I have a population of individuals who suffered depression over a 10 year period. I also have the mortality data and the number of times they were referred for counseling. Within this population there were multiple individuals who were lost to the study for a period of time (2-3 years). During this period of time they could not have been diagnosed with depression (under the study framework) and therefore not referred for counseling. As a consequence there is an obvious in the data. I am interested to explore if the rate of counseling impacts mortality however I am uncertain as to how to account for the time during which individuals could not have been referred (regardless as to if they were depressed or not). Apologies if this is a simple question! I am guessing it is some variation on Cox?