"We would have won if the ALJ wasn't so pro-employee," is a common refrain in the worker's compensation world. Undoubtedly the statement is at least partly true. However, there may be more going on than what at first meets the eye. Human beings evaluate the world using various heuristics. A heuristic is an experienced-based technique for problem-solving, learning, and discovery. Common heuristics include trial and error, rules of thumb, educated guesses, common sense, and stereotyping. Heuristics are useful because they allow us to take shortcuts when making decisions involving complicated questions. Unfortunately, heuristics often result in decisions and assumptions that are not optimal. It is critical to understand this behavioral condition when assessing why judges reach the decisions they do.One of the more common heuristics and the most vexing in the medico-legal world is the association of correlation with causation. The most frequent example involves judges and juries assigning causation because of temporal proximity. Assigning causation based on temporal proximity is a logical fallacy that, in the Latin, is called cum hoc ergo propter hoc ("with this, therefore because of this"). The logical or cognitive error is considering two independent events to be causally related because they occurred at or about the same time. In the claims arena, this frequently involves an employee who develops back pain at work without a precipitating event then concludes, with medical support, that because the back pain arose at work the work activities must have caused the back pain. Although we as humans have a natural bias toward explaining the world as a series of cause and effect relationships, imposing this bias on independent events simply because they occurred near in time erroneously establishes causation where none in fact exists.A well-known example demonstrating why the "correlation is causation" assumption is false involved claims that hormone replacement therapy (HRT) in post-menopausal women reduced heart disease. Initial reports indicated that women receiving HRT had a lower incidence of heart disease than women who did not receive HRT. However, after a large, randomized study specifically designed to test the effects of HRT in post-menopausal women, it was determined that HRT slightly increased the risk of developing heart disease in post-menopausal women. The reason for the counterintuitive result was the failure to account for confounding variables differentiating women receiving HRT versus those who did not. Put shortly, women receiving HRT were generally healthier to begin with than those women not receiving HRT. Once the confounding variables were controlled for, it was discovered that HRT had a slightly negative effect on cardiovascular disease rates. The HRT example effectively demonstrates that simply because two things are present at the same time does not mean they are related.The difficulty for claims professionals and attorneys is that they must find a way to convince judges and juries that the commonsense assumption that complaints arising during or shortly after a particular activity must be related is almost certainly false without sounding condescending. In the back pain example, most IME reports do a good job of explaining that the diagnostic imaging findings look more like the degenerative changes from aging than the changes from acute injury. However, this alone is usually insufficient to convince judges and juries that the experience of pain or disability during an activity or incident is not related to the activity or incident. Instead, the IME doctor will need to offer an analogy that judges and juries will understand. For example, we call rhinovirus infections "colds" because we used to believe that cold temperatures caused the ailment. We now recognize that "colds" are caused by viruses and that being cold has nothing to do with becoming infected. I think it is reasonable to ask independent medical examiners to use effective analogies to make it clear that medical science should trump assumptions about correlation and causation that are unreliable and probably false.
log in to commentBack to Blog