Cross sectional database analysis of antidepressant prescribing in general practice in the United Kingdom, 1993–5

Methodsand results

Cross sectional data for two periods of 12 months ending June 1993 and June 1995 were obtained from the database of the doctors' independent network (DIN-LINK); this is maintained by the independent company Compufile (Woking, Surrey), which provides general practice data mainly to the pharmaceutical industry. It contains data from 100 general practices (377 general practitioners, 750 000 patient records) across the United Kingdom that use AAH Meditel practice computer systems. Prescriptions for antidepressants linked to a general practitioner's diagnosis of depression were included in the study. Antidepressants that accounted for less than 2% of the total were excluded. The effective dose of tricyclic antidepressants was taken as 125 mg daily from the consensus guidelines1 and that of selective serotonin reuptake inhibitors as the value in the manufacturers' data sheets.

Since 1993 the number of prescriptions issued for the treatment of depression increased overall by nearly 33%. The prescription rate for all antidepressants increased, but the increase was much lower in the older tricyclic antidepressants (12.4%) than in the selective serotonin reuptake inhibitors (133.8%). The number of people receiving an effective dose increased from 41% in 1993 to 53% in 1995. For amitriptyline, clomipramine, and dothiepin the percentage of prescriptions for an effective dose and the average dose prescribed changed only marginally (table 1).

Comment

In this large population the number of prescriptions for antidepressants issued to patients with a diagnosis of depression increased substantially between 1993 and 1995. We emphasise, however, that antidepressants are far from being the only effective treatment for depression.

Prescriptions for all antidepressants increased but more for selective serotonin reuptake inhibitors than for tricyclic antidepressants. This suggests that general practitioners are not prescribing selective serotonin reuptake inhibitors to patients who have been taking tricyclic antidepressants long term.

Encouragingly, the percentage of patients receiving an effective dose of antidepressant increased, but this is accounted for entirely by the much greater increase in prescribing of selective serotonin reuptake inhibitors than tricyclic antidepressants. Although for the older tricyclic antidepressants the proportion of prescriptions for an effective dose and the average doses prescribed were unchanged, we note that the British National Formulary still has not adopted the consensus guidelines on antidepressant doses. However, Kerr found that only 52% of general practitioners would prescribe tricyclic antidepressants at a dose lower than that recommended by the consensus guidelines3 compared with 86% in our study. This may reflect a difference between actual prescribing practice and what general practitioners say about their practice. Doctors may consider that low doses of tricyclic antidepressants offer effective treatment. They may be unaware of, or do not believe, the consensus guidelines or the results of clinical trials showing that tricyclic antidepressants prescribed at these doses are not beneficial in the treatment of depression. They may also lack confidence in prescribing at higher doses or fear that patients will find effective doses intolerable because of side effects.

So far as we are aware low doses of tricyclic antidepressants are ineffective. Failure to treat depression effectively contributes to relapse and the development of recurrent and chronic depression1 and may be a contributory factor in suicide.4 One study found that a large proportion of patients receiving long term treatment with antidepressants at low doses were still moderately to severely depressed during follow up.5 Further research is urgently needed to determine why general practitioners continue to prescribe tricyclic antidepressants at low doses and to investigate the consequences of such prescribing on a large scale.