Digital Patient Engagement in Technology-Supported Neurorehabilitation (PSY-NeT-Engage)

PSY-NeT-Engage is an academic, non-profit, multicentre, cross-sectional observational study designed to characterize digital patient engagement during technology-supported neurorehabilitation and to examine factors associated with differences in engagement. The study will be conducted among patients already receiving standard neurorehabilitation involving robotic systems for upper-limb training and/or virtual reality-based rehabilitation technologies at participating clinical centres. It will not assess the clinical or therapeutic efficacy of these technologies, functional recovery, or the achievement of rehabilitation goals, nor will it introduce any experimental intervention or modification to the standard rehabilitation pathway. Because the data will be collected cross-sectionally, the findings will be interpreted as exploratory associations and predictive relationships within the specified statistical models.

Data will be collected once for each participant during the standard neurorehabilitation pathway, after the participant has completed at least three rehabilitation sessions using the technologies under study. Consequently, participants may be assessed at different stages of rehabilitation and after completing different numbers of technology-supported sessions.

Patient-reported data will be collected through a single computerized self-report questionnaire administered via the Qualtrics digital platform using a password-protected tablet. Questionnaire completion is expected to require approximately 15 minutes. Research staff will be available to provide procedural clarification and technical assistance without influencing participants' responses. Relevant sociodemographic, clinical, functional, and rehabilitation-related variables will be extracted from clinical documentation. Variables not available in the clinical records will be collected directly from participants. No follow-up assessment is planned.

In addition to the primary outcome measure of digital patient engagement, the study includes secondary outcome measures of overall satisfaction with the rehabilitation pathway, behavioural intention to continue using rehabilitation technologies, and the perceived psychosocial impact of rehabilitation technologies. Four other pre-specified exploratory variables will also be assessed: hope, perceived quality of the patient-healthcare professional relationship, technology anxiety, and disease-management self-efficacy. These variables will be examined as potential psychosocial, relational, and individual predictors of digital patient engagement. Disease-management self-efficacy will be evaluated as a potential moderator.

Hope will be assessed using the 2-item Brief Hopelessness Scale with Positive Valence (Brief-H-Pos), which measures positive expectations regarding the future and the perceived attainability of personal goals. Higher scores indicate greater hope.

The perceived quality of the patient-healthcare professional relationship will be assessed using the 8-item Nordic Patient Experiences Questionnaire (NORPEQ). The questionnaire evaluates relevant dimensions of the patient's care experience, including communication with healthcare professionals, perceived care and attention, and trust in the healthcare team. Higher scores indicate a more positive care experience and better perceived relationship quality.

Technology anxiety will be assessed using the 11-item Abbreviated Technology Anxiety Scale (ATAS), which evaluates negative affective responses to technology, including apprehension, confidence, and perceived competence in using technological tools. Higher scores indicate greater technology anxiety.

Disease-management self-efficacy will be assessed using the 10-item Self-Efficacy Self-Care Scale (SCSES), which evaluates participants' confidence in their ability to maintain their health, monitor their condition, and manage symptoms while living with a chronic condition. Higher scores indicate greater disease-management self-efficacy. This variable will be examined as a secondary individual exposure factor and as a potential moderator of the associations between hope, the patient-healthcare professional relationship, and digital patient engagement.

For measures without an available validated Italian version, a research-specific Italian translation will be developed using forward-backward translation followed by review by an expert panel. The translated measures and data-collection procedures will initially be piloted in a subgroup of 15 participants to assess item comprehensibility and the adequacy of questionnaire administration. These translations are intended exclusively for research use and will not constitute validated instruments for routine clinical practice.

The minimum target sample of 129 participants was established through an a priori power analysis for structural equation modelling, assuming a moderate effect size of 0.30, statistical power of 0.80, and a two-sided significance level of α = 0.05. Consecutive non-probability sampling will be used, and data collection may continue beyond the minimum target throughout the planned recruitment period to improve the robustness and generalizability of the findings.

The statistical analysis will include descriptive statistics, assessment of variable distributions and outliers, and evaluation of internal consistency using Cronbach's alpha. Exploratory factor analysis will be used to examine the factor structure of measures without a validated Italian version. Initial associations will be evaluated using Pearson or Spearman correlation coefficients, as appropriate. Structural equation modelling will be applied within a confirmatory-exploratory framework to examine the plausibility, direction, and magnitude of the hypothesized direct, indirect, and total associations. Exploratory moderation analyses will be conducted using regression-based interaction models, and multiple regression models will be used to examine additional associations specified in the protocol. Robust estimators will be considered when non-normality or influential outliers could materially affect the results.