&&*&* z&e *??uje*y Ca*7a7 **/> <5<Fs&/s?? d^*r?As/A/jj4/4t&s&#jzajexnP& ,, /?#&- &e7&Q#4Wfta. pf AA7>Sxj^/ls yy/> 0fr &/Stt/SiJrs0A^^^ S40?SVS* s&&*& azm%&as?& 7//e*A? Attunes*/, 2, A*7%> ts/SC &>& ??7 Azovas: 7V7> A&e09A>7~ a&g * eo-exj 9 ***srAb pa A#/jK&e/6 MO^ijSSj 777tf M&e# A77777S/2A& 27/2 /tt#r Az?/M*sz>&e& A&/* A<wSw/, 4'TAW/*2' A^'Vaf%^*5&'*'*acmtw!w/i" ^4$%54si ^ /asje^ a2g*-au/AU&&s 7* /ZuprA*'. /&a/bt/y^j., &*&&*?, k&a&vo& xttefjjsrzzr /t&eg &C7tG#<%^z> /rWW x/"A*ee /A*W*2S>Ws<(&eZw<l,j <w2t#?r7/y? Ay>7r/wA^5 ArvOrvezT/A ;s S,,.W. &A ,**<> OpaTtG?^ 7&7%/a2>71s*aas j A77?<u> p*7 ^q<> ^ /<^:r /sAa//?>*a?> 7z/f ST277pr?*t4> r*TM, A----7--k--#---t-e---&-----/-&- -`--<---s--y--!-r-r--z--y--?--? t fs#-sj S' ?- $sz*CY /vt^ ;?^r z&t/'sysWs&^ArrS*?S /i4^ag^^ sj&es&A da*&W7&& ^j^g/r^rwr . Anfrrsrrsjrtrfry'. - &^#,/,/ --2.0. 7- _A____A__*_&*&__4_ /7- _aA* tt**? **& ^,T1/#^^%Zg*J5^'7'^?P,1 *-/** **ver%>A*lKS^meiLe, 7** SAfCCJ/'/Y^ yCSWi^r At4G /S^ia^ AA/7& J^sr SJj*ssiS^}jOs*\ w ^^/-yy^sfTar-r 'SZVVSZK* ^e^xs^r ^ ----- '^ -- -- - --- - yjjjg sia0Xx-~A/~X*-&eSA *a-wrrz*->' 7#e #&a*hTM*u-> -A>' /tmaTt-&---a--sr*'s7 y '?=~&vs'rer "snt&weai <_'vl Zr'rws Afr/A/r ^x*r ysvu*2> 72 Jv~ Oat/OS 77/7/74AA. AtyAPX*fP&*4 7X447- XU7 yiVAtfr ^77) Xf^A ' x?rXA/2y)j%? ^4^/3^ APs??-A A'P&WA*-' /> 2232T 2A ,S7&iV-j!4'4jr At^ 4&X JZutr, aas>> Ss/^4/auu^ 7&ir Asatm*^ AZeo&&8s4sAY A^A A4R4 rs6su4&&, &&*#: #A'&e#AA& Aoe A A'pAa^a^ S&euSSae&AsAr S2u?r~ a*)ua^> A&&z*x~y 7PXAP&7 A S7XAS4*/r~ A7AA? *** S**#, 4V</ %>*&&& J/AAO&S 7X& /#ie?7C4/*4?se 3*4>Z2 &*V??A/7~ A&7*t? M**Z> *'s&usb>, A*rb stf/yw/AjL 7ZUA>o;-*sA! UtS^e# A&G*2*&t%>Ab 7# 2>7?A7VSAAC 774^ /d7AS4>A7 AffMAV^ 774^ XXe&?J*f fAtfU/AZ'. TtfAsAS, 7&A* AKA&ZKAZA 0*&AA~ ^ *sfrS*7&e 77/ 3SA>Zf, ^/sn*e zzva /vsxr steoae&uBe: castw tm/a*4<s/s, #&?- #X#S &#AA&aCA& /&&> 72 4S1/& s?*-~ 4 tee> ts J&r ys&v 0A 7Z& PAT# AW*> <43*^ /t&HSA&rJ'S 7&A7~ AM#*/?-AO& *7&75? Aa/J/ 70s\WA4?S$ /77*L4AA.y A&&&&&&%>. PetTAtSA AZ&?*?? ^XW AZ&& AZxAj / ^SQAS77AS4A04C& 2/9770 S7&A&S /&CA&XX? 77* 774? A/&/*S7?e. Me//A4*4 ASZ/AsP Aa/X>*^s<?>> sS 7&&4/77!ASP:`J> S*7 777^ yA&AgUA * ~ -^fi'SSSJV' ^ /3S7/7-* //TAk&S. AAAy PA ZW ATG&^ax,<0O?AU& * * ATCTTPaA 7$ SPAs77?Ok2et7'st** S&e P?e^A77A7*A/P) # DO 095??4 CONFTOFNTIAl ^4^-r #sme& pwstx? ,7?r? az>s77&a7p aap. **7-*pa, +* 4*4 TJS&e* 4M7P4, *??? APT??*#/*, Ate. Az7 +*+ ZZ*> Z>7&7>, 37Z>??S7?e/A*/, ^/7A**F *//*#&&/7? PdtA&7 p /?&2>//a ArzepsAMr ***/' ^^eKa/ /s/z>s&#7's m*&&&*. zhszz?A!y?xy no 09$??s OONF TDFNT T A PROCFnuPE FOR ASCERTAINING NORMAL I TV OF POPULATIONS T:1' TUESOAV, JULY ZB, l<>81 ! J procedure for ascertaining normality of populations 7:ts TUESDAY, JULY 2R, IRfll 2 I l -I IN J U i j n o j oa 3> u opnrEOijPF FpP ASrFRTAJNING NORMALITY OF POPULATIONS 7rt3 TUESDAY, JULY 28* 1981 3 S 8 ft Y U N J O I dNOO ^ ^ 9 6 0 on 7 ) rr-rr 1* * * 1 DO 0 9 F,??