Advanced Robotic Therapy Integrated Centers (ARTIC): an international collaboration facilitating the application of rehabilitation technologies

Hubertus J. A. van Hedel, Giacomo Severini, Alessandra Scarton, Anne O’Brien, Tamsin Reed, Deborah Gaebler‐Spira, Tara Egan, Andreas Meyer‐Heim, Judith V. Graser, Karen Sui Geok Chua, D. De Zutter, Raoul Schweinfurther, Jens Carsten Möller, Liliana Paredes, Alberto Esquenazi, Steffen Berweck, A. Sebastian Schroeder, Birgit Warken, Anne Chan, Amber Devers, Jakub Pětioký, Nam‐Jong Paik, Won‐Seok Kim, Paolo Bonato, Michael L. BoningerJournal of NeuroEngineering and Rehabilitation 2018 · DOI

TL;DR

This paper introduces the Advanced Robotic Therapy Integrated Centers (ARTIC) network, an international collaboration aimed at standardizing and advancing the science and clinical practice of robotic gait rehabilitation. By collecting data from a diverse patient population across multiple centers, ARTIC aims to provide real-world evidence on the application and effectiveness of robotic therapy, bridging the gap between controlled trials and clinical practice.

What Problem Does This Paper Address?

The field of neurorehabilitation has seen a significant increase in the use of robotic devices, particularly for gait training. While meta-analyses suggest benefits for some patient groups, the evidence remains less conclusive for others, such as those with spinal cord injury (SCI), traumatic brain injury (TBI), multiple sclerosis (MS), and cerebral palsy (CP). Several factors contribute to this uncertainty:

  • Small Sample Sizes: Individual studies often lack the statistical power to draw robust conclusions.
  • Variability in Protocols: Inconsistent training protocols, including duration, frequency, and specific exercises, make it difficult to pool results.
  • Strict Inclusion/Exclusion Criteria in Trials: Randomized controlled trials (RCTs) often select highly specific patient populations, limiting the generalizability of their findings to the broader patient population encountered in clinical settings. This "ecological validity" gap is a significant challenge for clinicians.
  • Lack of Real-World Data: There's a need for data that reflects how these technologies are actually used in diverse clinical environments with a wide range of patients.

ARTIC was established to address these issues by creating a collaborative network to gather standardized data on robotic gait therapy across various clinical settings and patient diagnoses.

Methods

ARTIC employs a pragmatic observational study design based on a practice-based evidence approach. This means the network collects data on how robotic therapy is currently being applied in real-world clinical care, rather than in a highly controlled experimental setting.

Key methodological aspects include:

  • Standardized Data Collection: A common database was developed to collect patient demographics, diagnosis-specific information, impairment severity indicators, core clinical assessments, and detailed training parameters for each Lokomat® session.
  • Common Technology: The network focuses on a single, widely used robotic device: the Lokomat® exoskeleton system. This system consists of a treadmill, a weight support system, and a driven gait orthosis for both legs, designed to guide patients through a physiological walking pattern.
  • Core Clinical Assessments: The 10-Meter Walk Test (10MWT) and Goal Attainment Scaling (GAS) are used as primary outcome measures. The 10MWT assesses walking speed, while GAS allows for individualized goal setting and measurement of progress.
  • Automatic Data Capture: Data from each Lokomat® training session (e.g., duration, distance, speed, bodyweight support, guidance force) are automatically collected, providing objective measures of training intensity.
  • Multicenter Collaboration: Data are gathered from numerous international centers, increasing the sample size and diversity of patient populations and clinical practices.

Datasets

The primary dataset is the ARTIC database, which prospectively collects data from patients undergoing Lokomat® therapy. At the time of analysis, the database contained data from 595 patients. The paper presents descriptive statistics for the following patient groups:

  • Cerebral Palsy (CP): n = 208
  • Stroke: n = 129
  • Spinal Cord Injury (SCI): n = 93
  • Traumatic Brain Injury (TBI): n = 39
  • Various other diagnoses: n = 126

This dataset is unique in its focus on real-world clinical application across multiple centers and diagnoses, differing from the highly curated datasets typically found in RCTs.

