Introduction Cross sectional therapeutic drug monitoring (TDM) data mining introduces new opportunities for the investigation of medication treatment effects to find optimal therapeutic windows. Medication discontinuation has been proven useful as an objective surrogate marker to assess treatment failure. This study aimed to investigate the treatment effects of escitalopram and pharmacokinetic influences on blood levels using retrospectively assessed data from a TDM database. Methods Data was collected from 134 patients longitudinally treated with escitalopram for whom TDM was requested to guide drug therapy. Escitalopram metabolism was estimated by the log-transformed dose-corrected concentrations and compared within subpopulations differing in age, gender, renal function, smoking status, body mass index, and comedication. Results Patients with a depressive episode who were treated with escitalopram and discontinued the treatment within the hospital stay showed lower serum concentrations compared to patients who continued escitalopram treatment with a concentration of 15 ng/mL separating both groups. Variability was high between individuals and factors influencing blood levels, including dose, sex, and age. Comedication that inhibits cytochrome P450 (CYP) 2C19 isoenzymes were further found to influence escitalopram pharmacokinetics independent of dose, age or sex. Discussion Medication switch is a valuable objective surrogate marker to assess treatment effects under real-world conditions. Of note, treatment discontinuation is not always a cause of insufficient response but may also be related to other factors such as medication side effects. TDM might not only be useful in addressing these issues but titrating drug concentrations into the currently recommended reference range for escitalopram will also increase response in non-responders and avoid treatment failure in underdosed patients.