& OONFTDFNTTA I i |- VAR 1 ABLE =WRCDIF !,, N mean -'STD DEV SKEWNFSS * uss U cv 14 T:MEAN=0 0:NORMAL i* KOMFNTS 84 -0.1702 38 2. 138 73 -0.2(n007 382.09 -1256.32 -0.729526 0.0795516 SUH WGTS SUM VARIANCE KIJRTOSIS CSS STD MEAN PROB>lTl PR0B5D stem leaf 53 4 A3 3 _3_133___________ 2 578 2 004 1 6677788999 1 01112344 0 6678889 0 344_______ -0 443210 -0 987 -1 4322111000 -1 87766655 -2 43332111 z2_8?6_6___ -3 40 -3 6 zh___ -4 7 -5 -5 -6 -6 -7 3 PRDCFDURF FOR ASCERTAINING NORMALITY OF POPULATIONS DIFFERENCE OF EACH^VALLIS TAKEN FOR ANALYSIS_____ UNIVARIATE Tit3 TUESDAY, JULY 28, 1981 4 84 -14.3 4.57416 0.525557 379.656 0.233355 0.467734 >0.15 3 3 10 8 7 3 6 3 _I0 QUANTI L SI OEF =41 EXTREMES 100* MAX 75* Q3 50* MED 25* 01 0* MIN RANGE 03-01 MODE 5.3 1.55 -0.25 -1.675 -7.3 12.6 3.225 -2.3 99* 95* 90* 10* 5* 1* 5.3 3.25 2.45 -2.65 -3.3 ' -* LOWEST -7.3 -4.7 -3.6 -3.4 -3 HIGHEST 3.1 3.3 3.3 4.3 5.3 HISSING VALUE COUNT t COUNT/NOBS 1 1.18 RDXPLOT I I +------------ + +------ _+L I ___ I ... t 5.25 + 2.75+ 0.25+ -2.25+ 1 -4.75+ NORMAL PROBABILITY PLOT + + *** *** ******** ***+ **+ *++ ** *** *** ******* ****+___ * ++ * + _+_____ "T '" 'NI -7.25+ --+- -2 --+-1 --+- -+-----++0 + 1 2 \L u PROCEDURE FOR ASCERTAINING NORMALITY OF POP UL A TIONS OIFFFRFNCE OF EACH VALUE TAKEN FOR ANALYSIS TUT TUESDAY, JULY 28, 198! VARIARLF=LVMU1F UNIVARIATE MOMFNTS 'N MEAN___ STD DEV SKEWNESS USS nr TtMEAN'O D!NORMAL BA 1.45238 11.764? 0.0427016 11664 "HW.TS9T 1.13151 0.0642366 SUM WGTS SUM VARIANCE KURTOSIS CSS______ sttfhttan PROB>IT1 PROB>D 84 122 138. 395 -0.223499 11486.8 f .28357 0.261098 >0.15 100X MAX 75X 03 50X MEO 25X 01 OX MIN RANGE J33-OL MODE QUANTILES(DEF==41 32 8.75 3 -6.75 -26 58 15.5 5 99* 95* 90* 10X 5X 1* 32 20.75 16.5 -16.5 -17.75 -26 EXTREMES LOWEST -26 -19 -19 -18 =rr HIGHEST 20 21 26 26 32 STEM LEAF 32 2 66 2 01 1 556788 1 00033444 0 5555556777888899 0 111133333444 -0 44322211000 -0 9998766555 -1 332000 -1 998777766 -2 -2 6 16 12 II 10 6 9 MULTIPLY STEM.LEAF BY 10**+01 MISSING VALUE COUNT X COUNT/N08S boxplot 32.5 + 1 1. 18 NORMAL PROBABILITY PLOT *--+-- J_-_--_-_--_4! t t 2.5* -27.5+ * ***** ****** *** ****+ ***** ***** * ** ****** +++ -2 -1 +0 1 +2 |? I) l r H 7? no 09SP29 CONFTDFNTTi 3> J f' jI ^hrspTATn;r=TfttrfTr PROCEDURE FrtR ASCERTA IN IMF, normality OF POPULATIONS DIFFERENCE OF EACH SET OF VALUES TAKEN FOR ANALYSIS JNIVARIATF 7:30 TUESDAY, JULY 28, 1961 1 *1 1 =; n MEAN .~5ttv dev . SKEWNPSS . uss ,xv . T:MEAN=0 D:NORMAL MOMFNTS e** 0*0011904B ~ D.3B7396 0. 6599A 3 12.45 32532.9 0.02B1T19 0.125503 ' ' ~~ SUM WGTS SUM VARIANCE KURTOSIS CSS STD MEAN PROBXTl PR08>0 e* 0.1 0.149949 2.15089 12.4499 0 .0422575" 0.977593 <0.01 loot MAX 75* 03 501 "Mf 0 25* Q1 0* MIN RANGE 03-01 MODE OUANT TLES(DFF ==4 1 1.4 0.2 0 -0.275 -0.9 2.3 0.475 0.1 991 95* 90S 10* 53! rr 1.4 0.674998 .4 -0.