The study cohort was identified during a 3-month
period. Teens were included if they had any diagnosiscode (ICD-9) or procedural code (CPT) indicating sexual
activity. These included ICD-9 or CPT codes for
an abnormal Pap smear, a laboratory test or diagnosis
code for a sexually transmitted infection, a pregnancyrelated
event such as an ectopic pregnancy or miscarriage,
or evidence of contraceptive use (Table 1). Since
our outcome measure was compliance with contraceptives,
we needed to exclude teens that were or became
pregnant. In a subsequent 3-month period or “washout”
we examined claims to exclude teens with an
ICD-9 code indicating pregnancy (Table 1). The following
12 months were the study interval to examine
contraceptive utilization and compliance (Figure 1).