Figure 1.

Study design. Four cardiometabolic user cohorts were analyzed by restricting the follow-up time only to periods when that specific drug category was used (all other time periods were excluded from analyses and not analyzed), to statin use in this example. Each statin use period ended either to discontinuation (indicated with arrow) or censoring. Statin use periods were divided into use and non-use (“NO AP”) of antipsychotics. Specific antipsychotics and antipsychotic use further to use of specific monotherapies (“OLA” as olanzapine, “CLZ” as clozapine and “RIS” as risperidone) in this example. Concomitant use of 2 or more antipsychotics was considered as antipsychotic polytherapy (“POLY”). First outcome event is assigned to no use of antipsychotics and the second to risperidone use.