*5 -0.6 -0.9 EXTREMES LOWEST -0.9 -0.9 -0.7 -0.6 ~ -0.6 HIGHEST 0.6 077 1 1.1 1 *9- * i it ii X 71 missing value Count * COUNT/NOBS . I 1.18 ;; ' " - '`''ft* " STEM L FAF 14 0 n tjt n 10 BOXPLOT 0 1.45 + NORMAL PROBABILITY PLOT ------- Tt 00 -- - T 0000 05555 ? 0000000 1 ooooooooooooooo 0 -0 000000000 -1 00000000000 _2 ooooo oooooooo -4 00000 -5 1550 -6 00 -7 0 -a 00 ------4------4-------4-------4 MULTIPLY STEM.LEAF BY 5 7 15 9 ii 5 a 5 3 2 1 2 10*4-01 4- -4 ! 4- -4--* 1 II 4- -4 1 1 1 1 1 1 +** 44* 4*** t 4444 1 ****** ! 4*44 1 1 44444 01 1 *4*444 44*44 1 44444 | * 44 j+ -H.85+ **** 4----4 -- -2 -1 +0 +1 +2 4m -H U SI SJ " H l : iT T on 095?3O r.O N FTD FN TTAl 9 I I J PRUCFOUR!: FOR AS CCP T A I N I NS NORMALITY OF POPULATIONS oiffepfncl of each set of values taken for analysis 7:30 TUOSOAY, JULY 28, 1981 2 ; VARI APIf = A L K o i r 1 1 Mni*FNT? N MEAN | STfi O FV SKEWNESS USS j cv T:MAN=0 D:NORMAL P4 2.2381 ~23.73 I T 0.0762609 47 166 1060*36 0.864348 0.0592763 SUM WGTS SUM VARIANCE K UR TO SI S CSS STD mean PROB>1T! PROB>D 84 1SB 563.196 0.274504 46745.2 2.58935 0.389688 >0.15 STFM tt?AF 5 QQ | 5 44 4 A 13 13 3 113 2 5699 2 012 1 556777+89 i 0134 0 56778689 | 0 11222222444 1 -0 444310 6 -0 8888766 -1 4330 -1 7766 -2 42210 -? 86 , -3 211 ] -3 8 1 -4 1 -4 -5 300 ------+------4--- # 2 2 2 3 4 3 9 4 B 11 6 7 4 4 5 2 3 1 1 3 MULTIPLY STEM.LEAF 8Y 10**+01 UNIVARIATE ; QUANTILES 1 OFF =4) EXTREMES 1 100* MAX 75* 03 50* MED 25* 01 0* MIN RANGE Q3-01 MODE 59 17 2 -13 -53 112 30 2 99* 95* 90* 10* 5* It 59 51.2498 31 -29.5 -40.25 -53 LOWEST -53 -50 -50 -41 HIGHEST 43 54 54 59 ----------- ------------- 59."' 10 u r Jl MISSING VALUE COUNT * COUNT/NOBS I 1.18 BOXPLOT I 1 1 1 t 1 1 1 4- -- 4- 11 |l 4- -+ -- * 1l !i 4- -- + 1 1 1 I 1 1 I 1 57.5 + 1 1 t 1 ! 1 1 1 1 1 2.5 + 1 1 t 1 j j ............... NORMAL PROBABILITY PLOT 4444 444 44 4 4444 44 444 444 444 44444 44 444 4444 4+t 444 44 44 44 44 444 | + f +* | 444 -52.5+ * ------*-------4----- -2 -i + 0 + 1 +2 r. . <.* .-.''V T. - ~ - B n u 33 a 1M -* O 30 SI ----------:------ -- <1 11 O ,, O 41 M S3 - ** ^ v O U ! 0> t! r* m ST M DO 0 9 5 P ^ i CONFTDFNJ y 1+ ! . 9 PROCEDURE FOR AFTER TA IN TNG NOR MAI T TV OF POPULATIONS 7:10 TUESDAY, JULY 28. 1981 3 DIFFERENCE OF FACH SFT OF VALUES TAKEN FOR ANALYSIS I[ vap t fiiji [tsho rr I- MOMENTS i; 1 , HE AN 84 -0.166667 "STD DEV 22.3444 !. SKEWNE SS 0.436539 uss 41442 rL tv -13406.7 T :MEAN=0 -0.06B3627 0:NORMAL 0.095467 SUM WGTS SUM VAR I ANCF KURTOSIS CSS STD UEAW PROB>IT I PROB>D 84 -14 ~4<5977T3 1.53298 41439.7 2.43798 0.945661 0.05T UNIVARIATE 100( MAX 75* Q3 50* MFD 25* 31 0* MIN RANGE Q3-Q1 MODE QUANTILES!DEE*41 69 11.75 -1 -13 -68 137 24.75 2 99* 95* 90* 10* 5* 1* 69 42.4999 26 -24.5 -35.5 EXTREMES LOWEST HIGHEST -60 38 -37 - 44 -37 49 -37 65 =31------- '------- 69 ns MISSING VALUE 1L COUNT * COUNT/NOBS 1 .18 STI^ L*=AT 6 5<* a BCIXPLOT 20 5 025 2 c 3 68 2| 1 2 0001345667 10 1 b my 1 122336 0 112222344455567788 -0 9887655444322110 6 18 16 +- + 11 *- -- * 3 -1 988833222110 12 +- r -2 98632221100 -3 7771 > ~4 11 1 4\ -5 r -6 8 10 t-------- --------- +--------- * MULTIPLY STEH.LEAF BY 10*4+01 NORMAL PROBABILITY PLOT 65 + * * 1+ 1 *** 1 +**4 1 ****** "1 ~ ** 1 ******* 1 ***** "1 ***** 1 ******* 1 ******** 44f+ | -654 * A-----+------- -F-------------A---------- +-------A------- --A---------- A--------- V--------- A -2 -1 *0 ! +2 |7 l ' 18 M 2< W ^ - r. H ') u n *J 3+ P 3$ 31 31 *> Xi 44 SI 'Ia l _ ri I Cl A .3 ppncrnijnF for ascfr ta[ n j ng normality cif populations 7:30 TUESnAY, JULY 2B, 19R1 DIFFERENCE OF FACH SET OF VALUFS TAKEN FOR ANALYSIS VAR 1 API C-=SGPT01F mdmfnt 2 It N k "FAN I" STO HFV SKEWNFSS h USS vr L T:MpAN=0 l. P:NORMAL i< B4 5.2619 35.3782 3.91T2B 106210 672.366 1.36316 0.189177 SUM WGTS SUM VAR[ANCE KURTHSIS css STO MEAN" PROB>1T\ PROB>D 84 442 12 5 i . 6 2 25.2055 103886 r. 86008 0.176519 <0.01 '* ST I.FAF !,, 24 5 H i r 22 rp* 20 18 i- 16 p* 14 t, 12 *1 10 * 8 6 H 60 l 4 6838 1- 2 133556739 i i 4 9 1" 0 123645668899223336655B 22 -0 8B755566310987 665566693 33322211000 34 -2 0076332210 *1 -4 I *j -6 1 10 1 1 MULTIPLY STEM.LEAF BY I0*+01 l1t '------------------^-- + ,, Ma M 12 U M UNIVARIATE QUANTIL FSf DEF=6> EXTRFHFS --j ' 100? MAX 75? 03 50? MFO 25? 01 0? MTN RANGE Q3-01 MODE 265 13.75 -1 -12.5 -71 316 26.25 -6 99t 95? 90? 10? 5? 1* 265 56.7699 35.9999 -22. 5 -29.25 -71 LOWEST -71 -61 -30 -30 -27----- HIGHEST " 53 58 60 86 365 HISSING VALIIF COUNT ? COUNT/NOBS 1 1. 18 BOXPLOT * 0 0 0 f-- +1 --+ *------ * t 1 0 2 50* i i i | i i 90 + 1 1 1 f J 1 1 -70+ * 4- - NORMAL PROBABILITY PLOT * + * + ++++**** + + 4-+***** +********** ************ * ++f4ff *+ -2 -1 +0 + 1 +2 * + -- ,, 14 - - il n X71 21 224! 7244 X 33 3: j, AT " ,, * Sss<oI < * sSB5*4I UH M K m m n n tl 3 D zO V o n C-<? osD 2 \'J)\ CO --< *> 3> PROC.FOUPE FOR ASCERTAINING NnpMALITY OF POPULATIONS T:30 TUFSOAY, JULY 28. 1981 8 ,1\ 1 0^T/SRTAF LT =GGTP TT DIFFFRCNCF OF EACH SET OF VALUES TAKEN FOR ANALYSIS UNIVARIATE A .N . MEAN skewness > us 5 u cv ,i T:MAN-0 s D:NORMAL 18 MOMENTS 64 Q.0230095 ' 20*0535 -0,R36R54 36994 a rsBA.s 0.0104643 0.172058 SUM WGTS SUM "VARIANCE" KURTOSIS CSS Sit) MEAN" PROB>|T| PRDB>0 84 2 434 .067 5. 