Anchoring bias is a well known cognitive bias in medicine. The term refers to overemphasizing early-acquired information. One example which occurs in medicine that looks like this: an obese patient sees his primary care provider for a rash; the doctor diagnosis a rash and prescribes a topical ointment with instructions to return if the symptoms persist. They do, and the patient sees another provider at the same office, complaining of joint pain. The second provider suspects it is related to his obesity. Three weeks later the patient is seen by a rheumatologist at another clinic, and is diagnosed with lyme disease: the rash was related to the tick bite; the joint pain was lyme arthritis. This is anchoring bias – it’s not for boats.

Practically speaking, if you are getting a second opinion in a case, it’s a good idea to not tell the second provider what the first diagnosis was. The reason there is that the earlier diagnosis will immediately jump to the top of the second provider’s differential diagnosis.

In a potential medical malpractice case, if both providers independently come to the same diagnosis that is later discovered to have been wrong, you will likely have a hard time finding an attorney to take your case. Because the otherwise erroneous conclusion looks reasonable when the same conclusion is independently reached by two providers.

Why is anchoring bias a problem, aside from the obvious example above?

We try to account for cognitive biases like anchoring bias in consulting with expert witnesses by trying to create a “blind read” by an expert. This means we will reach out to an expert via an anonymous email address, and not identify whom we represent. In radiology cases we will send a copy of the imaging studies, but not imaging report. In other cases, like an emergency room case, we may edit out the diagnoses reached by the prospective defendant, while keeping in all of the exam and test results. The idea is that the expert may be able to develop his or her own opinion regarding a correct diagnosis without the anchoring of the original diagnosis. Further, if there really is no case for liability to be made, that is a pretty good way to quickly reach that conclusion.

These cases are expensive to develop and risky to try. If we can get good experts supporting a case for negligence, then we are off. If, though, we can learn early on that there is no negligence case to be made, that will save our clients a lot of heartache.