As the Affordable Care Act passes through more phases of implementation, it sets off waves of estimates of its impact on workers’ compensation. The most popular contest among estimators is guessing if universal health insurance will result in fewer workers' compensation claims.

I went back to some earlier analysis I did of workers’ comp-claiming behavior. Myriad studies, conducted by independent researchers and government agencies in the past 20 or so years, mostly suggest that many workers with work-related injuries or illnesses do not file workers’ comp claims. 

How big is the pre-ACA era claiming shortfall? The answer is, a lot. The scale of historic underclaiming appears to be so large that any perturbations caused by the ACA could easily be buried and unobserved under preexisting patterns of nonclaiming.

Leslie Boden and Al Ozonoff, both at Boston University School of Public Health and with a long record of studying the issue, wrote a report for California in 2008 in which they estimated that from 21% of lost-time injuries to 40% of lost-time injuries fail to be recorded on OSHA-mandated logs.

These researchers, as many others have, looked at what can be called OSHA undercounting. That undercount and nonclaiming patterns are distinct but influence each other, as if in the same force field. 

I estimated in 2010 that roughly 40% of work injuries and illnesses do not result in a workers’ comp claim. That might amount to well over 1 million claims that never were submitted, each year.

The weight of the evidence provided through some 20 studies suggests there is a moderate to serious problem with nonclaiming of acute injuries and some nonfatal diseases (such as asthma) and a significant problem in failing to capture potentially fatal diseases (such as cancers).

When employees with a work injury or illness fail to claim workers’ compensation benefits, they forgo the protections of the system. In addition, the costs of medical care and disability are shifted to the injured worker’s household and/or other public and private benefit programs. The workers’ comp system may be dodging $10 billion or more per year due to cost shifting.

My estimate deserves challenges as to its logic and to the factual evidence. 

If a claim was not filed, who is to say that the injury would have been compensable? Most accepted workers’ compensation claims in the United States are reviewed by a physician who determines if occupational exposure is a materially contributing cause to the medical condition. For these nonclaimed injuries, there is no such review. Estimates of nonclaiming typically assume that injuries are compensable.

The factual evidence for nonclaimed injuries arrives through several distinct methods, of which three are described below.

(1) Household surveys of workers. The Integrated Benefits Institute analyzed household responses from the 2002 and 2006 versions of the General Social Survey. In the surveys, 8% of employed adults reported being injured on the job at least once in the previous year, in both 2002 and 2006. This frequency was higher than the injury rates reported by the Bureau of Labor Statistics.

The Medical Expenditure Panel, an annual national survey, records health care utilization and source of payment for respondents. It includes occupational injury as a descriptor of a medical problem. Researchers reported that, among Hispanic construction workers, workers’ comp paid for only 27% of work injury-related medical costs, private health insurers paid 46%, 23% was paid out of pocket, and 4% came from other resources. White construction workers ran 50% of their work injury-related medical costs through workers’ comp.

(2) Medical records and payment databases. Studies of health maintenance organizations (HMO) medical records indicate that treatment of many suspected occupational asthma conditions might have been paid for by commercial health insurers. One study estimated that 33% of HMO asthma cases were work-related. In another study, 24% of cases were estimated to be work related.

(3) Capture/recapture analysis. This method compares two or more ideally independent sets of data about the same subject (for instance, fish in a lake). By applying this method to parallel sources of injury data, researchers can estimate not only the total number of injuries, but also the size of the undercount for each of the sources and the number of injuries not captured by either source. A six-state study by Boden and colleagues using this method estimated that 10% of all injuries had been omitted from both data sources. In the median state, 39% of work injuries failed to be entered in the workers’ compensation system, with the range from a high of nonclaiming of 55% to a low of 15%. OSHA logs were estimated to suffer from an even more significant undercount, ranging from 63% to 29%.

I concluded that 40% of compensable injuries were not reported. Another reader of the same studies might reasonably say 25% or 50%. Unfortunately missing from most of the studies is a distinction between light, moderate and serious injuries. One would expect (and hope) that nonclaiming rates decline as severity of injury increases.

Why do workers not file claims? The answer varies according to the severity of injury, the state of labor-management relations, worker knowledge of her or his benefits and other factors. That’s a subject for another article.

Peter Rousmaniere is a consultant for workers' compensation claims administrators and vendors and a veteran observer of workers' compensation industry trends.