Objective: Non-response to treatment with antidepressants (AD) is a clinical problem.
Method: The algorithm for pharmacological treatment of the Dutch multidisciplinary guideline for depression is compared with four other algorithms.
Results: The Dutch algorithm consists of five subsequent steps. Treatment is started with one out of many optional ADs (step 1); in case of non-response after 4-10 weeks, best evidence is for switching to another AD (step 2); next step is augmentation with lithium as the best option (step 3); the next step is a monoamine oxidase inhibitor (MAOI) (step 4); and finally electroconvulsive therapy (step 5). There are major differences with other algorithms regarding timing of augmentation step, best agents for augmentation and role of MAOI.
Conclusion: Algorithms for AD treatment vary according to national and local preferences. Although the evidence for most of the treatment strategies is rather meagre, an AD algorithm appears to be an useful instrument in clinical practice.