Noncommunicable diseases (NCDs) such as cardiovascular disease, cancer, chronic respiratory disease, and diabetes are the leading causes of morbidity and mortality globally and are driven by modifiable behaviors including tobacco use, harmful alcohol consumption, unhealthy diet, and physical inactivity. Emergency department (ED) visits represent a potential teachable moment for behavior change because many patients present during acute illness or discomfort and may be receptive to health advice. The authors designed a pragmatic trial to evaluate whether a digitally augmented, theory-based general health-promotion approach delivered at ED discharge could increase abstinence from health-risk behaviors.
This was an assessor-blinded randomized controlled trial conducted at a major public hospital ED in Hong Kong. The trial compared a brief telephone-based AWARD-model health-promotion intervention augmented by weekly instant-messaging boosters to a control condition of brief telephone advice alone. The study is reported as an uncorrected proof in PLoS Medicine.
Adults aged 18–65 years who were triaged as semi-urgent (level 4) or non-urgent (level 5), reported at least one health-risk behavior, and had access to a smartphone were screened for inclusion. From 2,134 screened patients, 572 participants met eligibility criteria and were enrolled and randomized, with 286 allocated to the intervention group and 286 to the control group.
The intervention combined two components. First, participants received a brief telephone-based intervention delivered according to the AWARD model: Ask, Warn, Advise, Refer, and Do-it-again. Second, intervention participants received weekly mobile instant messages via WhatsApp or WeChat for 6 months to reinforce the initial counseling.
Control-group participants received brief telephone advice only, without the planned weekly digital booster messages.
The primary outcome was self-reported abstinence from at least one health-risk behavior at 6 months after randomization. Secondary outcomes included self-reported abstinence from at least one health-risk behavior at 12 months and reduction in the number of health-risk behaviors at 6 and 12 months. Follow-up included outcome assessment at 6 and 12 months; outcome measures were self-reported.
The trial report includes intention-to-treat analyses and relative risks (RR) with accompanying 95% confidence intervals and P values for the primary and secondary outcomes. Specific sample size calculations are reported in the source article; this summary reports the enrolled sample (572 participants) and the randomized allocation (286 per arm).
At 6 months, 30.1% of participants in the digitally augmented intervention arm reported abstinence from at least one health-risk behavior compared with 19.9% in the control arm. This corresponded to a risk ratio of 1.51 (95% CI, 1.13–2.02; P = 0.006), indicating a statistically significant increase in the likelihood of stopping at least one risk behavior during the active intervention period.
The intervention also increased the likelihood of a reduction in the number of risky behaviors at both 6 months (RR = 1.54; P = 0.01) and 12 months (RR = 1.48; P = 0.02) compared with control.
Among the individual behaviors evaluated, physical inactivity showed the largest improvement at 6 months: 31.7% of intervention participants versus 16.2% of controls achieved the predefined improvement for physical activity (P < 0.001). Other behavior-specific changes were reported in the trial but are summarized here as part of the overall reduction in risk behaviors.
The observed benefits were strongest during the 6-month active messaging period. Effects attenuated after booster messaging ceased, with weaker evidence of sustained differences at 12 months. The authors note that the absence of sustained effects suggests that extended support or maintenance strategies may be required to maintain improvements over time.
The trial authors identified several limitations that could affect interpretation and generalizability: reliance on self-reported outcomes rather than objective measures; single-center conduct in one Hong Kong public hospital; and loss to follow-up. These factors may introduce reporting bias and limit external validity. The dataset is not publicly deposited because consent and ethics approvals do not permit public sharing of participant-level data; data are available on request to qualified researchers subject to institutional requirements.
A low-cost, scalable ED-delivered program combining a brief telephone AWARD-model intervention with weekly mobile instant-messaging boosters for 6 months produced short-term increases in self-reported abstinence from at least one health-risk behavior and greater reductions in the number of risk behaviors at 6 months. The findings support the feasibility of using brief counseling plus common messaging platforms (WhatsApp or WeChat) to reach discharged ED patients and promote lifestyle changes that reduce NCD risk. Because effects weakened once messaging stopped and because the trial was single-center and relied on self-report, further research is suggested—specifically multicenter trials, longer follow-up, and testing of extended or maintenance support—to determine long-term effectiveness and generalizability.
The trial is registered at ClinicalTrials.gov (NCT06077565). The data underlying the study are held by the Nethersole Evidence-based Nursing Practice Unit (NENPU) at The Chinese University of Hong Kong and are available on request to qualified researchers, subject to ethical and institutional conditions. Funding was provided by the Fund for Evidence-based Practice Improvement Collaborative Projects, NENPU, Nethersole Group Hospitals, Hong Kong SAR, China. The authors declared no competing interests.