Key Insights and Ideas

  • Bridging the Evidence Gap: ARTIC's practice-based evidence approach is crucial for understanding how robotic therapies perform in diverse clinical settings, complementing traditional RCTs.
  • Standardization is Key: By focusing on a common technology (Lokomat®) and standardized assessments, ARTIC can generate more comparable and generalizable data.
  • Real-World Patient Characteristics: The study highlights how patient populations in clinical practice can differ from those in RCTs (e.g., higher prevalence of comorbidities, different age distributions), emphasizing the need for tailored treatment approaches.
  • Training Intensity Increases: Initial results show that training intensity (duration, distance, speed) generally increases from the first to the final Lokomat® session, suggesting patients adapt and progress.
  • Goal Attainment is Achievable: Despite initial functional deficits, a significant percentage of patients achieved their self-selected goals, underscoring the potential of robotic therapy.

Key Results

  • Patient Demographics: The database included 595 patients, with CP, stroke, SCI, and TBI being the largest diagnostic groups. Gender distribution varied by diagnosis (e.g., more males in SCI and TBI). Age distributions also differed significantly across groups.
  • Initial Walking Function: At the onset of therapy, average walking speeds were generally slow across all patient groups.
  • Training Progression: Training intensity, measured by duration, distance covered, and walking speed, increased from the first to the final therapy session.
  • Goal Achievement: A substantial proportion of patients achieved their goals, with percentages ranging from 56% (SCI) to 83% (TBI) for achieving at least their defined goal (GAS score ≥ 0).
  • Lokomat® Parameters: Guidance force showed less change during training compared to walking duration, distance, and speed, suggesting therapists might prioritize maintaining a physiological pattern while adjusting intensity.

Limitations and Future Work

  • Data Completeness and Standardization: While efforts were made to ensure data quality, initial data collection had some missing information, and continuous efforts are needed to improve completeness and standardization.
  • Lack of Control Group: As an observational study, ARTIC does not have a control group, making it difficult to definitively attribute improvements solely to Lokomat® therapy, especially since patients may receive other conventional interventions.
  • Generalizability of RCT Findings: The paper emphasizes that findings from RCTs with strict inclusion/exclusion criteria may not directly translate to the broader patient population treated in clinical settings.
  • Future Analyses: The authors plan more detailed analyses on specific patient groups, training parameters, and functional progress. They also suggest that the network can serve as a basis for future projects, including those investigating biomarkers or specific intervention outcomes, provided financial resources are available. The need for more detailed reporting of interventions in future studies is also highlighted.

Summary updated

Abstract

The original article [1] contains a small mistake concerning the ARTIC Team members mentioned in the Acknowledgements. The team member, Rocco Salvatore Calabrò had their name presented incorrectly. This has now been corrected in the original article.
BibTeX
@article{hedel2018advanced,
  author = {Hubertus J. A. van Hedel and Giacomo Severini and Alessandra Scarton and Anne O’Brien and Tamsin Reed and Deborah Gaebler‐Spira and Tara Egan and Andreas Meyer‐Heim and Judith V. Graser and Karen Sui Geok Chua and D. De Zutter and Raoul Schweinfurther and Jens Carsten Möller and Liliana Paredes and Alberto Esquenazi and Steffen Berweck and A. Sebastian Schroeder and Birgit Warken and Anne Chan and Amber Devers and Jakub Pětioký and Nam‐Jong Paik and Won‐Seok Kim and Paolo Bonato and Michael L. Boninger and Michael Boninger},
  title = {{Advanced Robotic Therapy Integrated Centers (ARTIC): an international collaboration facilitating the application of rehabilitation technologies}},
  journal = {Journal of NeuroEngineering and Rehabilitation},
  year = {2018},
  doi = {10.1186/s12984-018-0366-y},
}
Figure