- Design
- assessor-blinded randomised controlled trial, single centre
- Population
- 572 adults aged 18–65 discharged from a Hong Kong emergency department with at least one health-risk behaviour
- Primary outcome
- self-reported abstinence from at least one health-risk behaviour at 6 months
- Effect
- 30.1% vs 19.9%, risk ratio 1.51 (95% CI 1.13–2.02), p=0.006
An emergency visit is a moment when people are unusually open to changing something, and almost nothing is built to use it. This assessor-blinded randomised trial in a Hong Kong public hospital enrolled 572 adults aged 18 to 65, triaged semi-urgent or non-urgent, with at least one health-risk behaviour and a smartphone. The intervention was a brief structured telephone intervention after discharge followed by weekly WhatsApp or WeChat messages for six months; the control group received brief telephone advice only.
At six months, 30.1% of the intervention group reported abstaining from at least one health-risk behaviour against 19.9% of controls (risk ratio 1.51, 95% CI 1.13 to 2.02, p=0.006). The intervention also increased the likelihood of having fewer risky behaviours at six months (1.54) and twelve (1.48). Physical inactivity moved most: 31.7% against 16.2% at six months. The effects attenuated once the booster messaging stopped.
Two things make this worth acting on despite the limitations — single centre, self-reported outcomes, loss to follow-up. First, the control arm also received telephone advice, so the comparison is messaging versus advice alone rather than versus nothing. Second, the fade after cessation is the honest finding and the operational one: this is not a discharge intervention that keeps working, it is a support that works while it runs. Any department building this needs to decide who keeps sending the messages, and for how long, before it starts.
- Treat the discharge conversation as an opening, not a formality — brief structured advice was the control and still worked
- Plan who sustains follow-up messaging before starting a programme; the effect faded when it stopped
- Target physical inactivity, which showed the largest movement
- Expect self-reported abstinence to overstate the real effect
- Note the trial excluded urgent presentations — this applies to the patients you discharge, not those you admit
Why it matters
It puts a number on the teachable moment that emergency departments are always told to use and never resourced to.
Don't overread it
The outcome is self-reported abstinence in an unblinded participant group — the true effect is likely smaller than 30% against 20%.
The statistics, in plain English
A risk ratio of 1.51 on a 19.9% control rate translates to about 10 extra people in every hundred reporting abstinence — a large absolute difference, which is why the wide interval of 1.13 to 2.02 still excludes no effect. But the endpoint is what participants said, and participants who received six months of supportive messages have an obvious reason to report success. The attenuation after the messages stopped is the strongest evidence in the paper that something real was happening, because a purely reported effect would have no reason to fade.
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.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for emergency & critical care, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free