Accuracy of self-reported adherence and therapeutic drug monitoring in a psychiatric emergency ward

The aims of the study were: (1) the evaluation of the agreement between therapeutic drug monitoring (TDM) and a self-assessment of adherence to psychopharmacological treatments; (2) the identification of predictors of TDM results.Adherence in patients admitted into a psychiatric emergency service (PES) for a relapse of a schizophrenia spectrum disorder (SSD) or a bipolar disorder (BD; DSM-5) was assessed both directly with TDM and indirectly with a self-reported measure (Medication Adherence Report Scale -MARS- 10 items). The agreement between TDM and MARS was evaluated. Fifty-seven patients with SSD and 76 people with BD participated in the study. TDM was in range in about 50% of the global sample. No evidence of an association between MARS total scores and TDM results was found. Sensitivity, specificity, positive and negative predictive values of almost all MARS total scores were near to 50%. Smoking was strongly associated with a reduction of TDM results within the reference range. In the BD group, female sex was a predictor of TDM in range. In this clinical setting, self-assessment of adherence is neither reliable nor predictive. Furthermore, smoking is a strong predictor of poor adherence to psychopharmacological therapy.