P307-2 1.不分层计算结果如下 Chi-Square Tests alue 2-sided) 2-sided) 1-sided 80989 Continuity Corectiona80320 000 Likelihood ratio 82.761 Fis hers Exact Test Linear-by-Linear Associ ation 80988 000 N of valid cases 330372 a Computed only for a 2 2 table b 0 cells(0%)have expected count less than 5. The minim um expected count is 357 Ris k Estimate nterval Value Odds ratio for石棉(1/2) 2013 2351 For cohort癌死亡=1 2.010 1.721 2347 For cohort崩瘾死亡=2 999 998 Nof valid Cases 330372 计算得:x2=80.98,P<0.001; 相对危险度RR=2010,95%置信区间为(1.721,2.347) 因此认为石棉工人肺癌死亡率高于非石棉工人,是非石棉工人的2.01倍 2.分层后计算结果如下:
P307-2 ⒈ 不分层计算结果如下： Chi-Square Tests 80.989b 1 .000 80.320 1 .000 82.761 1 .000 .000 .000 80.988 1 .000 330372 Pearson Chi-Square Continuity Correctiona Likelihood Ratio Fisher's Exact Test Linear-by-Linear Association N of Valid Cases Value df Asymp. Sig. (2-sided) Exact Sig. (2-sided) Exact Sig. (1-sided) a. Computed only for a 2x2 table 0 cells (.0%) have expected count less than 5. The minimum expected count is 357. 85. b. Risk Estimate 2.013 1.723 2.351 2.010 1.721 2.347 .999 .998 .999 330372 Odds Ratio for 石棉 (1 / 2) For cohort 肺癌死亡 = 1 For cohort 肺癌死亡 = 2 N of Valid Cases Value Lower Upper 95% Confidence Interval 计算得： 2 =80.98，P＜0.001； 相对危险度 RR ˆ =2.010，95%置信区间为（1.721，2.347）。 因此认为石棉工人肺癌死亡率高于非石棉工人，是非石棉工人的2.01倍。 ⒉ 分层后计算结果如下：
Chi-Square Tests Asymp. Sig. Exact Sig. Exact Sig 年龄 2-sid 10.00 Pearson Chi-Square 6.110 Continuity Correction 020 Likelihood Ratio 6.388 1 Fis hers Exact Test 016 010 Linear-by-Linea 6.110 013 n of valid cases 163857 20.00 Pearson Chi-Square 41.378c .000 Continuity CorrectioN.564 000 Likelihood ratio Fis hers Exact Test Linear-by-Linear Association 41378 000 N of valid Cases 114453 30.00 Pearson Chi-Square 36.114d .000 Continuity CorrectioA35386 Likelihood ratio 000 Fis hers exact Test 0 000 Association 36.113 1 00 N of valid Cases 36361 40.00 Pearson Chi-Square 21.162e Continuity CorrectioA20411 Likelihood ratio 22.158 000 Fis hers Exact Test 000 000 Association 21.160 000 n of valid cases 15665 a. Computed only for a 2x2 table b. 0 cells (0%)have expected count less than 5. The minimum expected count is 22.70 C0 cells (0%)have expected count less than 5. The minimum expected count is 113.42 d. 0 cells (0%)have expected count less than 5. The minimum expected count is 127. 64 e0 cells(0%)have expected count less than 5. The minimum expected count is 68. 94
Chi-Square Tests 6.110b 1 .013 5.428 1 .020 6.388 1 .011 .016 .010 6.110 1 .013 163857 41.378c 1 .000 40.564 1 .000 41.486 1 .000 .000 .000 41.378 1 .000 114453 36.114d 1 .000 35.386 1 .000 36.346 1 .000 .000 .000 36.113 1 .000 36361 21.162e 1 .000 20.411 1 .000 22.158 1 .000 .000 .000 21.160 1 .000 15665 Pearson Chi-Square Continuity Correctiona Likelihood Ratio Fisher's Exact Test Linear-by-Linear Association N of Valid Cases Pearson Chi-Square Continuity Correctiona Likelihood Ratio Fisher's Exact Test Linear-by-Linear Association N of Valid Cases Pearson Chi-Square Continuity Correctiona Likelihood Ratio Fisher's Exact Test Linear-by-Linear Association N of Valid Cases Pearson Chi-Square Continuity Correctiona Likelihood Ratio Fisher's Exact Test Linear-by-Linear Association N of Valid Cases 年 龄 10.00 20.00 30.00 40.00 Value df Asymp. Sig. (2-sided) Exact Sig. (2-sided) Exact Sig. (1-sided) a. Computed only for a 2x2 table b. 0 cells (.0%) have expected count less than 5. The minimum expected count is 22.70. c. 0 cells (.0%) have expected count less than 5. The minimum expected count is 113.42. d. 0 cells (.0%) have expected count less than 5. The minimum expected count is 127.64. e. 0 cells (.0%) have expected count less than 5. The minimum expected count is 68.94
