Background: Accordons-nous, a smartphone app, was developed to support patients in the advance care planning (ACP) process. The app raises awareness and facilitates communication on this sensitive topic. It helps patients express their values and preferences for care and write their advance directives (AD).
Objective: Measure the impact of distributing Accordons-nous on patients' propensity to engage in the ACP process, compared with the distribution of a leaflet. A secondary objective was to test the effect of socio-demographic factors (age, health status, gender, level of education) on propensity to engage in ACP.
Methods: Pre-post randomized control study. Participants were patients approached in medical waiting rooms. They received the app (treatment) or an information leaflet (control). They responded to two questionnaires: one at recruitment and a second 3-4 weeks later. Improvement on four variables relevant to ACP was measured: reported contemplation of an event relevant to ACP; decision about treatment in case of that event; discussion about it with relatives or health care professionals; writing advance directives. Statistical analysis included between-group comparisons of pre-post differences with 2-sample tests for equality of proportions and logistic regression models.
Results: Four hundred seventy three participants were recruited and full responses obtained from 312. Overall, the intervention (control and treatment together) had a positive effect on the mean reported ACP engagement for all variables: new or renewed contemplation 54%; increase in decision 8%, discussion 11%, and writing 1%, compared to the baseline. Compared to the control (leaflet), the treatment group (app) had a larger effect size for all variables: between-group difference in contemplation + 11% (logistic regression, p = .05), decision + 1% (but p > .05 on this variable), discussion + 5% (p = .05), and writing AD + 5% (p = .03). Moreover, greater age was positively correlated with having written AD at inclusion (21% among retired compared to 2% among young adults) and with the propensity to write AD after our intervention (logistic regression, p = .001). Other factors tested (frequency of consultations, gender, level of education) had no effect on participants' ACP engagement.
Conclusions: When distributed without specific counselling, the tool increased reported ACP engagement, although effect sizes remain modest. Further studies are needed to investigate whether the app could generate greater ACP engagement if used by professionals in dedicated ACP consultations.
