@Tyersome's answer is correct if you consider that the only mutations that occur will result in substitutions. This is false for biology, if true for how the subject is commonly talked about. (I realize that I'm in the uncomfortable position of saying that Rich Lenski is being imprecise about his evolutionary terminology). In fact probably a rather larger proportion of mutations will result in other kinds of DNA modifications.
For those unfamiliar with the different kinds of mutations, substitutions (also called point mutations) are only one class of mutation. Other kinds include transpositions, inversions, insertions, deletions, and many many more.
In other words, the "mutation rate" is probably >2X higher than the "substitution mutation rate", and I suppose since I'm getting very involved with my terminology, we could then say that there is an "accepted substitution rate" (the same as the "substitution rate" that people talk about) and an "accepted mutation rate" which includes all substitutions but then also all other kinds of mutations that are incorporated as laid out in the quote above. (I have made up all the terms other than "substitution rate" and "mutation rate" to emphasize that I think the usage of terminology from the Barrick and Lenski article is extremely misleading).
We tend to focus on substitutions because they are easy to understand and our math works well for them. Thus when people write about the "mutation" rate they have a tendency to use it to mean the "mutations that lead to substitutions" rate. But that's an approximation for mathematical convenience rather than reflecting real biology.
It's good to recall that when Hugo de Vries wrote "The Mutation Theory" >100 years ago the mutations he was talking about were not substitutions but changes in ploidy- those are absolutely mutations, but they have nothing to do with substitution. In plants changes in ploidy are among the most phenotypically dramatic class of mutation, but those are invisible to analysis of substitutions.
Of course, the approximation for mathematical convenience has more or less taken over the field, so from that perspective @Tyersome's answer is perfectly correct, and it does directly address your specific question. But I think that there are problems with how the field in general thinks about this.
(Forgive me for being pedantic, I wrote a review on the topic: https://www.cell.com/trends/genetics/fulltext/S0168-9525(19)30012-5)