Click on image to view larger version.

Figure 1 Study design for each phase. Phase I (top) was a crossover design (n = 248); Phase II (bottom) was a parallel design (n = 333). CRT = cardiac resynchronization therapy; Rand = randomization time.