Steenland K, Burnett C, Osorio A M
Am J Epidemiol. 1987 Aug;126(2):247-57. doi: 10.1093/aje/126.2.247.
Commercial city directories, currently produced in 1,250 United States cities, potentially provide yearly information on occupation and employer for all city residents over age 18 years. To investigate the usefulness of these occupational data, the authors have conducted a case-control study of male bladder cancer mortality in Hamilton County, Ohio (which includes Cincinnati). A total of 731 bladder cancer cases who died during 1960-1982 were matched on age, sex, race, date-of-death, and residence at death to two controls per case. Risks of bladder cancer death were calculated by occupation, industry, and specific employer, using both city directories (multiple statements) and death certificates (single statement). Four companies showed a significant excess bladder cancer risk when using city directories. Only one would have been identified using death certificates, which ask for usual lifetime type of industry rather than a specific company name. Using city directories, significant positive associations were found between bladder cancer and occupation as an engineer, tailor, carpenter, furnace operator, blending machine operator, chemist, pressing machine operator, house cleaner, or salesman. For industry, the authors found significant positive associations for the textile, chemical, grain mill, foundry, petroleum, building service, entertainment, and advertising industries. A significant increase in risk for those with 20 or more years of employment was seen for those employed as truck drivers and furnace operators, or those employed in the railroad industry. A check of the validity of city directory data indicated that 77 per cent of the listings agreed with Social Security earnings reports for employer in any given year. One limitation of Hamilton County city directory data was the fairly large number of yearly listings without any occupational data (15 per cent for occupation, 36 per cent for employer). While city directory data do provide work history over time, unlike death certificates, such data are available only for years of residence in the city in question.
目前在美国1250个城市编制的商业城市名录，有可能提供18岁以上所有城市居民的职业和雇主年度信息。为了调查这些职业数据的实用性，作者对俄亥俄州汉密尔顿县（包括辛辛那提）男性膀胱癌死亡率进行了一项病例对照研究。1960年至1982年期间死亡的731例膀胱癌病例，按照年龄、性别、种族、死亡日期和死亡时的居住地，以1:2的比例与对照进行匹配。使用城市名录（多条记录）和死亡证明（单条记录），按职业、行业和具体雇主计算膀胱癌死亡风险。使用城市名录时，有四家公司显示出显著的膀胱癌风险过高。使用死亡证明时，只会识别出其中一家，因为死亡证明要求填写通常的终身行业类型，而不是具体的公司名称。使用城市名录时，发现膀胱癌与工程师、裁缝、木匠、熔炉操作员、搅拌机操作员、化学家、压机操作员、房屋清洁工或推销员等职业之间存在显著的正相关。对于行业，作者发现纺织、化工、谷物磨坊、铸造、石油、建筑服务、娱乐和广告行业存在显著的正相关。受雇为卡车司机和熔炉操作员或在铁路行业工作的人，如果工作20年或更长时间，风险会显著增加。对城市名录数据有效性的检查表明，在任何给定年份，77%的记录与社会保障收入报告中的雇主信息一致。汉密尔顿县城市名录数据的一个局限性是，每年相当多的记录没有任何职业数据（职业方面为15%，雇主方面为36%）。虽然城市名录数据确实提供了一段时间内的工作经历，但与死亡证明不同，此类数据仅适用于在所涉城市居住的年份。
