Design limitations to bipolar II treatment efficacy studies: A challenge and a revisionist strategy.