Abstract：
BACKGROUND: Our aimed to conduct a meta-analysis of cohort studies on risk of genitourinary (GU) cancers in workers exposed to welding fumes (WF).
METHODS: We performed a systematic review of studies published on Pubmed, Scopus and Embase following PRISMA criteria. Two researchers selected cohort studies on WF exposure. From 2582 articles, 7 non-overlapping studies were included. Quality of studies was scored according to CASP. We run a random effects meta-analysis to calculate the relative risk (RR) and 95% confidence intervals (CI) of GU cancer, overall and stratified by cancer, country, and quality score.
RESULTS: We included seven studies reporting results on GU cancers, including prostate, bladder and kidney cancer (PC, BC, and KC). The RR was 1.19 (95% CI = 1.07-1.32, 16 risk estimates) for GU cancer; 1.13 (95% CI = 0.90-1.42, 4 risk estimates) for PC; 1.26 (95% CI = 0.98-1.60, 7 risk estimates) for BC and 1.28 (95% CI = 1.12-1.47, 5 risk estimates) for KC. Heterogeneity was present in all meta-analyses (p < 0.001). The increased risk was more pronounced in North American than in European studies (respectively, OR = 1.35, 95% CI = 1.18-1.55; OR = 1.13, 95% CI = 1.01-1.27 p heterogeneity = 0.03). There was no heterogeneity according to quality score (p = 0.4). Data were insufficient to investigate associations by industry or welding type. Publication bias for each cancer was excluded.
CONCLUSIONS: This meta-analysis suggests increased risk of KC and BC, but not of PC, in workers exposed to WF. Confounding by other occupational and non-occupational risk factors could not be excluded. Data were not adequate to address the risk of specific exposure circumstances.
METHODS: We performed a systematic review of studies published on Pubmed, Scopus and Embase following PRISMA criteria. Two researchers selected cohort studies on WF exposure. From 2582 articles, 7 non-overlapping studies were included. Quality of studies was scored according to CASP. We run a random effects meta-analysis to calculate the relative risk (RR) and 95% confidence intervals (CI) of GU cancer, overall and stratified by cancer, country, and quality score.
RESULTS: We included seven studies reporting results on GU cancers, including prostate, bladder and kidney cancer (PC, BC, and KC). The RR was 1.19 (95% CI = 1.07-1.32, 16 risk estimates) for GU cancer; 1.13 (95% CI = 0.90-1.42, 4 risk estimates) for PC; 1.26 (95% CI = 0.98-1.60, 7 risk estimates) for BC and 1.28 (95% CI = 1.12-1.47, 5 risk estimates) for KC. Heterogeneity was present in all meta-analyses (p < 0.001). The increased risk was more pronounced in North American than in European studies (respectively, OR = 1.35, 95% CI = 1.18-1.55; OR = 1.13, 95% CI = 1.01-1.27 p heterogeneity = 0.03). There was no heterogeneity according to quality score (p = 0.4). Data were insufficient to investigate associations by industry or welding type. Publication bias for each cancer was excluded.
CONCLUSIONS: This meta-analysis suggests increased risk of KC and BC, but not of PC, in workers exposed to WF. Confounding by other occupational and non-occupational risk factors could not be excluded. Data were not adequate to address the risk of specific exposure circumstances.
摘要：
背景：我们的目的是对暴露于焊接烟雾（WF）的工人的泌尿生殖系统（GU）癌症风险的队列研究进行荟萃分析。
方法：我们对发表在Pubmed，Scopus和Embase遵循PRISMA标准。两名研究人员选择了关于WF暴露的队列研究。从2582篇文章中，纳入7项非重叠研究。根据CASP对研究质量进行评分。我们运行随机效应荟萃分析来计算GU癌症的相对风险(RR)和95%置信区间(CI)。整体和按癌症分层，国家,和质量得分。
结果：我们纳入了7项报告GU癌症结果的研究，包括前列腺,膀胱癌和肾癌(PC，BC,和KC）。GU癌症的RR为1.19（95％CI=1.07-1.32，16个风险估计值）；PC为1.13（95％CI=0.90-1.42，4个风险估计值）；BC为1.26（95％CI=0.98-1.60，7个风险估计值），KC为1.28（95％CI=1.12-1.47，5个风险估计值）。所有meta分析均存在异质性(p<0.001)。北美的风险增加比欧洲的研究更明显(分别是，OR=1.35，95%CI=1.18-1.55；OR=1.13，95%CI=1.01-1.27p异质性=0.03)。根据质量评分，没有异质性（p=0.4）。数据不足以按行业或焊接类型调查协会。排除每种癌症的发表偏倚。
结论：这项荟萃分析提示KC和BC的风险增加，但不是PC，在接触WF的工人中。不能排除其他职业和非职业危险因素的混淆。数据不足以解决特定暴露环境的风险。
方法：我们对发表在Pubmed，Scopus和Embase遵循PRISMA标准。两名研究人员选择了关于WF暴露的队列研究。从2582篇文章中，纳入7项非重叠研究。根据CASP对研究质量进行评分。我们运行随机效应荟萃分析来计算GU癌症的相对风险(RR)和95%置信区间(CI)。整体和按癌症分层，国家,和质量得分。
结果：我们纳入了7项报告GU癌症结果的研究，包括前列腺,膀胱癌和肾癌(PC，BC,和KC）。GU癌症的RR为1.19（95％CI=1.07-1.32，16个风险估计值）；PC为1.13（95％CI=0.90-1.42，4个风险估计值）；BC为1.26（95％CI=0.98-1.60，7个风险估计值），KC为1.28（95％CI=1.12-1.47，5个风险估计值）。所有meta分析均存在异质性(p<0.001)。北美的风险增加比欧洲的研究更明显(分别是，OR=1.35，95%CI=1.18-1.55；OR=1.13，95%CI=1.01-1.27p异质性=0.03)。根据质量评分，没有异质性（p=0.4）。数据不足以按行业或焊接类型调查协会。排除每种癌症的发表偏倚。
结论：这项荟萃分析提示KC和BC的风险增加，但不是PC，在接触WF的工人中。不能排除其他职业和非职业危险因素的混淆。数据不足以解决特定暴露环境的风险。
