Dialysate leaks from the exit site can heal spontaneously in up to 50% of cases, particularly if they occur early (within the first month of catheter insertion) and if the PD can be safely interrupted for a period of time by relying on residual renal function or transferring temporarily to HD (at least 48 hours, but preferably 1-2 weeks). If the PD is continued, the chance of spontaneous cessation of the leak is low.

I am presuming that the fluid discharge from the exit site has been confirmed to be dialysate (ie high glucose concetration) rather than serum.