Document 85q2aeK24kNJg9ww75Gn7Kmne

EH-1029 CSfMU9DMP2 E. 3. d u Po n t d e Ne mo u r s 2* Co mp a n y INCORPORAteO Wil min g t o n , De l a w a r e t9898 ENGINEERING DEPARTMENT November 17, I972 A,' Azar - Haskell Lab D. R. Diggs - Orchem,Wilm R. C. Butler " W. E. Bettoney- Pet Lab A, J, Pahnke 11 " E, S. Jacobs "" W. E. Morris "" J. N, Whelan R. W, Kennard D, W. Marquardt RC 13 - Applied Statistics M. MC LAUGHLIN (3) EMPLOYEE RELATIONS HASKELL LABORATORY A. L. LINCH (3) MEDICAL DIVISION CHAMBERS WORKS DU PONT FIVE-YEAR Pb STUDY - Pb DETERMINATION QUALITY CONTROL This letter summarizes the analysis of the urine Pb and blood Pb percent recovery data collected by the Chambers Works Medical Division. These data serve as a quality control check on the Pb determination methods used in analyzing the urine and blood samples collected in the Du Pont Five-Year Pb Study. No significant trends could be detected in either the urine Pb or the blood Pb percent recovery data* The urine Pb percent recoveries varied from 90 to 98% (average = 93.2%) with the exception of three months (November 1971 77.3%, December 1971 85.2%, February 1972 81.7%), while the blood Pb recoveries varied from 57 to 95% with an average of 81%. It is concluded that there is no evidence in these percent recovery data that significant negative trend in average urine Pb levels in the Du Pont Five-Year Pb Study (see letter from R, D. Snee to M. McLaughlin, "Du Pont Five-Year Pb Study", 9/6/72) can be attributed to differences in the method of urine Pb determina tion over the five-year period. ENGINEERING SERVICE DIVISION N 27020 R. D. Snee Applied Statistics RDS:chm Attachments RECEIVED 4 1972 Haskell Laboratory DUP040005915 DU PONT FIVE-YEAR Pb STUDY Analysis of Urine Pb and Blood Pb Percent Recovery Data I. Objective In the analysis of the urine Pb and blood Pb data from the Du Pont Five-Year Pb Study we found a significant negative trend in the average urine Pb levels of the subjects (letter from R, D. Snee to M, McLaughlin, 9/6/72). It's possible that this negative trend may have resulted from differences in laboratory procedures over the five-year period 1967-1972. The percent recovery of the analytical method of urine Pb determination was routinely checked from 9/68-4/72 (Table 1), The percent recovery of the blood Pb analysis was checked less frequently from 8/66 to 4/72 (Table 3). Any significant trends in the urine Pb and blood Pb percent recovery data would be an indication of differences in the analytical method, II. Statistical Analysis and Results Time plots of the. urine Pb and blood Pb percent recovery data are shown in Figures 1 and 2, The urine Pb recoveries varied from 90 to 98% with the following exceptions (Figure 1), Month % Recovery November 1971 December 1971 February 172 77.3 85.2 81.7 There were no significant trends. These three abnormal months cannot explain the significant negative trend we observed in the average urine Pb levels. It should be noted also that the standard deviation of replicate urine Pb percent recoveries within a month increased in 1971 and 1972, compared to previous years (Table 2). There also were no significant trends in the blood Pb percent recovery data (Table 4, Figure 2) which varied from 57 to 95%, with an average of 81%. It is of interest that approximately two-thirds of the variability in blood Pb percent recovery is variation between replicate blood Pb percent recoveries run on the same day (Table 4). N 27020.01 DUP04000S916 Month 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 TABLE 1 Urine Pb Percent Recovery Data Year Number of Det 68 7 68 24 68 16 68 17 69 f9 69 69 69 69 9 69 69 -i 69 69 69 16 29 24 35 21 6 19 20 20 31 24 24 70 24 70 26 70 16 70 15 70 17 70 15 70 9 70 70 3 70 7 70 7 70 9 71 11 71 10 71 12 71 14 71 13 71 6 71 7 71 8 71 15 71 12 71 3 71 5 72 7 72 7 72 7 72 8 N 27020.02 Average 9-5.7 92.8 93.8 9 X.4 92.2 94.4 93.7 91.6 93.6 91.0 92.7 91.4 91.9 94.1 94.6 96.7 96.2 9^.0 94.0 90.3 90.2 96.3 93.1 96.0 93.1 95.4 96.0 91.3 94.8 93.7 93.3 93.2 97.7 91.7 91.2 93.7 92.2 77.3 85.2 94.6 81.7 95.7 90.0 DUP040005917 wCM O .tv On Year 1968 (1) 1969 1971 1972 (3) TABLE 2 Urine Pb Percent Recovery Data Number of Determinations 64 269 148 '116 29 Average 93.0 '93.4 94.3 92,4^ 90, .5 ^ Standard Deviation^' 7.56 5.37 5.86 8.73 15.11 Based on 4 months of data - September to December ^No data for August (3)'Based on 4 months of data - January to April ^Standard deviation of replicate determinations within a month 'Significantly low because of the low % recoveries in Nov & Dec ^Significantly low because of the low % recoveries in Feb N 27020.03 DUP040005918 Month 8 4 7 2 3 4 2 3 6 7 8 10 11 4 TABLE 3 Blood Ph Percent Recovery Data \0o0 Year 66 67 67 69 69 7.1 71 71 71 71 71 71 72 Number of Determinations 2 16 4 2 15 5 2 4 9 16 9 8 16 11 Average 82; 5 83,2 76.O 92.5 87,1 Q5.0 72,0 57,0 74.0 86.4 84,8 76.5 78.6 73,8 Total Number of Determinations = 119 Average Percent Recovery = 81.0 Standard Deviation of Replicate Determinations on the same day - 12.87 N 27020.04 DUP040005919 Source Total Year-Month Days (Y-M) Replicates TABLE 4 Blood Pb Percent Recovery Data Analysis of Variance df SS MS 118 28590.9 13 6322.2 486.3 45 12332.5 274,1 60 v 9936.2 165.6 F 1.42 1.65* Source Year-Month Days Replicates Total Variance Component 17.6 62.1 165.6 245.3 Percent of Total Variation 7.2 25.3 67.5 100,0 Standard Deviation 4.19 7.88 12.87 Significant at the .05 probability level N 27020.05 DUP040005920 N 27020.06 U R lA )9 T b f o R&COy&gV 'b/VF?) o o v' e- O Z' ^ -,Xi irv O T"- r*- DUP040005921 FIGURE 2. BLOOb TW % RetWERY b m \ N 27020.07 Oo <s- ,>$3*00311 % 3 DUP040005922