Background: It is widely believed that most antidepressant medications exhibit a delay of 2-4 weeks before clinically relevant improvement can be observed among patients. During this latency period, patients continue to be symptomatic and functionally impaired. Thus, time to onset of effect is an important attribute of a new pharmacotherapy. We assessed the onset of effect for duloxetine, utilizing analytical methods previously recommended in the literature.
Method: Efficacy data were pooled from two identical, but independent, randomized, double-blind, placebo-controlled, 9-week clinical trials of duloxetine (60 mg QD). Efficacy measures included the 17-item Hamilton Rating Scale for Depression (HAMD(17)), HAMD(17) subscales (Maier, core, and anxiety), and the Clinical Global Impression of Severity (CGI-S) and Patient Global Impression of Improvement (PGI-I) scales. In each individual study, duloxetine demonstrated statistically significant advantages over placebo on multiple outcomes. The present analysis utilized pooled data to more accurately and fully characterize the onset of effect for duloxetine.
Results: Median times to sustained improvements of 10% and 20% in the HAMD(17) total score among duloxetine-treated patients were 14 days and 21 days, respectively, compared with 34 days and 49 days, respectively, for placebo-treated patients (p < 0.001 for both results). The median time to sustained 30% improvement in HAMD(17) total score was 35 days for duloxetine-treated patients, while the median time for placebo-treated patients was not estimable since less than half of the patients met this criterion by the end of the trial. For duloxetine-treated patients, median times to sustained 10%, 20%, and 30% improvements on the Maier subscale of the HAMD(17) were the same as those for the HAMD(17) total score: 14, 21, and 35 days, respectively. However, in other analyses, changes in core emotional symptoms as measured by subscales of the HAMD(17) were somewhat faster than changes in overall symptomatology. The probabilities of achieving a sustained 30% improvement (Maier subscale) at Week 1 for duloxetine- and placebo-treated patients were 16.2% vs. 4.8%, respectively (p < 0.001). The corresponding probabilities of sustained improvement at Weeks 2 and 3 for duloxetine were 32.5% and 45.4%, respectively, compared to 12.8% and 21.4% for placebo ((p < 0.001 for both comparisons).
Conclusion: The absence of an active comparator limits the conclusions which can be drawn regarding the rapidity of onset of clinically meaningful improvement. However, results from the present investigation may be useful to clinicians in consideration of treatment options for individual patients.