71657 36094 iT.lt 753" 0.991676 <0.01 100? MAX 75* 03 507 MEO 257 01 07 MIN RANGE Q3-01 MODE =4QUANTIL ES(DEF J 7S 99* 0 95* 0 907 -5 107 -79 5? IS 157 13 -2 78 33.4998 20 -18.5 -37 =7T EXTREMES LOWEST -79 -76 -44 -39 =TT HIGHEST ' 23 37 40 48 - - - - - - - TB- - - - - - - I 'T__________ ,, i sirs TTAF ~ 78 w6 nt "j 17 L- *' 34 * II jr I *' I[ ^ *3 08 7 2 1123 1 22234455669 0 112334444556677889 -o 98 r66555444433332222221111000000 -l 6554322 -2 741 -3 91 -4 4 -5 --6 -7 96 --------- +----- ---- 4---------- ~MUL"T rPi"Y 5TTP LFAF RY 10**401 a H l 2 1 4 11 lfl 32 7 3 2 1 2 MISSING VALUE COUNT 7 COUNT/NOBS I 1.18 " BDXPLOT * 4 0 ] 1 *--+--* +----- -- 0 0 0 * 754 NORMAL PROBABILITY PLOT 1 45a 1 15+ 1 | -15+ | ** ++* 4 +++**** +++****** ******** ********** ****++ **++++ 1 -454 4 **4 ** | I---------------------------- -75+ * * +--------- +.--------- 4--2" +0 + 1 +2 * + Ii I] "1 1 ! m 1 j I J_______________ tJ (0 - It 3 SI k") 34 34 o 37 W 41 iJ .41 41 m 4> ' T U u 4! U M J 1 Ky U H ni .'4 gl PROCEDURE FOR ASCERTAINING NORMALITY OF POPULATIONS 7:30 TUESDAY, JULY 2B, 1981 DIFFERENCE OF EACH SET OF VALUFS TAKEN FOR ANALYSIS UNIVARIATE VAR1AFU F=StOTOIF MOMENTS QUANTIL ESIDEF =41 N '`V,, VW- VS- J t *64 i HEAN,,- -^JJ,42|571 tt ntiEr^ 3 7.9^^a SKEWNESS 5.23923 USS J 19614 SUM WGTS 84 1001 MAX .. SUM 36 751 03 VAR I ANCE 1440.95 1 501 "MEEf KURTOSIS 44.7611 25* 01 CSS 119599 01 MIN TDiiWAN?^A.lAL75%r ' J "fljgfeV1- ROtt> (T 1*0 .917635 ' ' `":K 'RRAANGE- .03-01,, TTotTE 294 T -1 -9 -123 991 951 90* 101 51 EXTREMES 294 , . LOWEST 25.75 177T----- -41 -25 -40 -36 -123'". -38 , -. -&* .,t* -...\Ai ,u*\ tA-iaLr^ HIGHEST . Z5 26 29 33 HISSING VALUE p^jCOUNT KCOUNJ^NOBS BOX PLOT * 290* ' '6 <* 1 ^ ,, 2.1,55693 , * -A 10 ,,v ' . -8*. * ^^65^556446333221110 34 j^^^rJu?T I PL 1 ST EM. ILEEA. F BY 10**+01 \ "1 jfj^RSPT^1T5E0* o 4?-*. Ij PROCEDURE FOR ASCERTAINING NORMALITY OF POPULATIONS DIFFERENCE OF EACH VALUE TAKFN FOR ANALYSIS UNIVARIATE 7:51 TUESDAY, JULY 2B, 19B1 Terr VARIABLE=RUNDIF MOMENTS HEAN& 8A -1.16667 STD DEV 5.022AA SKEWNESS -0.0618393 USS 2208 *30.495 -2.12898 L^iO 1.07779 53 SUM WGTS SUM VARIANCE KURTOSIS CSS STD PROB> I f 1 PROB>D j 8A -98 25.22A9 0.05A5A9 2093.67 O.Tj362195 ^ T15. 100* MAX * . 75* 03 50* MED 25* Q1 0* MIN - RANGE <&rm MODE OUANT IL ES I DEF==Al 12 3 -2 -4 -1A 99* 95* 90* 10? 5* 26 ..t -2 " 12 ` ',6.7*998 EXTREMES WZT LOWEST ' , HIGHEST , , I: -1* - '* y v.l :.......* ,.;cdunt^ :% ''N^.-^7- MISSING VALUE COUNT/NOBS . 1.18>;^|f; 4>`- , STEM leaf /. : 1 i2o II T.Loo boxplot t NORHAL probability plot i&l 7 00 V 6 OOO 5 000 7300000" "000000 ,000 1 0000000 0. -0 000000 ItOOOAi# ' 2,-Loooooooooo 3.00000 `` %-> (.000000000 es., w 'I* /, nnnn E>0 -10 0 -11 0 5-;--i2 ; MH-ttfjsSu fe-.vr,***.** V'1- -*&rpWf .V*JI T. JU* . . 9$ ' v IT At SbG SfcA>SF&2>jj&A?y': ^ S7SS> w ^vysiSV/? &XA2V ^9/ sW&CS*&%> x? ^`/^-04r ^ *2X2/7? /4V^^> /R5^ft!6C?^r A/PS/Z/Qf ^oc ^ z$w2> <5aa^0^\*^^*^Z7>4>as& , 77&<&Q.0a&&cfje y*//.jya>&)> SX/SWO/SA//? &/&& &&VZ&A/XS A2><& &A>AS sG&A/A^VPAt^AAy* f /as Sax*/ &&<#/*&& 0^&<*>sesAs&-J, sV'&tUt&essO^g A//pp/ &&a?7V i/4/z>e<e * ^ ax^e/t 2(%'0&yg&> Tifcfsfrc *&& SS&Z?AsA* ^WT /ZSJSTVi/r A**sZ> ^^A^JSS ^S- '&''?