.. some extra comments to hammer a few points home ..
(a) Vmca/Vmc in the AFM is the certification figure - the "real world" figure on the day for a given configuration should be a little lower
(b) Inertial effects are therefore usually irrelevant.
Vmca generally is worked up progressively to determine the static value (a bit like the typical pilot Vmca demonstration thing). However, once the TP folk have defined what they reckon is the static value, the dynamic value is determined with an engine failure. This may or may not cause the first value to be revised to end up with the book figure
(c) VMCA must be demonstrated/determined within a constraint of not using more than 5 degrees of bank.
Unless the book says something different, presume that the value is with 5° bank into the live engine(s). More importantly the variation of Vmca with variation of bank is SIGNIFICANT - if you are near to the book figure and playing engine failures, make sure that you crank in the bank .. lest you find the Vmca for your actual bank to be something faster than your actual speed. The 5° limit is to prevent folk cheating and ending up with an unrealistically low Vmca by using excessive bank angles.
As I recall, for a well-known many-engined bomber, the Vmca goes up something like 40-45 kt if the aircraft is banked the "wrong" way ... (or was it 30-35 kt ? ... either way, a very significant increase),
Also, the 5° figure is linked with Vmca handling, not OEI climb performance. Generally the optimum bank for OEI climb is around 2-3° and the performance wings level is much the same as that for 5° into the live engine(s). As a consequence, unless you really need that last bit of performance, the usual OEI climb technique is wings level.
The oft cited "requirement" for 5° bank during OEI climb (which one sees in every second GA operations manual) is tied up with the Industry's wealth of OWT rather than any objective engineering validity.