The nurse lost the test, and the patients still got better

Researchers at Karolinska Institutet randomly assigned 172 primary-care patients in Stockholm with mild-to-moderate stress-related ill health to one of two routes¹. Eighty-five got nurse-led collaborative care: two to six visits over up to twelve weeks, built on problem-solving, explanation and care coordination¹. Eighty-seven got guided self-help cognitive behavioural therapy: twelve weekly modules with support from a psychologist, online or as a book with sessions¹. The study was published in BMJ Mental Health and summarised by Medscape on Wednesday¹.

The test the nurses failed. The question was whether the nurse route was “not worse” than the therapy by more than a set margin. It was not shown to be: the difference favoured the therapy (effect size 0.22), and the upper end of its confidence interval, 0.52, overshot the 0.3 margin¹. That is a precise way of saying the trial cannot rule out that the nurses’ route is meaningfully weaker.

The numbers both groups share. On the ten-item Perceived Stress Scale, the nurse group fell from 24.9 to 19.4 at twelve weeks and the therapy group from 25.8 to 18.4¹. A year later the scores were 16.3 and 15.2¹. Desk arithmetic: that is a drop of 8.6 points for the nurse route and 10.6 for the therapy, measured from where each group started. Patients in the nurse group needed on average 3.68 sessions¹.

What the study cannot say. There was no untreated group, so part of the improvement may be time itself¹. It was one centre, and most participants were middle-aged, well-educated women¹. Satisfaction ratings were similar in both arms¹.

The constructive reading. The authors’ own conclusion is modest: the nurse route is “safe, relatively easy to administer” and may still be useful in primary care¹. That is the kind of finding that rarely makes a headline and often makes a service work. Not every patient needs the best-scoring option on day one; many need a competent first step that is available this month.