Clinical science is more "bedside" than bench, e.g. what effect does this experimental drug have when we give it to people? How do we evaluate the effect of the drug, how do we assess side effects and determine is adverse events are due to the drug, the disease we're treating, or just random events related to neither the disease nor the drug?
Docs (without PhDs) but with research-intensive fellowship training (sub-specialty training) might also do bench research. I suspect that this is particularly common in the surgical specialities where some techniques are tried in animals before being used in humans.