The role of the economist is to evaluate therapeutic options so as to identify both the value of what is given up to provide care (the cost) and the value of what is gained, in terms of improved health status (the benefit). The function of the economic evaluation is to inform choices: these will be determined by clinicians and managers. By informing the choice process, the economic calculus makes decisionmaking more ‘transparent’ and makes those who allocate resources more accountable to their peers, purchasers, and patients.

General practitioners (GPs) work in an environment different from that of most of their specialist colleagues. Some of these differences stem from being a doctor of first contact, some from working in one part of a multi-tier system for providing medical care, and some from the referral system which deals with the movement of patients between these tiers. Other differences stem from the GP's National Health Service contract to provide continuity of care 24 hours a day, and to make early diagnoses as well as monitoring normal development on one hand and chronic impairment, handicap, and disability on the other. The British GP works with a defined list of registered patients who can present symptoms representing disorders of a wide range of severity, originating in any bodily system. These symptoms can result from any type of pathogenic stimulus, including disordered perception: even temporary variation from the patient's usual subjective experience may arouse natural concern.

The adverse effects of antidepressants include their pharmacological side-effects, toxic effects, interactions with other drugs, and symptoms occurring during their withdrawal. The focus of this paper is on the pharmacological side-effects.

This paper does not attempt to review the whole topic of clinical efficacy but will consider general concepts, particularly in the light of the advent of the serotonin (5-HT) uptake inhibitors (SSRIs).

The aetiology of depression is complex. In most cases there is some form of predisposition based on problematical family dynamics, disordered family communication and/or genetic factors; each may be present to a variable extent. On top of this predisposition, some form of stimulus - ‘stress’, an external problem, loss of control, a physical disorder, adverse life circumstances - appears to affect those who are already primed. There are a variety of possible outcomes from this combination of predisposition and stimuli: low mood, depression in the full syndromic sense, and depression with biological features. It seems likely that, particularly with severe depressions (the major syndromic forms and those with biological features), there are neuroendocrine and neurochemical changes to mediate the effects of stress into those of the clinical symptoms and signs. It has been hypothesised that stressors induce adaptive neurochemical changes, failure of which may engender behavioural disturbances (Anisman & Zacharko, 1992). However, there are also many individuals who may not undergo this process, but who probably have some cognitive or learned change leading to dysphoria; this aspect needs to be considered carefully in relation to efficacy.

In psychiatry, the law, and economics, there are many areas of overlap where our knowledge and appreciation are growing and developing. Many doctors ‘understand the words but not the music’, and the same is true for the other professions involved. This is exemplified in relation to suicide.

Concern over the cost of health care is playing an increasing role in Great Britain, but evaluation of benefit is generally inadequate. This is particularly true in the case of depression, for which the cost of drugs is 1.9% of the National Health Service pharmaceutical budget. Since differences in effectiveness between antidepressant drugs are difficult to demonstrate, quality-of-life studies may help to identify outcome differences. At present, the worst outcome of depressive illness - suicide - absolves carers from further costs. An assessment is needed which takes into account both the direct and indirect costs of depression; this should include the costs of investment into improving diagnosis. Even without financial analyses, it is clear that some cases of suicide can be prevented by prescribing less toxic drugs, but research is needed to investigate whether use of these drugs also reduces the costs of overdose.

Antidepressants, as the term is currently understood, were introduced into clinical practice in the late 1950s. The first of these was imipramine, shortly to be followed by amitriptyline and then by a series of related compounds - all chemically ‘tricyclics’. These subsequently became known as the ‘first-generation’ antidepressants; imipramine and amitriptyline in particular have since become the standards against which others have been compared.