Abstract
BACKGROUND: Immersive virtual reality (VR) is promising in stroke rehabilitation; it is believed to promote motivation and intervention adherence among patients. However, existing work often lacks a structured assessment of user experience over a longer period of time. OBJECTIVE: This study aimed to assess the feasibility, user experience, and preliminary effectiveness of a VR-based rehabilitation program designed for patients with stroke to train upper limb and cognitive functions. METHODS: Thirty-two chronic (n=19, 59%) or postacute (n=13, 41%) patients with stroke (mean age 60, SD 11 years) were enrolled. All participants performed 4 weeks of training, performing exercises in the Virtual Supermarket for Stroke (VSS). The VSS is an ecological VR-based application allowing customization of difficulty to make the task of "doing the shopping" more challenging throughout the sessions. Subjective outcomes were assessed after the first and last sessions. Clinical scales were administered at baseline and at the end of the treatment. RESULTS: Of 32 participants, 31 (97%) completed the training. Flow (median 4.56, IQR 3.94-4.72; 5-point scale), sense of presence (Spatial Presence: median 3.44, IQR 12.85-3.85; Engagement: median 3.79, IQR 3.26-4.191; Naturalness: median 3.60, IQR 3.20-4.40; all 5-point scales), and affective state-related variables (Positive and Negative Affect Schedule; Positive Affect: median 4.60, IQR 4.00-5.00; Negative Affect: median 1, IQR 1.00-1.00) were satisfactory after the first session. Perceived ease of use was rated as very high (median 6.75, IQR 6.00-7.00; 7-point scale). No severe symptoms of cybersickness were recorded (Simulator Sickness Questionnaire [SSQ-TS]: median 11.22, IQR 0-20.57). At the end of the intervention, no significant differences were recorded in any subjective variable. Regarding clinical outcomes, significant improvements were recorded in balance (Berg Balance Scale pre: median 30, IQR 14.00-45.75; post: median 33.5, IQR 17.00-47.00; P=.02), upper limb motor functions (Motricity Index pre: median 45, IQR 15.25-69.00; post: median 46, IQR 32.00-77.00; P=.02; Box and Block pre: median 0, IQR 0-11.5; post: median 0, IQR 0-28; P=.005), and functional mobility (Time Up and Go pre: median 18, IQR 13.25-34.50; post: median 14, IQR 9.00-26.00; P=.005). No significant differences were recorded for general cognitive abilities (Mini-Mental State Examination pre: median 27, IQR 25-28; post: median 28, IQR 26-29), spasticity, and pain (visual analog scale pre: median 0, IQR 0-3.5; post: median 0, IQR 0-2). CONCLUSIONS: The study showed the preliminary feasibility of a rehabilitation program using the VSS. It addressed the essential topic of assessing VR-based rehabilitation user experience throughout the entire training period, shedding light on the features that can contribute to an optimal psychological experience. Clinical outcomes suggested that the VSS promoted neuroplasticity and that the recorded improvements could translate into meaningful functional gains in daily activities. Further studies with larger samples and patients with more severe disabilities are needed to confirm these results.