**' ^^aeAszaAAAy&s 5^^^fi5V/`^/<^ JSZX7S7?<ZS A/S Ak<##/&& ^/42^V/ ^ c*ma*m/s em. , a&/T says?/ ?&/ &/z//>/ xit^ OaArtsea*, TPmas as passm/e s?sa*>*&ss ws*# 's&7&t* ^t^s s*//> as w* ja&*sey~: -sxvaa y&ex/yZ's&S&e. as SXXves a^esas S/AAtSASSOy AOktfs 72A4 ATXA?AS- StF&7~ OA>*Ac2> SS SSseA&GAP&b, WZ> At S'-SSAASS Ot/3X*?AlSS2> SX&JA7 75MT ASSS&e Atea&dtSaaaS*S MO/SS fs&s SPA'S 9) S<?as 0a&7? 7?/* Stc/?dr" resr ae+ts S&ssX&tsaazZi aWS> a? ^SAtx*/#a*mAA*> /5S^ Z&<4ASS. Sa/ca>o&> as a? xast~ as sM? sssssss &e*&*u& s#e wr /?SSea/>A/A77vA' saSatpaa/SS Ats S&ass? 77?<4AS ^Z7~J S?& SS^/'AAX/Ar S/CaPS S*G0SZ><2AKAs. do CONP 09s?-08 tdfnt r ai  O1 ur ULt-LHULml V A |\ i o L t :> rnun UUL 6 I l UN A1 KL i> PAbt 1 variable: - R1TH ABNORMALITY L AT UbLU""""*,""ta" R ESPONSE NUMBER/TOTAL RESPONSE NUNBER/TOTAL PAIR 1 PAIR PAIR 2 7 PAIR A PAIR PAIR A PAIR 7 PAIR PAIR 9 PAIR IP PAIR U PAIR 12 PAIR 13 " PAIR 14 PAIR 15 PAIR 16 PAIR T7..... PAIR 16 PAIR 19 PATR '2 0 PAIR 21 PAIR 22 ----------- PATR~~ 23 PAIR 24 PAIR 25 PAIR Pfe PAIR 27 PAIR 28 PAIR T9' PAIR 30 PAIR 31 PATR' T'2 ' PAIR 33 PAIR 34 j----------- PAIR' 35 PAIR 36 PAIR 37 1 PAIR 38 PAIR 39 PAIR 40 PAIR T1 PAIR 42 PAIR 43 PAIR 44 PAIR 45 PAIR 46 -----------PATR 47 PAIR 48 NO NO NO NO NO NO NO NO NO YES* ' NO NO NO yTT* NO NO ------------ NO NO NO NO NO YES** -------- RTS------------YES** NO NO NO YES** ------------ m--------- NO NO FTo NO NO Tsro NO NO NO NO NO -------------NO--------NO NO --ye:?** NO NO ----- NO--------NO ' 'r:-.4 - ^ ' 1/2 1/2 1/2 1/0 r;f * ' "L'' NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO No v T ' ' * -- " - - .'':i ' ' 'T'-'STT*- NO " .. - --r- NO NO NO NO \ ^ j' < iioTV-i-v: ..NO. \ NO NO NO . NO ' . no"^ , -^*X. t-S- - JK. no : ^ * Ht -' ^ NO YES** 1/2 NO NO . ' T".` ~ NO NO . NO NO YES** I^q ' . " ' ' ', 1/2 jV * .-y \ r ->%. .* NO ' '/' ' *? T < :,VT : NO NO NO 'Vy\ ' T / r'1 DO 095739 CONFIDENTIAL  Wbw I i 1 |\|. v) VARIABLE - BIRTH ABNORMALITY t. A r Ui>tl/ RESPONSE NUMBER/TOTAL RESPONSE" NUMBER/TOTAL ~ 1 J34 I PAIR PAIR PAIR PAIR PAIR PAIR PAIR PAIR PAIR PAIR PAIR PAIR PAIR PAIR PAIR PAIR PA" ITT PAIR PAIR PAIR PAIR PAIR PAIR PAIR PAIR PAIR PAIR PAIR PAIR PAIR PAIR PAIR PAIR PAIR PAIR PAIR 49 5U 51 5? 53 54 55 56 57 58 P/9 60 61 62 63 64 SF 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 ' NO NO NO NO NO NO NO NO NO NO NO YES** NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO YES** 1/2 's'- P' i \ 1/4 NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO YES* NO NO NO io -- '' ' ` -VS. sy 7SST~ s*? -. .y/3^ TOTAL POS RES 8 REP 9.5X TOTAL POPULATION SIZE - 84 6/12 e 3 3.5X 2/4 r ` v *;' fev* ' *7_' k.aC*f** '.** t ' V . : .