This is a practical, clinically oriented management guide that summarizes the side effects of psychotropic drugs and presents guidelines for successful management. The editor points out that the incidence of side effects is difficult to estimate because it depends on dose, concomitant medication, age, individual sensitivity, and other factors.

The issue of whether advance counseling about side effects causes patients to develop side effects by means of suggestion and thus should be discussed only after they occur is confronted (but not resolved). Should orthostatic hypotension be discussed when giving a tricyclic antidepressant to a 30-year-old or to a 70-year-old? What is frustrating here is that there are so few hard data as to the benefits of different ways to introduce such issues to patients, monitoring side effects, or facilitating patient-physician cooperation. We depend on reasonable conjecture rather than demonstrated fact.

There is a certain commonsense quality about many recommendations, therefore, which highlights the need for good clinical sense and experience but at the same time demonstrates the lack of useful factual knowledge. For instance, in the discussion of sexual side effects, the use of adjunctive pharmacological agents (e.g., yohimbine, sildenafil, cyproheptadine) is mentioned, but, in fact, it is really gossip. That various psychotropic medications are relatively safe during pregnancy (e.g., antipsychotics, tricyclic antidepressants, and selective serotonin reuptake inhibitors) is asserted, but the database is not specified and is quite slim.

Since the database is often slow to develop or persistently absent, it is understandable that reasonable clinicians may differ about correct practice. For instance, I give nifedipine to my patients who are taking monoamine oxidase inhibitors, if they have no evidence of coronary insufficiency, to carry with them and use if they have a sudden, pounding, occipital headache. My judgment is that the potential benefits far outweigh the possible risks. I could be wrong, but the advice to go to an emergency room if you get a headache seems more like a gesture than a useful procedure to me.

This excellent book forcefully brings to our attention that the current direction of psychiatric research support, combined with the lack of a useful system for postmarketing surveillance plus systematic monitoring of clinical interventions and outcomes, ensures continued ignorance about important practical issues. Further, given the rapid development of knowledge regarding drug interactions, any book must rapidly fall out of date.

Some day a Web site will be developed that will interactively answer questions about drug indications, side effects, and interactions on the basis of systematic clinical data. (But who will take the responsibility?) Until then, the efforts of Dr. Balon and his collaborators are very welcome.