Oleinick A, Gluck J V, Guire K E
Department of Environmental and Industrial Health, School of Public Health, University of Michigan Ann Arbor 48109-2029, USA.
Am J Ind Med. 1996 Nov;30(5):556-68. doi: 10.1002/(SICI)1097-0274(199611)30:5<556::AID-AJIM4>3.0.CO;2-R.
The current BLS Annual Survey of Occupational Illnesses and Injuries and several recent analyses of factors affecting missed worktime in occupational back injuries rely on ANSI-based injury codes derived from injury narratives to classify occupational injuries and estimate incidence and outcome. No population-based studies of the concordance between back injury codes and clinical diagnoses have been reported. Back injury cases were identified in two large work-injured populations totaling almost 80,000 cases in the states of Michigan and Minnesota. In both populations, cases had been coded by the single nature-of-injury and part-of-body-injured codes assigned by an ANSI-based injury-coding system and by as many as four (Michigan) or three (Minnesota) clinical diagnoses according to the International Classification of Diseases-Clinical Modification, 9th Revision. Concordance was measured by the sensitivity and predictive value positive (PVP, aka PV+ or PPA) of the injury coding scheme for related diagnostic groups. We also used an algorithm based on the limited clinical information available to corroborate the diagnosis of displaced/herniated disc for cases that underwent spinal surgery. Cases identified by the algorithm were then used to obtain a lower bound estimate of the fraction with disc injury. The injury coding scheme had PVPs of 82.9-90.1% and overall sensitivities of 69.7-75.9%. Sensitivities for individual diagnostic groups show that their distribution in ANSI-coded injury groups is skewed slightly toward cases with sprain and disc displacement/herniation, but these shifts are modest. The lower bound estimate of the fraction of cases with disc displacement/herniation in a population of cases with back injuries producing at least 1 day of missed worktime is 5.8%. The demographic comparisons indicate that, as the time between injury and cohort ascertainment increases during the first 8 days of missed worktime following injury, the proportion of younger workers in an injury cohort decreases. The relationship between increasing age and increasing missed worktime disability, reported in various outcome studies, is also present during the first few days following injury. The use of ANSI injury codes underestimates the contribution of back injuries to missed worktime because 24-30% of cases are missed by the ANSI coding system. However, the distribution of diagnostic groups in the injury-coded groups approximates that observed with all diagnosed cases and supports the use of such data to study outcome. Our estimate, and one from Quebec, suggest that disc displacement/herniation occurs more frequently in the subset of occupational back injuries compared to the set of back injuries from all sources.
美国劳工统计局当前的职业疾病和伤害年度调查以及最近几项关于影响职业性背部损伤缺勤时间因素的分析，均依赖于基于美国国家标准学会（ANSI）的损伤编码（这些编码源自损伤描述）来对职业损伤进行分类，并估计发病率和结果。目前尚无基于人群的关于背部损伤编码与临床诊断一致性的研究报告。在密歇根州和明尼苏达州的两个总计近80,000例的大型工伤人群中识别出背部损伤病例。在这两个人群中，病例均已根据基于ANSI的损伤编码系统所指定的单一损伤性质和身体受伤部位编码，以及按照《国际疾病分类 - 临床修订版，第9版》进行的多达四种（密歇根州）或三种（明尼苏达州）临床诊断进行了编码。通过损伤编码方案对相关诊断组的敏感性和阳性预测值（PVP，又名PV + 或PPA）来衡量一致性。我们还使用了一种基于可用有限临床信息的算法，以证实接受脊柱手术病例的椎间盘移位/突出的诊断。然后使用该算法识别出的病例来获得椎间盘损伤病例比例的下限估计值。损伤编码方案的阳性预测值为82.9 - 90.1%，总体敏感性为69.7 - 75.9%。各个诊断组的敏感性表明，它们在ANSI编码损伤组中的分布略微偏向于扭伤以及椎间盘移位/突出的病例，但这些偏移幅度不大。在导致至少1天缺勤的背部损伤病例人群中，椎间盘移位/突出病例比例的下限估计值为5.8%。人口统计学比较表明，在受伤后缺勤的前8天内，随着受伤与队列确定之间时间的增加，损伤队列中年轻工人的比例会下降。在受伤后的头几天也存在各种结果研究中所报告的年龄增长与缺勤工作残疾增加之间的关系。使用ANSI损伤编码会低估背部损伤对缺勤时间的影响，因为ANSI编码系统会遗漏24 - 30%的病例。然而，损伤编码组中诊断组的分布与所有确诊病例中观察到的分布相近，并支持使用此类数据来研究结果。我们的估计以及来自魁北克的一项估计表明，与所有来源的背部损伤相比，椎间盘移位/突出在职业性背部损伤子集中更为常见。
