Brett Bouchard, 17, was working at Violi’s Restaurant in Massena, NY last month. Press reports indicate he was cleaning out a pasta-making machine when the equipment severed his right arm at the elbow. He was rushed to a local hospital which later transferred the young worker to Massachusetts General Hospital.
Nationally, there are thousands of work-related amputations each year. The Bureau of Labor Statistics (BLS) estimates that 5,100 U.S. workers suffered an amputation injury in 2012. But we’ve written here before about the limitations of BLS’ estimates for work-related injuries. Those limitations include that its “Survey of Occupational Injuries and Illnesses” (SOII) relies exclusively on self-reported data by employers and some categories of workers are not covered by the reporting requirement (e.g., individuals working on small farms and some government employees.) These limitations and others result in an undercount by BLS of the true magnitude of work-related amputations (and, for that matter, all other injuries.)
A paper by researchers at Michigan State University, for example, calculated that the BLS system misses about 70 percent of work-related burns in their state. The researchers also looked at amputations and they identified more than twice the number than the BLS’ figure. Other limitations in BLS's estimate, some of which also relate to the self-reporting component of the survey, are employers falsifying their records, and some not understanding what is supposed to be recorded. In addition, some injuries may not be included on employers’ injury logs because workers are reluctant and/or discouraged from reporting injuries to their bosses. Others who’ve compared BLS’ annual estimates to alternative data sources (e.g., hospital records, workers compensation claims) also find that the government-reported figures substantially understate the number of work-related injury and illness case (e.g., here, here, here, here, here, here, here.)
In recent years, BLS has been more willing to acknowledge the limitations of its estimates of work-related injuries. Attention from Members of Congress compelled BLS to provide grants to several States and university-based researchers to confirm and quantify the undercount. Three papers funded by those grants have been published recently in the American Journal of Industrial Medicine (AJIM). One written by Les Boden, PhD of Boston University School of Public Health, another by Sara Wuellner, MPH and David Bonauto, MD, MPH of the Washington State Department of Labor and Industries, and the third by a collaboration of researchers from the Massachusetts Department of Public Health (MDPH), UCLA, Boston Children’s Hospital, and Boston University School of Public Health. Leticia Davis, ScD, EdM, director of the MDPH’s Occupational Health Surveillance Program is the lead author.
This third paper addresses work-related amputations in Massachusetts. Davis and colleagues set out to calculate the magnitude of the undercount of amputations reported in BLS’s SOII estimate. They linked and compared the number of amputations cases in the BLS system to hospital and workers’ compensation records. For the period 2007-2008, using the self-reported employer data, BLS estimated there were 210 cases of work-related amputations in Massachusetts. Using the multi-source dataset, the authors identified 787 cases for the same time period. Davis and colleagues report that 41 percent of the actual amputation cases in SOII were not classified properly by the employers as amputations. Instead they were reported as “other injuries,” such as cuts, lacerations or crushings. We can speculate on the reasons for those errors, such as employers trying to misrepresent the severity of the employees’ injuries, or needing better instructions on how to classify injuries on their OSHA logs.
The authors classified each of the 787 case as “SOII eligible” (n=406), “SOII ineligible” (n=96), and “unknown SOII eligibility” (n=285.) They calculated the ratio of the multisource “SOII eligible” cases to BLS’s published estimates. They reported a 48 percent minimum estimate of the undercount.
Davis and colleagues reiterate, as other researchers have also done, that a multi-source surveillance system would provide much better data than BLS’s current system. They note, however, the practical challenges in attempting to design and implement such a system nationwide. Each state, for example, has different workers’ compensation laws and data collection systems, and policies regarding release of hospital records.
BLS called on researchers (and provided funding) to confirm the existence of an undercount and quantify its magnitude. Three of those studies are now published in AJIM. BLS is also sponsoring interviews with employers to learn their practices for recording work-related injuries on OSHA logs, and a study of historical trends in the undercount comparing SOII and workers' compensation data. The results of these assessments will likely appear also in the literature.
In last year’s news release announcing the most current SOII data, BLS wrote:
"Several studies by outside researchers conducted in the mid 2000's questioned the completeness of BLS injury and illness estimates from the SOII.
And added that the agency:
"established an ongoing research program to explore potential undercounting of workplace injuries and illnesses."
At some point in the next few years, I hope we'll see BLS do something else in addition to funding studies. When the agency releases its SOII data, BLS should mention the conclusions of these studies. I’d suggest, for example, something like the following for the paper by Davis and her colleagues:
"a study examining work-related amputations in Massachusetts found that SOII underestimated by at least 48 percent the number of amputation cases."
That would be a further step to inform the public and policy-makers on the validity of SOII with respect to the true incidence of work-related injuries.
Seventeen year-old Brett Bouchard’s disabling injury is a reminder that thousands of work-related amputations occur in the U.S. every year. With better data on the magnitude of the problem, there might be more interest in preventing them.