Seasonality, smoking and history of poor treatment compliance are strong predictors of dropout in a naturalistic 6 year follow-up of bipolar patients.

Bipolar disorder is a highly recurrent disease which requires long-term treatment. Dropout is a major problem, poorly understood. The objectives of this study were to know the risk of dropout of a cohort of bipolar patients under ambulatory treatment and to identify the clinical profile of patients more likely to abandon the follow-up. A sample of 285 BD I and II patients was followed up for a mean of 2.87 years. A significant proportion of patients failed regular follow-up. The dropout rates were 6.3 % at three months, 12.7 % at 6 months, and 17.6, 27.2, 37.3, 44.0, 47.2 and 49.0 % at 1, 2, 3, 4, 5 and 6 years respectively. Very few variables at baseline predicted dropout. Patients under 35 years of age were more likely to dropout than older cases. Seasonality, smoking and specially history of poor treatment compliance were strong predictors of dropout. Given the magnitude of dropout, additional early clinical interventions should be considered for high-risk patients.