Ris k Estimate 95% Confidence 年龄 alue 10.00 Odds Ratio for 石棉(有/无) 1.154 3.966 For cohort肺癌 死亡=死亡 1.153 3.964 For cohort肺癌 死亡=存活 1000 1000 1000 Nof Valid Cases 163857 20.00 Odds Ratio for 石棉(有/无) 2300 1772 2987 For cohort肺癌 死亡=死亡 2296 1.769 For cohort肺癌 死亡=存活 998 Nof Valid Cases114453 30.00 Odds Ratio for 石棉(有/无) 2.111 645 2.709 For cohort肺癌 死亡=死亡 2.100 1639 2689 For cohort肺癌 死亡=存活 995 Nof valid cases 36361 40.00 Odds Ratio for 石棉侑有/无) 2278 1588 3266 For cohort肺癌 死亡=死亡 2261 1581 3234 For cohort肺癌 死亡=存活 993 990 996 N of valid cases 15665 Tests of Conditional Independence Chi-Squared (2-sided Cochrans 000 Mantel-Haensze 103698 000 Under the conditional independence assumption, Cochrans statistic is asymptotically dis tributed as a 1 df chi-s quared dis tribution, only if the number of strata is fixed, while the Mantel-Haenszel statistic is always asymptotically distributed as a 1 df chi-squared dis tribution. Note that the continuity orrection is removed from the mantel-Haenszel statistic when the sum of the differences between the observed and the expected is 0
Risk Estimate 2.139 1.154 3.966 2.138 1.153 3.964 1.000 1.000 1.000 163857 2.300 1.772 2.987 2.296 1.769 2.980 .998 .998 .999 114453 2.111 1.645 2.709 2.100 1.639 2.689 .995 .993 .996 36361 2.278 1.588 3.266 2.261 1.581 3.234 .993 .990 .996 15665 Odds Ratio for 石 棉 (有 / 无) For cohort 肺 癌 死 亡 = 死 亡 For cohort 肺 癌 死 亡 = 存 活 N of Valid Cases Odds Ratio for 石 棉 (有 / 无) For cohort 肺 癌 死 亡 = 死 亡 For cohort 肺 癌 死 亡 = 存 活 N of Valid Cases Odds Ratio for 石 棉 (有 / 无) For cohort 肺 癌 死 亡 = 死 亡 For cohort 肺 癌 死 亡 = 存 活 N of Valid Cases Odds Ratio for 石 棉 (有 / 无) For cohort 肺 癌 死 亡 = 死 亡 For cohort 肺 癌 死 亡 = 存 活 N of Valid Cases 年 龄 10.00 20.00 30.00 40.00 Value Lower Upper 95% Confidence Interval Tests of Conditional Independence 104.465 1 .000 103.698 1 .000 Cochran's Mantel-Haenszel Chi-Squared df Asymp. Sig. (2-sided) Under the conditional independence assumption, Cochran's statistic is asymptotically distributed as a 1 df chi-squared distribution, only if the number of strata is fixed, while the Mantel-Haenszel statistic is always asymptotically distributed as a 1 df chi-squared distribution. Note that the continuity correction is removed from the Mantel-Haenszel statistic when the sum of the differences between the observed and the expected is 0
Mantel-Haenszel Common odds Ratio Estimate 2.211 Std Error of In(Estimate) As ymp. Sig (2-sided) As ymp. 95% Confidence Common Odds Lower Bound Interval Upper Bound In( Common Lower Bound 637 Odds Ratio) Upper Bound 949 The Mantel-Haenszel common odds ratio estimate is asymptotically normally dis tributed under the common odds ratio of 1.000 assumption. So is the natural log of the estimate (1)10岁~组 x2=6,11,P=013;RR=2.138,95%置信区间为(1.153,3.964) (220岁~组: x2=41.378,P<0.001;RR=2.296,95%置信区间为(1769,2.980) (3)30岁~组: x2=36.114,P<0.001:RR=2.100,95%置信区间为(1.639,2689) (4)40岁~组: x2=21.162,P<0001;RR=2.261,95%置信区间为(1.581,3,234) 分层后计算x2=103698,P<001;调整相对危险度RRa=2.21l 95%置信区间为(l.891,2.584)。说明经年龄分层后石棉工人肺癌死亡 率仍高于非石棉工人,是非石棉工人的2.211倍。各年龄组石棉工人肺癌 死亡率均高于非石棉工人,调整相对危险度RRa值、分层后各年龄组 RR值与不分层相对危险度RR值大致相同,说明年龄不是混杂因素
Mantel-Haenszel Common Odds Ratio Estimate 2.211 .793 .080 .000 1.891 2.584 .637 .949 Estimate ln(Estimate) Std. Error of ln(Estimate) Asymp. Sig. (2-sided) Lower Bound Upper Bound Common Odds Ratio Lower Bound Upper Bound ln(Common Odds Ratio) Asymp. 95% Confidence Interval The Mantel-Haenszel common odds ratio estimate is asymptotically normally distributed under the common odds ratio of 1.000 assumption. So is the natural log of the estimate. ⑴10岁～组： 2 =6.11，P=0.013； RR ˆ =2.138，95%置信区间为（1.153，3.964） ⑵20岁～组： 2 =41.378，P＜0.001； RR ˆ =2.296，95%置信区间为（1.769，2.980） ⑶30岁～组： 2 =36.114，P＜0.001； RR ˆ =2.100，95%置信区间为（1.639，2.689） ⑷40岁～组： 2 =21.162，P＜0.001； RR ˆ =2.261，95%置信区间为（1.581，3.234） 分层后计算 2 =103.698，P＜0.001；调整相对危险度 RR ˆ a=2.211， 95%置信区间为（1.891，2.584）。说明经年龄分层后石棉工人肺癌死亡 率仍高于非石棉工人，是非石棉工人的2.211倍。各年龄组石棉工人肺癌 死亡率均高于非石棉工人，调整相对危险度 RR ˆ a值、分层后各年龄组 RR ˆ 值与不分层相对危险度 RR ˆ 值大致相同，说明年龄不是混杂因素