stratified cox (PWP-TT) with interactions

Question

im modeling recurrent events (hospital readmission) and i have some trouble with my pwp-tt model. My objective is to compare HR between strata. I have CP data format, it looks like this:

id	start_date	stop_date	terap_dis	event	tstart	tstop	status
1	16078	16377	1	1	0	299	1
1	18019	18213	0	2	299	2135	0
2	15719	16037	0	1	0	318	1
2	16049	16129	0	2	318	410	1
2	16212	16275	0	3	410	556	0

terap_dis is a binary variable that is 1 when a pacient reach a therapeutic discharge, event is the stratification variable, and status is the censoring variable (1 when and individual have a subsequent event.

coxph(Surv(tstart,tstop,status==1) ~ terap_dis+ terap_dis:strata(event)+cluster(id)+strata(event), data = data)

Model results

I think if i want to know the HR for terap_dis = 1 vs terap_dis = 0 for the second strata, i should calculate exp(-1.03320+2*0.74589) = 1.71. Is this approach correct? because it dont make sense that a pacient have and increased risk of readmission if they reach therapeutic discharge. One posibility is that previous admissions have greater risk of readmissions. I'm very confused.

Its there a way to plot survival and hazard plot for diferent strata? Pd: This is just a simplified model, because i have more covariates, but the effect is the same.

Thank you very much for your time