A Model Is a Model: A Decision Analysis for Rectal Cancer

Extract

Many years ago, when I was a medical intern, I had a patient who presented with a very early stage breast cancer. The resident wished to get a bone scan on the patient. Given the very low risk of a positive test, I inquired whether this was really necessary. His response was, “Well, what if it is positive?” Although not usually stated that blatantly, many of the decisions we as clinicians make are often based on a similar low level of clinical insight into the possible benefits of a procedure. Clinicians are routinely faced with the question of how much should be done to a cancer patient. The question is faced at all phases of patient management—with components of a physical examination, treatment, or, as discussed by Telford et al. (1) in this issue of the Journal, initial evaluation of the patient with cancer. Unfortunately, there are often minimal data to provide objective guidance in the initial evaluation. More diagnostic studies clearly supply more information. However, sometimes this additional information is not useful, and sometimes it is downright confusing. More information may lead to more tests and greater uncertainty for the clinician regarding proper patient management.