Final State Interactions and epsilon'/epsilon: A Critical Look

We critically analyze recent attempts to include final state interaction (FSI) effects in the calculation of the CP violating ratio epsilon'/epsilon. In particular the approach of using the Muskhelishvili-Omnes equation for the decay amplitudes is examined. We find that the values of the dispersive correction factors are very sensitive to initial conditions which are, at present, out of control, and demonstrate that the claimed large enhancement of epsilon'/epsilon through FSI is questionable. We propose, instead, a different implementation of this approach which may be useful to lattice determinations of weak amplitudes.