The significance of two normally independent findings in a single patient is often difficult to assess. One simple measure of the likelihood of random coincidence is the joint probability, based on the assumption that the findings are independent. Unfortunately, this technique is often used wrongly.…