After low-risk chest pain, a referral letter is not the treatment

Design

randomised three-group comparative effectiveness trial, 12-month follow-up

Population

375 adults discharged from six emergency departments with low-risk chest pain and at least moderate anxiety

Primary outcome

change in GAD-7 anxiety score over 12 months

Effect

peer-supported internet therapy 1.22 points better than referral (95% CI 0.01-2.43); 2.8 points in severe anxiety

PACER randomised 375 adults across six university emergency departments who had been discharged with low-risk chest pain and had at least moderate anxiety on the GAD-7 or a positive panic screen. Mean age was 39.9 and 70.7% were women. All three arms got psychoeducation; they were then assigned to a recommendation to follow up in primary care, to peer-supported internet cognitive behavioural therapy, or to therapist-delivered cognitive behavioural therapy by telehealth. The outcome was change in GAD-7 over twelve months.

Every arm improved: 4.32 points with the primary care recommendation, 5.54 with peer-supported internet therapy, 5.13 with therapist-delivered telehealth. The internet arm beat the referral by 1.22 points (95% CI 0.01-2.43, effect size 0.25), and by 2.8 points in those with severe anxiety. The odds of patients reporting global improvement were about three times higher in both therapy arms. Depression, somatisation and disability improved similarly across all three.

The between-group difference is small and its confidence interval nearly touches zero. What is not small is the reframing. These are the patients we spend a great deal of money ruling out coronary disease in, and then discharge with a letter, and a third of whom return. Naming the anxiety and offering something structured is a defensible step even where the trial's specific internet programme is unavailable - which in India it is. The transferable finding is that the referral recommendation alone was the weakest of three options, not that this particular platform should be bought.

• Screen for anxiety with the GAD-7 before discharging a patient with low-risk chest pain.
• A referral recommendation alone performed worst of the three options tested.
• The benefit was largest in severe anxiety - 2.8 points - so prioritise those patients.
• Name the anxiety explicitly to the patient; all three arms received psychoeducation and all improved.
• The specific internet programme is not available in India; the principle of offering structured treatment is.

Why it matters

It challenges what happens after the troponin comes back negative, which is currently a letter and nothing else.

Don't overread it

A small between-group difference whose confidence interval nearly includes zero - this supports offering treatment, not any particular programme.

The statistics, in plain English

A 1.22-point advantage with a confidence interval running from 0.01 to 2.43 is statistically positive by the narrowest possible margin, and an effect size of 0.25 is small. The more informative numbers are the within-arm improvements, which were substantial everywhere, and the subgroup with severe anxiety, where the difference more than doubled - a pattern consistent with the general rule that treatments show most where there is most to treat.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.