A randomised controlled trial to evaluate the effectiveness and cost-effectiveness of counselling patients with...

To examine the effectiveness and cost-effectiveness of short-term counselling in general practice for patients with chronic depression or combined depression and anxiety, compared with general practitioner (GP) care alone.

A randomised controlled trial and economic evaluation with an initial assessment at randomisation and follow-ups at 6 and 12 months.

Nine general practices that were well-established participants of the Derbyshire counselling in general practice scheme, and already had a counsellor in the practice team.

Patients were screened at GP practices, and asked to participate if they scored >/= 14 on the Beck Depression Inventory (BDI), had suffered depression or depression/anxiety for 6 months or more, were aged 18-70 and had no history of drug or alcohol abuse, psychoses or suicidal tendencies.

The experimental group received usual GP treatment and were also referred to an experienced, well-qualified counsellor attached to their general practice. Of the eight counsellors, two practiced cognitive behavioural therapy (CBT) and six had a psychodynamic approach. The controls were referred back to their GP for routine treatment. There were no restrictions regarding the treatment that could be used, except that GPs could not refer controls to practice counsellors.

The main outcome measure was the BDI. Others included the Brief Symptom Inventory, the Inventory of Interpersonal Problems and the Social Adjustment Scale. All tests were given at initial, 6- and 12-month assessments. Comprehensive costs were also estimated, and combined with changes in outcomes to examine between-group differences and whether counselling was more cost-effective than standard GP care.

The trial recruited 181 patients. There was an overall significant improvement in the actual scores over time but no difference between groups or between CBT and psychodynamic counselling approaches at either 6 or 12 months. However, fewer experimental group patients were still 'cases' on the BDI than controls. This difference was statistically significant at 12 months and neared significance at 6 months (using logistic regression with the initial score as a covariate). In addition, most patients were very positive about the counselling and considered it helpful. Visual inspection of the outcomes suggested that more patients with mild or moderate depression at study entry had improved and ceased to be cases, and that more of these patients had become 'non-cases' in the experimental than the control group. However, a multiple regression analysis indicated no significant interactions between group and initial severity of depression. This could be partly due to there being no difference in outcome between the experimental and control group patients who were initially severely depressed and few of these patients ceasing to be cases at follow-up. There were no significant differences in the mean total costs, aggregate costs of services, or any of the service-group costs, except for primary care, between the experimental and control groups over time. The cost-burden to GP practices was significantly higher in the experimental than the control group at 6 months.

Although patients were generally appreciative of the counselling received, there was only limited evidence of improved outcomes in those referred to counselling. Stricter referral criteria to exclude the severely depressed may have yielded more conclusive results. It is also difficult to estimate the effect of recruitment by screening rather than GP referral, which may limit the applicability of the results to routine clinical practice, and may have interfered with the normal working alliance established between the GP, patient and counsellor. A patient preference trial may, therefore, have been more appropriate. The results indicated that there were similar improvements for both CBT and psychodynamic counselling, but a