EXPERIMENTAL INTERVENTION

The intervention group received usual practice rehabilitation (as described for the control group) and participated in repetitive training using the robotic device.

ReachMan was located in the rehabilitation gym .

The use of ReachMAN intervention had three distinct phases

1. Initiation of practice. The physiotherapist and engineer (Mr Che-Fai Yeong) worked with the patient to establish an appropriate physical set-up and task demand for the patient. They prepared a schedule and instructions that formed the basis of regular practice.

2. Establishing practice. The physiotherapist then worked with a rehabilitation assistant to set up the patient each time the patient wishes to practice, checking that the schedule clearly described how to set up both device and patient.

3. Maintaining practice. The rehabilitation assistant set up the device with patient when requested.

Intervention quality was monitored on a weekly basis by the physiotherapist.

The experimental invention had to fit around routine care and not interfere with therapy sessions, ward rounds, meal times, or medical investigations. In addition, when

participants were using the ReachMAN device their experience of fatigue, their

preferences for timing of using the device and the visits of their family and friends were taken into consideration. All these factors could potentially reduce the amount of time the subjects spent on the device. Consequently, the research therapist paid particular

attention to communication with clinical staff and participants regarding the flexible timing of intervention.

164

The aim was that subjects would have a maximum of 30 (5 week days over six weeks) half-hour robotic therapy sessions during the intervention period, but less if they were discharged earlier than six weeks.

During the robot rehabilitation session, the subject sat comfortably on a wheelchair or on a standard chair while the impaired hand was rested on the arm support. The height and distance of the robot were adjusted such that the elbow was flexed at 90⁰, shoulder abduction at about 35^o and shoulder flexion at 0⁰ (all with a 10⁰ tolerance). The hand grasped the robot handle and was secured with Velcro band. The sequence and number of sets were adapted to the performance of the subject, but all subjects started with the reaching exercise followed by pronosupination, grasping, then the combination of

pronosupination and reaching. Each sets contained 10 trials, and maximally 10 sets were completed for each exercise. However, if there was no movement at all, the particular exercise was limited to three sets. The first session was for the subjects to familiarize with the robot. Subjects could try the exercise modules freely without having their performance assessed. During the second and last sessions, subjects were assessed with the robot on all four exercise modules, using a standard difficulty level (Seven.) From the third session onwards, each subject progressed from lowest level of difficulty on each of the exercises.

Subjects could progress of maximally one level up on each day, if their score was over 90% with 10 hits thrice in a row.

ReachMan was programmed with eight difficulty levels (one the easiest up to eight the hardest). Each level had an increased range of motion and resistance to the movement. . At the lower levels participants were only required to put in relatively little effort to move.

For example, only 2N was required to trigger the reaching movement, 0.05Nm for the pronosupination and about 1N for grasping. However, in contrast level 8 required higher forces to trigger a movement, about 5N for reaching movement and 0.15Nm for

pronosupination.

The protocol followed for the intervention phase can be found in Appendix II . This was devised following piloting of the intervention in three subjects. This pilot has been reported in Yeong et al²⁴⁴.

165 6.4.3 TRUNK RESTRAINT

A modified form of trunk restraint was used with all participants when using the device (A photograph of this is shown in Photograph 6.1). Studies that have investigated normal reach to grasp pattern have found that stroke patients use compensatory trunk movements when reaching for an object within reach (as compared with normal controls who only use a trunk strategy when objects are placed beyond their reach).¹²³Trunk restraint (this is where the subject trunk movements are limited by a strap or belt) has been shown to result in an immediate improvement in active range of movement and inter-joint co-ordination in moderately to severely affected stroke subjects.⁴⁹

Most trials using robotic devices use some form of trunk restraint ( ie a harness attached to the patient) ^108;182(In the case of the MiTManus device this is in the form of a 3 point seat belt,¹⁹³ or two shoulder and a waist belt seat belts which limit trunk movement.)

The use of the non- assistive interface (p ReachMAn as described in Chapter 5) showed that subjects used compensatory trunk movements when using the device. Therefore to maximise arm movements alone a form of modified trunk restraint was used with all subjects who used ReachMan. A summary of this procedure with instructions as to how this was practically achieved can be found in Appendix II

Photograph 6.1-Trunk Restraint used in the study.

166 6.5. OUTCOME MEASURES

Chapter Four has described in detail the scale selection strategy used and selection of scales. Chapter Five has then discussed how these scales were tested for

appropriateness and reasons why some scales were not used in this exploratory trial. The study therefore used the following outcome measures: STREAM, CAHAI, ABILHAND, FMA, ARAT, EQ5D, Barthel Index, NIHSS and DASH

In addition to the scales used in Phase One, one additional scale was used in this phase of the study: the Ashworth score³¹³. The Ashworth score was used in the study as it was felt it would be useful to record any increased tone participants had on baseline and whether this changed with the treatment interventions. The score is the most popular and consistently used clinical measure of spasticity. ³¹⁴The most common version used of this scale in regards to the upper limb is the modified version, and this was the version used in this study.

167 The modified Ashworth score is a five point scale : 0 No increase in muscle tone

1 Slight increase in muscle tone, manifested by a catch and release or by minimal resistance at the end of the range of motion when the affect ted part(s) is moved in flexion or extension

1+ Slight increase in muscle tone, manifested by a catch, followed by minimal resistance throughout the remainder (less than half) of the ROM

2 More marked increase in muscle tone through most of the ROM, but affected part(s) easily moved

3 Considerable increase in muscle tone, passive movement difficult 4 Affected part(s) rigid in flexion or extension

Other studies looking at robotic intervention have used this scale to look at increased tone in the upper limb with this use of a robotic device.²⁰⁰ Tests of inter and intra-rater reliability have had conflicting results and the scale is unable to distinguish between the reflex and non-reflex components of increased tone. Despite the limitations with this measure, due to the ease of competition, and repeatability with other studies it was chosen for use in the exploratory trial.