shows baseline covariates for our population during 2001, which represented the full-coverage period immediately before the policy change and occurred midstudy. The average age of current antidepressant users was 76; 69% were women. A total of 64% of patients had received at least one mental health prescription in the prior year, 9% had had at least one mental health specialty visit, and less than 1% had been hospitalized for mental health reasons. The subset of patients who were new users of anti-depressants during this period and had a depression diagnosis in the previous six months was similar to the full group but used fewer mental health medications and fewer medications overall.
Baseline population characteristics of seniors receiving full prescription coverage in British Columbia in 2001
From 1997 to 2001, antidepressant dispensing increased at a rate of 857 imipramine-equivalent milligrams per month per 1,000 seniors (95% confidence interval [CI]=792 to 922) (). Observed and predicted values are shown in , and model parameter estimates are reported in . The median daily dose of 45 mg of imipramine per day during this period equates to an additional 19 patient days’ worth of anti-depressant dispensed.
Antidepressants dispensed per 1,000 seniors per month in British Columbia between January 1997 and July 2006a
Number of seniors in British Columbia initiating antidepressant use per 1,000 seniors per month between July 1997 and July 2005
Model parameter estimates of antidepressant dispensing per 1,000 seniors per month in British Columbia during the move from full prescription coverage, to copay, to income-based deductible (IBD) policies between 1997 and 2005
The implementation of the copayment policy in January 2002 was associated with a drop in dispensing by 1,910 mg per month (CI=–7,111 to 3,290; 42 patient-day decrease in days’ supply per month), but the growth rate of dispensing increased by 375 mg per month after the policy change (CI= −109.7 to 860.2; change in slope=decrease of eight patient-days per month). The subsequent implementation of the income-based deductible policy resulted in a nonsignificant decrease in the dispensing level and a significant decrease in the rate of dispensing growth by 626 mg per 1,000 seniors per month (CI=−1,132.8 to −119.3; change in slope=decrease of 14 patient-days per month). When the income-based deductible period was compared directly with the baseline period, there was no change in dispensing level, but dispensing growth slowed by 283 imipramine-equivalent milligrams per month (CI=−466.3 to −99.6).
The use of selective serotonin reuptake inhibitors (SSRIs) increased substantially over the study period. SSRIs accounted for 42% of imipramine-equivalent milligrams dispensed in January 1997, with this proportion increasing to 63% by December 2005. Use of new and second-generation agents as a proportion of total use also increased, from 11% to 22%, whereas tertiary amine use as a proportion of total use fell from 37% to 15%. Within the SSRI class, citalopram market share increased rapidly from the medication’s first use in April 1999 to December 2005, when it accounted for 50% of total SSRI use. [A list of anti-depressants included in the evaluation is provided as an online supplement to this article at ps.psychiatryonline.org.]
Rates of antidepressant initiation are shown in . Antidepressant initiation increased from 4.3 starts per 1,000 per month in January 1997 to 5.0 starts per 1,000 per month in December 2001. The implementation of the copay policy was associated with a significant .38 per 1,000 drop in initiation level but no change in the rate of increase over time (). Growth in initiation rates slowed minimally by .03 starts per month (CI=−.066 to .004) with the introduction of the income-based deductible and coinsurance policy. Relative to the baseline period, the initiation level was reduced by .3 starts per 1,000 seniors (CI=−.55 to −.06), and the growth in initiation rates slowed by −.028 per 1,000 seniors (CI=−.038 to −.018).
Among patients with a recorded depression diagnosis in the prior six months and no use of antidepressants during that period, the baseline frequency of antidepressant initiation was 315 starts per 1,000 seniors in January 1998, which increased by .8 starts per 1,000 seniors per month over time (CI=.64 to .97) (). [A figure showing the number initiating antidepressants per 1,000 seniors per month with a recorded depression diagnosis in the previous six months is provided in an online supplement to this article at ps.psychiatryonline.org.] The copay policy resulted in a significant drop of 12.3 starts per 1,000 seniors (CI=−21.9 to −2.7). The income-based deductible policy resulted in a significant trend toward decrease (−1.2 starts per 1,000 seniors per month, CI=−2.2 to −.3). A similar decreasing trend and no change in number of starts per 1,000 seniors with depression were observed when comparing the coinsurance period directly with the baseline period.
At baseline 13% of antidepressant users stopped their medication each month. Rates of stopping decreased by .03% per month over time (−.06% to −.01%) ( and ). The implementation of the copayment and income-based deductible policies did not have a significant effect on stopping rates at the population level.
Proportion of depressed seniors in British Columbia who discontinued their antidepressant medications in each month between January 1998 and July 2005