Is this the same as the confidence interval listed for the intercept in parameter estimates?
slope and intercept are different (independent) parameter of a model, so they will have a different confidence interval