Factors affecting early attrition and later treatment course of antidepressant treatment of depression in naturalistic settings: An 18-month nationwide population-based study

BackgroundEarly attrition can impede treatment success of depression; its contributing factors and impacts on subsequent treatment course need further clarification.MethodsAll Taiwanese adult patients prescribed with antidepressants for depression (n = 216,557) in 2003 were identified through a total population health insurance claims database; their initial contact patterns could be classified into three types of attrition: non-attrition, returning attrition and non-returning attrition. Demographic and clinical characteristics associated with each attrition type were described and relationships between attrition type and subsequent treatment course over an 18-month follow-up period were examined with these demographic/clinical confounders being controlled for.Results41.6% of Study subjects had early attrition; among them, 35.3% returned to treatment later. Type of depression, medical/psychiatric comorbidities, painful physical symptoms and past treatment history, as well as prescribing physician specialty and choice of antidepressants, were associated with early attrition. Three types of follow-up pattern over the 18-month follow-up period were identified: sustained treatment-free, continuous treatment and late re-contacts. Patients remaining engaged with treatment within the first three months had higher odds of achieving sustained treatment-free (OR = 1.21; 99% CI: 1.16, 1.27) and lower odds of having late re-contacts (OR = 0.20; 99% CI: 0.19, 0.21) over the 18-month period, compared to those who returned after early attrition.ConclusionsEarly attrition is a significant barrier for depression treatment in daily clinical practice and has negative impacts on later treatment course and/or outcome. Early attrition needs to be minimized through shared decision-making, exchange of treatment preferences and proper patient–physician communication.