Embedded motivational interviewing combined with a smartphone application to increase physical activity in people with sub-acute low back pain: a cluster randomised controlled trial.

The trial did not reach its target of 60 participants and was stopped early for resourcing reasons; 46 were enrolled and 6 did not complete the follow-up assessment, so it was likely too small to show a difference in physical activity. The control group received more physiotherapy on average (4.6 versus 3.2 consultations), which the authors say makes the between-group findings difficult to interpret. 26 of the 83 outcome assessments were completed by a study coordinator and were therefore unblinded. The groups were not alike at the start: mean age was 39.2 years in the experimental group and 49.5 years in the control group, and symptom duration also differed. One hospital site recruited only two patients, after an unexpected fall in referrals for low back pain. The app was meant to be used every one to three days, but 35% of the experimental group never used it and no physiotherapist reported using the online portal meant to link the app with consultations, so the intervention was not delivered as intended. Physiotherapists' motivational interviewing skill was assessed in a 20-minute conversation with a study coordinator about the coordinator's own health behaviour, not in a real low back pain consultation, so how well they embedded it in practice is unknown. The two parts of the intervention, the in-session motivational interviewing and the app, were given together and cannot be separated from each other. The text and the tables report the number of patients per group differently (20 control and 26 experimental in the text, reversed in the tables), so the exact group numbers are unclear. People scoring in the severe range for depression or anxiety were excluded, so these findings do not speak to that group.

The paper states that the author declares no conflicts of interest. No funding source is reported in the text provided, and nothing indicates that the developers of the MiMate app, or any company, funded, designed or wrote up the study. The app is described as newly designed for this trial.

Do not read the significant differences in disability, function and confidence as proof that motivational interviewing makes patients worse. Those differences favoured the control group, but the control group received more physiotherapy sessions on average, the trial was underpowered for the physical activity outcome, and the two groups were not otherwise delivered the same care. Equally, do not read this as evidence that motivational interviewing does not work; it tested one particular way of embedding it, with poor app use and less motivational interviewing delivered than in earlier studies.