Work-Related Amputations: Who’s Counting?

Knowing how many, who and where injuries or disease are occurring is a basic premise of preventing injuries and illnesses. If we don’t have accurate information on injury/illness occurrence, we don’t know how many resources to devote, what action(s) to take or whether the action we do take is effective.

New findings from Michigan State University and the Michigan Department of Community Health on work-related amputations, published in the Journal of Occupational and Environmental Medicine, is the latest in a series of efforts to find better ways to identify work-related injuries and use that information to prevent similar injuries from happening in the future. In this latest work we identified 616 work-related amputations in Michigan, which was two and a half times more amputations than identified by the national system for tracking workplace injury (616 vs. 250). Not only can we identify more amputations but, unlike the Federal system, the information can be used for efforts to prevent such terrible injuries in the future, including use by OSHA to conduct enforcement inspections at the facilities where the amputations occurred.

How is it possible that such an obvious injury is missed in the official statistics? When we in our program first began our work we assumed like many others that although the official statistics missed many chronic diseases such as lung cancer from asbestos, which the employer may not be aware of  because it occurred years after a person retired, that the current national  system worked for an acute obvious injury such as an amputation. The U.S. system for tracking work-related injuries and illnesses is based on reporting by a sample of employers selected by the Bureau of Labor Statistics (BLS). Selected employers receive a survey on which they are required to report injuries and illness recorded on their OSHA log. In contrast, our work in Michigan is based on collecting medical records from hospitals, emergency departments and clinics. We perform a census, not a sample, and our reports are not dependent on an employer knowing about the injury, recording the injury, accurately completing the OSHA log and being sampled by BLS.  If a doctor states an individual has had amputation, then those records are reported and reviewed.

A multi-source system such as ours has other advantages.  As stated above, for example, our data can be used by OSHA for enforcement.   The names of employers in the BLS survey are kept confidential. This means that OSHA can use the information to identify trends and high risk industrial sectors but cannot use the BLS information to conduct an inspection at the facility where the amputation occurred. Additionally, since BLS does not collect information on race, they cannot address the issue of health disparities.

In retrospect, we should not have been surprised by our findings.  In a 1988  National Academy of Medicine review of the BLS employer based survey, the NAS funded two studies, one in New Jersey and one in Texas that found that the BLS employer based survey missed 50% of work-related acute traumatic fatalities. In response to the undercounting reported in these studies, BLS initiated a multi data source system that includes death certificates, medical examiner reports, police reports and newspaper clippings. After the new system was implemented, the number of acute traumatic work-related deaths doubled in the United States — not because of a true increase in the number of deaths but because now we had accurate counts. A multi-data source system was not instituted for non-fatal injuries.

The advantages of a multi-data source surveillance system to track work-related injuries and illness are obvious and needed across the country.  We in our program will continue multi-data source tracking of work-related injuries and illnesses in Michigan to generate accurate information, examine health disparities and to initiate interventions. Last year we reported that there were 1,461 work-related burns in Michigan which was three times the official BLS statistics (1,461 vs. 450) (Journal of Occupational and Environmental Medicine 2012; 54:642-647). We are now in the process of finalizing data on work-related skull fractures.

Do you think a similar system would be useful in your state?  If you are interested in setting up a similar multi-data source surveillance system in your state we in our program would be happy to provide you with more information (rosenman@msu.edu).  See our website at www.oem.msu.edu for information about the conditions being tracked in Michigan.

Kenneth D. Rosenman, M.D.

Dr. Rosenman is a Professor of Medicine and Chief of the Division of Occupational and Environmental Medicine at Michigan State University. 

This blog is the first in what we hope will be a series of blogs from the states offering their unique perspective on occupational safety and health issues.