* ' " ft -^Ar. v. - / ^fJSSem D0 09S240 CONFfOFNTTAU l-JUN-81 ANALYSIS OF DEPENDENT VARIABLES FROM QUESTIONAIRES PAGE I' " VARIABLE - MISCARRIAGES r v ppcrrn ------------------ AAklTD Al ` /T RESPONSE NUMBER/TOTAL RESPONSE NUMBER/TOTAL' s. i PAIR PAIR pair 1 9 * PAlk 4 PAIR "5 PAIR 6 PAIR ----------- PAIR 7 n PAIR 9 PAIR 10 ----------- fair: 11 PAIR 12 PAIR 13 ----------- FTTP 1A PAIR 15 PAIR 16 ----------- PATH" rr PAIR 18 PAIR 19 ----------- PTrnr- "2 0 PAIR 21 PAIR 22 ----------- FAIR 23 - PAIR 24 PAIR 25 FAIR- 26 PAIR 27 PAIR 28 -----------PATR- 29 PAIR 30 PAIR 31 ----------- FAIT! 32 PAIR 33 PAIR 34 r--PAIR 35 PAIR 36 PAIR 37 'iL PATR 38 PAIR 39 PAIR 40 FAIR "41 PAIR 42 PAIR 43 Pair 44 PAIR 45 PAIR 46 NO NO NO YES NO YES NO NO"' ' YES NO HO--------NO NO YES YES NO ~ fro--------NO NO NO ..... NO NO m-- NO NO NO NO NO ms--------- NO NO ----- FTO NO YES --fro-------- YES NO NO NO YES ----------- JJO NO NO fro NO NO NO 1/2 1/1 2/2 2/2 1/1 ._ , - 1/3 1/4 1/2 NO NO NO NC NO NO NO NO NO NO YES NO NO NO NO NO NO NO NO NO NO NO NO YES YES NO NO NO NO NO NO YES NO YES NO NO NO NO NO NO NO NO NO NO NO NO NO NO 1/0 ' ft*** . 1/2 2/2 7 `' 1/4 1/2 V* sVviv**^. l-JUN-81 ANALYSIS Lit- ULKLNULNl VAKlABLtS rKUrt WUtS I 1UHAJ Rt-i' rni)L *1. *1 **1* 11a 1 *;.;i '!!'! '0 11 :'M :!:;i Is"o !< P'i VARIABLE - MISCARRIAGES -EXPOSED- RESPONSE NUMBER/TOTAL PAIR 49 PAIR 50 PAIR 51 PAIR 52 PAIR 53 PAIR 54 PAIR 55 PAIR 56 PAIR 57 PAIR 58 ------ PAIR-~^r PAIR 60 PAIR 61 'PAIR 62 PAIR 63 PAIR 64 NO YES NO NO NO NO NO NG NO NO NO NO NO NO ' NO NO 1/3 -CONTROL- RESPONSE NUMBER/TOTAL NO NO NO NO NO NO NO NO YES YES NO NO NO NO NO NO 1/2 1/1 , -- , j'jjf ----- ' ? */; |S2 S3 ' PAIR PAIR PAIR PAIR PAIR 80 31 82 83 84 NO NO NO fio NO Isa i-r, to 1 TOTAL POS RES 11 percentage pos RES 13.OX TOTAL POPULATION SIZE - 84 13/21 - lin-t tl>!" NO NO NO NO . NO 9 10.7% 11/16 -I*' ' ' ' y v 4 * y ' - &W--' 5 1` -in. *?' v*4fy^,V' V A VARIABLE - CHILDREN STILLBORN A HIULO ) rwr. UULO I JUWMIAL r aul h PAIR 1 PAIR 2 PAIR 3 PAIR A PAIR 5 PAIR 6 PAIR 7 PAIR " 8 PAIR 9 PAIR 10 FSTR II " PAIR 12 PAIR 13 ----------- P7QR"" FT PAIR 15 PAIR 16 ------ :--FTTIK""17-- PAIR 18 ---------PpAAIIRR 19 "28 PAIR 21 PAIR 22 ----------- PAIR "23-- PAIR 24 PAIR 25 ----------- prnr 26 PAIR 27 pPAA1IRR 28 29 PAIR 30 | PAIR 31 PITT"32~ PAIR 33 PAIR 34 PTTRT 35 PAIR 36 PAIR 37 PAIR 36 PAIR 39 PAIR 40 PAIR 41 PAIR 42 PAIR 43 TTTTR 44 PAIR 45 PAIR 46 ---------------- EXPOSED-------------------- RESPONSE NUKBFR/TOTAL NO NO NO YE r NO NO NO NO NO NO "ATO NO NO NO NO NO R7T~ NO NO "NTT NO YES ~ TTO-----NO NO NO NO NO NO NO NO NO NO NO no ~ NO NO NO ... NO NO RTJ------- NO NO ~pto NO NO 1/2 1/2 ' : *'' }; " '' ti> ' ** , ,2'.',-*.'V,, > -----------------CONTROL------------------- RESPONSE NUMBER/TOTAL NO NO NO NO NO : 4- -w' NO NO NO -- NO NO NO NO . -* NO NO NO NO NO NO NO ' ' - ,.v -v^tj,v?