Estimation of intervention effects using recurrent event time data in the presence of event dependence and a cured fraction