Health Outcomes Following the Health-Promoting "Reflective STRENGTH-Giving Dialogue" Intervention Among Community-Dwelling Older Adults.

The design was quasi-experimental and not randomised, so baseline differences between groups could explain the results. The sample was small: 34 intervention participants and 26 control participants were included at baseline, and only 27 intervention participants and 20 control participants were analysed because data at both time points were required. Attrition was substantial: seven intervention participants were lost and six control participants were lost. The intervention group was older and used more care services than the control group at baseline, including more medicine, safety alarms, weekly home-care visits, and assisted living. The GDS-20 had critically low internal consistency in this sample (0.42), so depression findings are weak. The NRS was used for well-being and impact of health problems, although it was validated for pain intensity, not these targets. The immediate effects were measured before and after dialogues in the presence of the health-care professional, which may introduce bias. Attention to intervention participants may have influenced results, although control participants also received attention during data collection. Use of the SelfSTRENGTH app and reflection tools may have been limited by health-care professional comfort and older adults' willingness or ability to use digital tools. Only participants who provided data at both time points were analysed, which may affect generalisability.

The authors declared no conflicts of interest. Funding came from the School of Health Sciences, University of Skövde, Sweden; the Skaraborg Institute for Research and Development; Sparbanksstiftelsen Lidköping; the Gösta Svensson Foundation; and the Agneta Prytz-Folke and Gösta Folke Foundation.

Do not conclude that the STRENGTH intervention, including the dialogues, reflection tools, and SelfSTRENGTH app, produced lasting health improvement. The immediate changes were measured right after each dialogue, and the 14-week follow-up changes were not significant.