Fig. 1

Effectiveness and safety of escitalopram treatment guided by dose personalization based on quantification of drug plasma concentration. (a) Schematic representation of study protocol; patients who achieved and maintained TRR with 10 mg/day of escitalopram throughout the trial were labeled as unadjusted dose group; patients who required dose adjustment at week 4 to achieve TRR were labeled as adjusted dose group; and patients who failed to achieve or maintain TRR at week 8 even after dose personalization were labeled as inadequate drug level group. (b) Hamilton Rating Scale for Depression (HAM-D) score decreased by 55% (95% CI: 47–64%, p < 0.001) from baseline to week 8, without difference in HAM-D score reduction between groups (p > 0.1). (c) The relative change in HAM-D score from baseline did not correlate with escitalopram plasma concentration (p > 0.1) at week 8. (d) Hamilton Rating Scale for Anxiety (HAM-A) score decreased by 52% (95% CI: 43–62%, p < 0.001) from baseline to week 8, without difference in HAM-A score reduction between groups (p > 0.1). (e) Clinical Global Impression of Severity (CGI-S) score decreased by 48% (95% CI: 39–54%, p < 0.001) from baseline to week 8, without difference in CGI-S score reduction between groups (p > 0.1). (f) Of 92 patients, 56 responded to treatment and 43 achieved remission; more responders (p = 0.0074) and remitters (p = 0.036) belonged to the unadjusted dose group compared to inadequate drug level group. (g) After week 4 and week 8 of escitalopram treatment, 39 and 47 out of 92 patients reported adverse drug reactions, respectively, without difference between groups (p > 0.1). (h) Frequency distribution of reported adverse drug reactions at week 8 is depicted from the most to the least frequent. (i) Escitalopram plasma concentration increased the probability for the occurrence of ADRs after week 8 by 3.2% (95% CI: 0.1–6.3%, p = 0.041) (j) QTc interval from baseline to week 8 was prolonged by 5.5 ms (95% CI: 1.8–9.3 ms, p = 0.0041), without differences in QTc interval prolongation (p > 0.1) between groups. In addition, (k) the QTc interval prolongation was not correlated with escitalopram plasma concentration (p > 0.1). Results are presented as bar charts, scatter plots and line graphs with annotated estimated marginal means ± 95% CI.