--v/ y->v <r NO NO NO NO NO ;' ' \ *',+ ,, >* i,*-, $*'*$: ' "v"' v,,> J*sy ** s' ' `1iv, \ . ; '' <* \" 4f ' NO NO NO NO "NO NO NO ,, NO NO NO NO . > ' '' '^Vs vV ' ' : V , - NO I.,'.. NO NO NO NO NO NO NO NO NO NO " UkJ 111 ~o J. HIIHLI ur L/L r l_ IVLf L 111 I VAIUHDLLA r (\ U 1*1 liUti I lOIMAJKLi & ARIABLE - CHILDREN STILLBORN if , PAIR . V * . 1V- *'*' ^< *.49 , EXPOSED---------- -------- .RESPONSE . v ""V m'*-s/ "A * i NUMBER/TOTAL NO .i*' . . -----------------CONTROL RESPONSE . NUNBER/TOTAL -v . 4- NO -"'H PAIR 50 NO NO PAIR 51 NO NO PAIR 52 ; PAI*R'` 53 NO . NO NO - V: `NO PAIR . 54 NO NO PAIR 55 PAIR 56 NO NO NO NO *- PA IB '57 NO NO PAIR 58 NO NO PA IB ^59. . NO NO PAIR ^ 6 0, - . -.NO , NO '< 4 ^ no PAIR 62 NO NO PAIR 63 NO NO PAIR - 64 NO NO mm fPATR PAIR PAIR PAIR 65 66 67 68 69 . tv. NO ' NO NO NO NO , '. . -' NO ; 5,-' j Z7- NO : i NO NO NO i ; 7Z&7~: /- .JZ PAIR 70 NO NO PAIR 71 NO NO PAIR 72 NO NO PAIR 73 NO NO PAIR 74 NO NO PAIR 75 NO NO PAIR 76 PAIR 77 NO . NO NO NO PAIR 78 NO NO PAIR 79 NO NO PAIR 80 NO NO PAIR 81 NO NO PAIR 82 --PATF" 83 NO tfo NO NO . i ij , Sw.4r *. PAIR 84 NO NO ' * >Y-' ' ; TOTAL POS RES 2 PERCENTAGE POS RES 2-3% TOTAL POPULATION SIZE - 84 2/4 0 0.01 0/0 4 * -------- . * mP.L i o ur uc-r LIMULivl VAKlAbl.LS PROM QUESTIONAIRES VARIABLE - DEATH UNDER 6 MONTHS PAGE 1 KtSPUNSE NUMBER/TOTAL RESPONSE NUHBER/TOTAL PAIR PAIR PAIR PAIR PAIR PAIR PAIR PAIR PAIR PAIR "Pa i R PAIR PAIR PAIR PAIR PAIR FRTR PAIR PAIR "PAIR PAIR PAIR 1 2 3 A 5 6 7 1 9 10 II 12 13 14 15 16 11 18 19 20 21 22 NO NO NO NO NO NO NO NO NO NO " NO NO NO NO NO NO NO NO NO " ' "ARTNO NO PAIR 24 NO PAIR 25 NO PAIR-"56------------ NO PAIR 27 NO PAIR 28 YES PAIR 29 RTS PAIR 30 NO PAIR 31 NO "PAIR T5------------ RTT" PAIR 33 YES PAIR 34 NO 35 RTJ PAIR 36 NO PAIR 37 NO PAIR "31 NO PAIR 39 NO PAIR 40 NO PAIR *t~ ; . NO PAIR 42 NO PAIR " PAIR 44 V-'.v -NO, YfcV PAIR 45 NO PAIR 46 NO , t 1/2 1/3 '-sj-k"! " \. NO NO NO NO NO NO NO NO NO NO NO NO 'NO NO NO NO -- NO NO no NO NO YES NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO , , .k ' -`Vir- v ,,rt*' . j,.a'i'x- -,v,` .* ;v.: . 2/2 r . - *? - ,*? - k-i. H**W3NO.' - '- -- VTA'M 1 No -,' ,. r V>;i 1 ''V S/WIK O J* hi'*/i u i o i o ur uu r uv i/ v- rrrw UL.v I JLlSiMAirtLO r A15L VARIABLE - DEATH UNDER 6 MONTHS ^" LrvnDl vA*.ciLrun--to*aR>Ba RESPONSE NUMBER/TOTAL RESPONSE NUMBER/TOTAL PAIR 49 PAIR 50 PAIR 51 PAIR 52 PAIR 53 PAIR 54 PAIR 55 PAIR 56 PAIR 57 PAIR 58 ----------- FATF""59 PAIR 60 PAIR 61 PAIR 62 PAIR 63 PAIR 64 r*---pa nr 65 PAIR 66 PAIR 67 PAIR 68 PAIR 69 PAIR 70 PAIR n PAIR 72 PAIR 73 PAIR 74 PAIR 75 PAIR 76 PAIR "77 PAIR 78 PAIR 79 PAIR 80 PAIR 81 PAIR 82 pair 83 PAIR 84 NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO NO YES NO NO NO NO NO NO 1/1 - / TOTAL POS RES 3 prRtnmrcr p o s RES 3.5% TOTAL POPULATION SIZE - 84 3/5 2 2.3X j.. ... 3/3 ' ' ' : f v- A C 86 10 G 7Zr&r : / t