The author shouldn't just be considering his individual risk level, though, which is the parent comment's point. Even if the individual's risk is extremely low, the expected negative societal impact can still be high given a) many other people are also taking the same risk at the same time, b) the system has a non-linear "breaking point" / constraint (e.g., a limited supply of hospital resources that are all near capacity), and c) crossing the breaking point causes major problems for people in and out of the system (i.e., doctors, people with QOL procedures that have to be cancelled due to overflow, etc.).