Current issues of ACP Journal Club are published in Annals of Internal Medicine


Problem solving improved major depression in primary care

ACP J Club. 1995 July-Aug;123:12. doi:10.7326/ACPJC-1995-123-1-012

Source Citation

Mynors-Wallis LM, Gath DH, Lloyd-Thomas AR, Tomlinson D. Randomised controlled trial comparing problem solving treatment with amitriptyline and placebo for major depression in primary care. BMJ. Feb



To determine whether a brief psychological treatment (problem solving) was as effective as antidepressant drugs and more effective than placebo in improving symptoms in patients with major depression in primary care.


Randomized controlled trial with 12-week follow-up.


26 general practitioners working in 15 practices in Oxfordshire.


91 patients (mean age, 37 y; 77% women) with major depression (low mood with at least 4 key symptoms of depression and a score of ≤ 13 on the Hamilton rating scale for depression). Exclusion criteria were diagnosis with another psychiatric disorder before the onset of depression, receiving psychological or antidepressant drug treatments, current psychotic symptoms, suicidal intent, history of schizophrenia, recent drug or alcohol misuse, or physical problems that precluded taking amitriptyline. 82 patients (90%) were included in the analysis.


30 patients were assigned to problem-solving treatment that consisted of 6 treatment sessions focusing on the rationale of problem solving and dealing with the patient's problems using the problem-solving method. 31 patients were assigned to amitriptyline, 25 mg for 2 nights then increased by 25 mg every third night until 150 mg (6 capsules) was being taken, plus standard clinical management. 30 patients were assigned to placebo, with the same schedule as that for amitriptyline, plus standard clinical management.

Main Outcome Measures

The Hamilton rating scale for depression, the Beck inventory, the modified social adjustment scale, the present state examination, and number of patients who recovered.

Main Results

Problem-solving treatment was superior to placebo at both 6 and 12 weeks for the adjusted mean differences in the scores on the Hamilton rating scale (5.31; 95% CI, 1.62 to 9.00 and 4.69; CI, 0.41 to 8.96, respectively), the Beck inventory (6.20; CI, 0.90 to 11.50 and 7.88; CI, 1.95 to 13.81), and the modified social adjustment scale (0.45; CI, 0.13 to 0.77 and 0.45; CI, 0.09 to 0.80). The differences between the problem-solving and amitriptyline groups were not statistically significant. Groups did not differ for results of the present state examination. More patients in the problem-solving group than in the placebo group were considered to have recovered at 6 and 12 weeks, (12 vs. 1; P = 0.003 and 18 vs. 8; P = 0.03).


Problem-solving treatment improved the symptoms in patients with major depression in primary care.

Sources of funding: Wellcome Trust and Warner-Lambert (drugs).

For article reprint: Dr. L.M. Mynors-Wallis, Department of Psychiatry, University of Oxford, Warneford Hospital, Oxford, United Kingdom 0X3 7JX. FAX 44-1865-793-101.


Mynors-Wallis and colleagues report results of a randomized, double-blind comparison of problem solving, amitriptyline, and placebo for the treatment of depression. The findings show a significant benefit for problem solving compared with placebo but not compared with amitriptyline. Because the study did not show any beneficial effects of amitriptyline compared with placebo, the relative benefits of this drug compared with problem-solving therapy remain tentative. The dose of amitriptyline, 150 mg/d, was not adjusted individually and may have been inadequate for some patients. This may explain why the response to amitriptyline was not superior to the response to placebo.

Mynors-Wallis and colleagues do not address the possibility that problem solving exerted its effect by being a stronger placebo. Talking about one's problems may be perceived as a more helpful activity than talking about the medications. Although the problem-solving intervention may be given by primary physicians, the feasibility of changing the practice patterns of primary physicians is uncertain. Also, although most physicians feel comfortable adjusting medications, many primary physicians may feel uncomfortable using a "new" psychotherapeutic intervention at successive office visits. A final concern is the high dropout rate for the placebo group.

On the basis of these data, common sense, and good clinical judgment, it hardly can be denied that attempting to assist patients with their problem solving is reasonable, particularly when they are depressed. Many primary physicians probably will feel uncomfortable trading proven drug therapies for problem solving. Others will question the need for problem solving as an adjunct to antidepressant drugs (i.e., doing both may increase the cost of treating depression). Indeed, the study did not test whether combining the 2 treatments results in additional benefit to either alone. Thus, the applicability of the findings in practice is uncertain.

John Absher, MD
Bowman Gray School of Medicine Winston-Salem, North Carolina