Robotic neurorehabilitation covers several kinds of electromechanical system used to help or resist movement, support body weight, guide the limbs, or record data from the device. Gait trainers, wearable robotic frames, robots for training the arm, and positioning devices do not have the same purpose or the same evidence.
These devices may make certain training repeatable and adjustable in level, but they do not treat the nervous system, do not guarantee the adaptation of the nervous system or recovery, do not predict the result, and do not take the place of clinical staff. The value depends on the diagnosis, the functional goals, the device, the treatment used for comparison, the participation of the patient, and the other parts of the rehabilitation programme.
The important points
Robots are one instrument in restoring the nervous system, not the whole treatment. The decision should begin from meaningful tasks, such as transferring, standing, walking, reaching, or grasping, and then ask whether the device adds training that is safe and in which the patient participates more than a suitable non-robotic alternative.
No device guarantees walking by oneself, the return of the use of the hand, returning to work, or a definite timeline. The number of repetitions and the graphs from the machine are data about the process, not evidence of being able to function in real life.
The evidence differs according to the task and the diagnosis
After cerebrovascular disease, the evidence and the recommendations do not support every kind of robot. The NICE guideline on rehabilitation after cerebrovascular disease (opens in a new tab) recommends repetitive task training and gait training where suitable, but does not recommend using robots to help train the arm as part of an arm rehabilitation programme, because increased strength does not consistently translate into better use of the arm or better daily routines when compared with therapy of a similar intensity.
For walking after cerebrovascular disease, the Canadian Stroke Best Practice Recommendations (opens in a new tab) do not recommend robotic gait training devices over general gait training. Summaries of the evidence state that some patients may walk by themselves more at the end of treatment that combines robots with physiotherapy, but the long-term benefit, the capacity to walk, and superiority over general training given in a similar amount are still uncertain.
The recommendations for other neurological diagnoses may differ. The NICE guideline for chronic neurological disease (opens in a new tab) states that robotic treadmill gait training may be considered where the equipment is available, as part of care of movement that sets goals. That guideline does not cover the rehabilitation of adults after cerebrovascular disease, so evidence from one diagnosis, device, or outcome should not automatically be applied to another.
The main rehabilitation is wider than the device
The rehabilitation plan should care for the activities and health problems that matter to the patient. According to necessity it may include specific task training in which the patient exerts effort, walking on the ground, transferring, balance, strength and endurance, the use of the arm, daily routines, communication, swallowing, cognition, vision, mood, fatigue, continence, equipment, training the carer, and preventing a further cerebrovascular event.
Training with robots should supplement, not crowd out, these. The assisting force should be reduced or adjusted where possible so that the patient exerts their own effort. Skills trained in the machine must be carried over to real tasks and environments. Movement that is smoother when the machine guides it does not always mean that walking will be safe or that the arm will be used better outside the training.
Patients and devices must be selected individually
Selection begins from a stable diagnosis, clear goals, and baseline functional data. The team may need to assess the control of movement, the range of the joints and contractures, pain, spasticity, balance, the skin and wounds, the strength of the bones and joints, body size and fit with the device, cardiopulmonary endurance, the blood pressure response, cognition, communication, vision, behaviour, fatigue, seizures, and any relevant medical lines or tubes.
The instructions for use of the device, the limitations of fit, and the number of trained staff are also important. A diagnosis or a label of severity alone does not confirm suitability. Those who do not suit one kind of robot may be able to use another approach, while those who can be fitted into a device may not gain a meaningful benefit from that device.
There is no single standard robotic programme
The amount of training, the assisting force, the resistance, the speed, the support of body weight, the rest periods, and the increase in level should be chosen from the goals, the quality of the movement, the effort the patient exerts, the symptoms, and the response. More repetitions are not automatically better if the machine is doing most of the work instead, the alignment of the joints is not suitable, or fatigue reduces the quality.
The team should set criteria for stopping and for review before beginning. If the function does not change, the goals, the settings of the device, the whole programme, and the diagnosis should be assessed again, instead of following a fixed course merely because it has been bought or booked.
The risks and the criteria for stopping
The harms that may arise are fatigue, pain in the muscles or the joints, pressure or friction on the skin, discomfort from the support straps, abnormal joint alignment, increased spasticity, dizziness, changes in blood pressure or heart rate, falls or collisions, and device malfunction. Some neurological diseases carry added risks, such as acute autonomic dysreflexia, or fractures in those whose bones are brittle. The device's manual and clinical assessment should be the main basis for determining the precautions, not general lists online.
Stop the training and assess when there is new or great pain, dizziness, feeling faint, difficulty breathing, chest discomfort, palpitations, a severe headache, new weakness or numbness, abnormal spasm, injury to the skin, the device slipping, or being unable to maintain safe body alignment. Trained experts should supervise, know the method of emergency release, and have a plan for transferring safely or dealing with a fall.
Consent, costs, and data from the device
Informed consent should explain the specific goals, the additional benefit expected, the uncertainties, the reasonable general alternatives, the risks specific to the device, who supervises, how the assisting force is adjusted and the decision to stop is made, and what will be done when the machine malfunctions. It should also cover the costs, conflicts of interest, and the conditions of a package that may limit stopping.
Some systems collect performance or health data. It should be asked what is collected, whether it is used for treatment or for research, who can access it, how long it is kept, and whether refusing to allow the data to be used for secondary purposes affects the treatment. Data from the device may help in reviewing the result, but must be interpreted together with the function that can be observed and the patient's experience.
How should progress be judged
Measure outcomes so that they match the goals, such as the help needed in transferring or walking, safety, speed or endurance, falls, using the arm to do tasks, daily routines, participation, pain, and fatigue. Record adverse effects and check whether the progress transfers to real conditions.
Changes in force, symmetry, or the number of repetitions shown on the machine may be useful, but are not enough by themselves. Continue, adjust, or stop according to meaningful benefit, the burden, and what the patient regards as important.
New neurological symptoms are an emergency, not a matter of rehabilitation
The stroke signals of the United States Centers for Disease Control and Prevention (opens in a new tab) are sudden weakness or numbness of the face, the arm, or the leg, abnormal speech or understanding, changed vision, loss of balance or coordination, and a sudden severe headache. Call the local emergency service immediately even if the symptoms improve. Do not test the symptoms by continuing the training.
Chest pain, severe difficulty breathing, loss of consciousness, a fall with possible injury to the head or the neck, a hot swollen calf, fever together with a new wound or skin injury from the device, or rapid deterioration of unknown cause must also receive an urgent assessment.
In summary
Robotic systems may help certain training under supervision, but the outcome depends on the patient, the diagnosis, the device, the treatment used for comparison, and the whole rehabilitation plan. Current guidelines do not support presenting robots as superior to general treatment after cerebrovascular disease, and results should not be generalised across devices or diseases. Responsible use maintains the main rehabilitation in which the patient participates and which is specific to the task, screens fit and risk individually, obtains consent specific to the device, measures real function, and stops immediately when harm or new neurological symptoms arise