Oleinick A, Guire K E, Hawthorne V M, Schork M A, Gluck J V, Lee B, La S
Department of Environmental and Industrial Health, University of Michigan, Ann Arbor 48109.
Am J Ind Med. 1993 Feb;23(2):231-52. doi: 10.1002/ajim.4700230202.
National and state estimates of the severity of occupational injuries and illnesses (severity = lost work time = missed work days+restricted work days) have come from the annual Survey of Occupational Injuries and Illnesses (Survey) produced by the U.S. Bureau of Labor Statistics. However, we show that the Survey practice of collecting injury information soon after the accident year reduces substantially the accuracy of missed work day estimates, which constitute 85.3% of the Survey lost work time estimate. To develop an independent estimate of missed work days, the research team created the Michigan Comprehensive Compensable Occupational Injury Database (Michigan Database) by linking state files with injury characteristics to files with workers' compensation information for injuries occurring in 1986. The measure of missed work time (days, weeks, or years) is the cumulative duration of compensation from the "date disability commenced," noted on the first payment form, through follow-up to March 1, 1990. Cumulative missed work time has been calculated or estimated for 72,057 injured workers, more than 97% of the 73,609 Michigan workers with compensable occupational injuries in 1986 identified through the close of the study. Our "best" estimate of missed work days, to follow-up, attributable to both fatal and nonfatal compensable occupational injuries and illnesses is 7,518,784, a figure four times that reported for Michigan by the Survey. When insurance industry data on disbursements are also considered, the estimate of missed work days increases to 8,919,079, a figure 4.75 times that reported by the Survey. When insurance data on reserves for future payments are also considered, the estimate of missed work days increases to 16,103,398, a figure 8.58-fold greater than that obtained for Michigan in the Survey. The Michigan data suggest that the national Survey may have failed to identify almost 373 million of 421 million missed work days in the private sector that have resulted, or will result, from 1986 occupational injuries. The present federal/state system for estimating occupational injury severity by measuring lost work days seriously underestimates the magnitude of the problem. The current policy of obtaining incidence and severity data from the same Survey should be reconsidered. We recommend that national estimates of injury severity be obtained from representative states by using state compensation data and that such estimates be used to evaluate current prevention and rehabilitation strategies. The redesigned occupational safety and health Survey (ROSH Survey) should be revised to permit linkage to compensation data.
美国劳工统计局开展的年度职业伤害与疾病调查（以下简称“调查”）提供了全国及各州关于职业伤害与疾病严重程度（严重程度 = 损失的工作时间 = 误工天数 + 受限工作日数）的估算数据。然而，我们发现，该调查在事故年份后不久就收集伤害信息的做法，大幅降低了误工天数估算的准确性，而误工天数在调查的损失工作时间估算中占比达85.3%。为了独立估算误工天数，研究团队通过将包含伤害特征的州档案与1986年发生伤害的工人赔偿信息档案相链接，创建了密歇根综合可补偿职业伤害数据库（以下简称“密歇根数据库”）。误工时间（天数、周数或年数）的衡量标准是从第一张支付表格上注明的“残疾开始日期”起，至1990年3月1日随访期间的累计赔偿时长。已对72,057名受伤工人的累计误工时间进行了计算或估算，这一数字超过了通过该研究结束时所确定的1986年密歇根州73,609名有可补偿职业伤害工人中的97%。我们对误工天数的“最佳”估算结果显示，截至随访时，致命和非致命可补偿职业伤害与疾病导致的误工天数为7,518,784天，这一数字是该调查所报告的密歇根州误工天数的四倍。若同时考虑保险业支出数据，误工天数估算增至8,919,079天，是该调查所报告数字的4.75倍。若再考虑未来支付储备金的保险数据，误工天数估算增至16,103,398天，是该调查中密歇根州估算数字的8.58倍。密歇根州的数据表明，全国性调查可能未能识别出1986年职业伤害已导致或将会导致的私营部门4.21亿误工天数中的近3.73亿天。当前联邦/州通过衡量误工天数来估算职业伤害严重程度的系统严重低估了问题的规模。应重新考虑从同一调查中获取发病率和严重程度数据的现行政策。我们建议通过使用州赔偿数据从具有代表性的州获取伤害严重程度的全国性估算数据，并将此类估算数据用于评估当前的预防和康复策略。重新设计的职业安全与健康调查（ROSH调查）应进行修订，以允许与赔偿数据相链接。
