A Study Report on the Impact of Task Complexity on Digital Health Metrics Evaluation of Upper Limb Function

Research Background

In patients with neurological diseases such as multiple sclerosis (PWMS), the prevalent upper limb dysfunction significantly affects the completion of daily living activities, increasing dependence on caregivers. To enhance our understanding of the types and potential mechanisms of upper limb dysfunction, and to provide sensitive and reliable endpoints for evaluating the efficacy of pharmacological or rehabilitation interventions, it is essential to use effective evaluation tools in clinical studies. Currently popular and easy-to-use evaluation tools, such as ordinal scales describing movement quality or recording the time to complete functional tasks, have limitations like ceiling effects and low sensitivity. Thus, the research community widely acknowledges the urgent need for novel, more complementary, and sensitive evaluation endpoints to provide detailed insights into upper limb dysfunction mechanisms and the impact of therapeutic interventions.

Research Source

This study was conducted by Christoph M. Kanzler, Tom Armand, Leonardo Simovic, Ramona Sylvester, Nadine Domnik, Antonia M. Eilfort, Carola Rohner, Roger Gassert, Roman Gonzenbach, and Olivier Lambercy and published as an original research article in the Journal of Neuroengineering and Rehabilitation, 2024, Volume 21, Page 125 (DOI link).

Research Objectives and Methods

The researchers aimed to investigate the impact of virtual reality (VR) and task complexity on the evaluation characteristics of digital health metrics describing upper limb function. The study relied on the Virtual Peg Insertion Task (VPIT), a haptic VR-based assessment involving virtual manipulation tasks. To evaluate the effects of VR and task complexity, the researchers designed two new tasks derived from VPIT—VPIT-2H (a VR environment with simplified task complexity) and PPIT (a physical task with simplified task complexity). The study included 27 healthy participants and 31 multiple sclerosis patients, and through an observational longitudinal study, compared the kinematic and dynamic metrics of these tasks, their clinical measurement properties, and the usability of the evaluation tasks.

Research Results

Results indicated that task complexity had a significant impact on participant variability, increasing significantly with complexity (coefficient of variation increased by 56%), and variability was 27% higher in VR environments compared to physical environments. Higher task complexity did not lead to significant differences in measurement error and test-retest reliability. Notably, in multiple sclerosis patients, the responsiveness to longitudinal changes was significantly higher in complex VR tasks compared to simple physical tasks. Additionally, patients rated the usability of PPIT higher than VPIT.

Research Conclusions

The study demonstrated that both VR haptic-based and physical task evaluations could provide metrics with appropriate clinical measurement properties. For longitudinal assessments in multiple sclerosis patients, VR haptic-based evaluations may be advantageous due to their enhanced responsiveness, while physical task evaluations might be more suitable for routine clinical use due to higher usability. These findings underscore the necessity to further validate the effectiveness of VR and physical task evaluations in practical use cases.

Research Highlights

This study provides new insights into the evaluation characteristics of digital health metrics for upper limb function through haptic VR and physical conditions, altering the perception and evaluation paradigm of upper limb movement metrics. The study finds that increasing task complexity can maintain good clinical measurement properties, although it impacts movement kinematics and variability. Furthermore, for specific clinical applications, VR environments and task complexity have different levels of importance on the clinical measurement properties of digital health metrics.