Covariate‐adjusted borrowing of historical control data in randomized clinical trials