Background Remission is a common outcome of short-term trials and the main goal of acute and longterm treatment. The longitudinal stability of remission has rarely been investigated under naturalistic treatment conditions. Methods Naturalistic multisite follow-up study. Three-year symptomatic long-term outcome of initially hospitalized tertiary care patients (N = 784) with major depressive episodes. Remission rates as well as the switch rates between remission and non-remission were reported. Results After one, two and three years 62 %, 59 % and 69 % of the observed patients met criteria for remission. During the follow-up 88 % of all patients achieved remission. 36 % of maintained remission from discharge to 3-years, 12 % of all patients never reached remission and 52 % percent showed a fluctuating course switching from remission to non-remission and vice versa. There was considerable transition between remission and non-remission. For example, from discharge to 1 year, from 1 to 2, and from 2 to 3 years 25 %, 21 % and 11 % lost remission. Conclusion Cumulative outcome rates are encouraging. Absolute rates at predefined endpoints as well as the fluctuations between these outcomes reflect the variable and chronic nature of major depression.
