RESUMO
INTRODUCTION: Post hoc analyses of clinical trials have shown that early improvement around day 14 is highly predictive for later response. More-over, evidence has been given that sufficiently high concentrations of antidepressant drugs in blood are required to attain response. In this study, we determined cut-off levels for citalopram serum concentrations and clinical improvement during the early phase of treatment to predict later response and the predictive power of these measures either alone or in combination. METHODS: Inpatients with depressive disorder according to ICD-10 who received citalopram were included. Psychopathology was assessed by the 17-item Hamilton Depression (HAMD-17) rating scale, and serum concentrations of citalopram were measured in weekly intervals. RESULTS: The analysis included 55 inpatients. Receiver operating characteristics analysis revealed for citalopram a serum concentration of 53 ng/ml on day 7 and a clinical improvement of 24% on the HAMD-17 scale on day 14 as significant cut-off values to predict response after 5 weeks of treatment. Both measures taken together predicted response on week 5 with 73% sensitivity and 85% specificity with an odds ratio of 14.6. DISCUSSION: It is concluded that treatment with citalopram should be guided by symptom rating at baseline and on day 14 and serum concentration determination on day 7.
