Clinical application of simulators when learning to use upper limb prostheses in children

Children with an upper limb deficiency often choose to have fitted a prosthesis
to improve functionality. Despite all technical developments and improvements
in upper extremity prostheses, the rejection rate of the prosthetic devices is
high (Meurs, Maathuis, Lucas, Hadders-Algra, & van der Sluis, 2006; Postema,
van der Donk, van Limbeek, Rijken, & Poelma, 1999; Scotland & Galway, 1983).
The functional use of upper extremity prostheses is not only determined by its
function, the technical possibilities, but also by its functionality, the way
the amputee is able to handle the prosthesis. As has been shown, this latter
aspect can be enhanced by training (Carter et al., 1969; Lake, 1997; Weeks et
al., 2003). Our hypothesis is that prosthetic training can be improved by using
the concept of *transfer of learning*. Transfer of learning is the finding that
a motor skill learned at one side of the body transfers to the other side of
the body. As such, the level of skill of the untrained side of the body
improves. This concept has been extensively studied (Hicks et al., 1983; Karni
et al., 1998; Kumar & Mandal, 2005; Lee et al., 2010; Mier & Petersen, 2006;
Pereira et al., 2011), but not in the use of myo-electric upper limb
prostheses. As such, we propose to start prosthetic training in the unaffected
limb. With an upper limb prosthetic simulator, as developed earlier in
Groningen, the unaffected arm can be trained. With a prosthetic simulator it is
possible to mimic the effects of a myo-electric prosthesis because it*s
functioning is similar to the functioning of a prosthesis. A prosthetic hand
can be opened and closed with a motor driven by electrical signals that are
produced by muscle activation. The simulator is placed over the arm, and the
prosthetic hand is placed before the sound hand (see figure 1 of the research
protocol) and then operates in the same way as the prosthesis. The training
with the simulator is therefore comparable to the training with the
myo-electric prosthesis. By using the prosthetic simulator in this study it can
be shown whether the effects of transfer of learning are present in healthy
participants. Transfer of learning in adult amputees will be investigated in a
separate research project executed by our team. If the transfer of learning is
detectable in adults, the next step is to reveal if these effects are also
present in children. Until now it is unclear what the effects and application
of transfer of learning in children are. Literature states that in children the
transfer of learning increases from 5 years old on (Byrd et al., 1986; Parlow &
Kinsbourne, 1989; Uehara, 1998). The first step, therefore, is to reveal if the
transfer of learning effects are present in children between 5 and 7 years old.
By using able-bodied children we do not need to bother children with an
amputation with research and we will have the ability to study more
participants. It is important that we study transfer effects in the execution
of a complex task. The results of the study in healthy children will give
information about the development of transfer of learning at a young age.
Thereby, when the effects are found to be present, this can have important
consequences for children with acquired upper limb amputation or a congenital
upper limb deficiency.
In case of congenital deficiencies there is no need to start prosthetic
training within a short period of time. However, the transfer of learning
effect might also be beneficial to children, because if transfer of learning
takes place, the child can start at a higher functional level when the actual
prosthesis is delivered. Apart from that, the training with a simulator using
the unaffected arm provides information on the capacities of a child to handle
a myo-electric prosthesis. As a consequence, the rehabilitation process might
be shorter and more satisfactory to the child, which may diminish the rejection
rate of these expensive devices.

The objective of this study is to establish the effects of transfer of learning
in children with upper limb deficiencies after practising with an upper limb
prosthetic simulator.

In children, transfer of learning effects have hardly been studied. The scarce
literature, which is mainly based on the execution of simple tasks, states that
transfer of learning can only be detected from 5 years on (Uehara, 1998). While
the child gets older the transfer of learning effects seem to improve (Byrd,
Gibson, & Gleason, 1986; Parlow & Kinsbourne, 1989).
Prosthetic simulators have never been used in children before. Training with a
simulator before the actual prosthesis is delivered may have the advantage of
the transfer of learning effects, which is subject of the present study.
Moreover, using the simulator might give children and their parents an idea of
the possibilities of a real myo-electric prosthesis. Furthermore, when a child
trains his unaffected arm with a simulator, the rehabilitation team gets an
impression of the capacities of the child to handle a myo-electric prosthesis.
If the child*s capacities are adequate, the team has more evidence that a
myo-electric prosthesis is the right choice for the child. This can reduce the
high rejection rate (Meurs, Maathuis, Lucas, Hadders-Algra, & van der Sluis,
2006; Postema, van der Donk, van Limbeek, Rijken, & Poelma, 1999; Scotland &
Galway, 1983) in children.
With this project the first step is made by analyzing the transfer of learning
effects in able-bodied children.

There is a pre-post design used, where half of the group trains in between (see
table 1 of the research protocol).

To measure the transfer of learning effects, a pretest, posttest and a
retention test will be performed. At the start of the experiment, the
functional level of the *test hand* (which resembles the affected hand) will be
measured in all children (pretest). The children are pseudo-randomly divided in
a control and an experimental group, in such a way that the amount of boys and
girls is equal in both groups. Then the children that form the experimental
group will train the *training hand* (which resembles the unaffected hand)
using the simulator. The other half of the children comprises the control
group. These controls do not train with the simulator. After five days, the
functional level of the *test hand* will be measured again in all experimental
and control children (posttest). After one week, all children will be measured
again to evaluate the consolidation of the transfer of learning effects
(retention test). With a retention test, the lasting of the learning effects
can be determined (Schmidt & Lee, 2005).

The training will be performed over five days to mimic a real learning process
in rehabilitation. It is found that the distribution of training over several
days is important for the consolidation of learning (Park & Shea, 2003;
Savion-Lemieux & Penhune, 2005; Siengsukon & Boyd, 2009). Furthermore, Pereira
et al. (Pereira et al., 2011) found improvement in complex dexterity skills in
the untrained limb after practicing with low intensity (20 min a day) spread
over 5 days. As such, training on several days seems to be important for the
transfer of learning as well.
For the training the Southampton Hand Assessment Procedure (SHAP) will be used
(see also paragraph 7.3 from the research protocol). With the SHAP we can offer
a standardized training procedure with sufficient variation in tasks (26) in
order to be able to promote strong learning (Schmidt & Lee, 2005). This
variability will improve skill acquisition, retention and transfer (Stokes et
al., 2008). Half of the children will train their dominant hand and half will
train their non-dominant hand.

The burden is expected to be minimal. The children will execute three tests and
half of them is training for five days. It is expected that the children like
to execute the tests and training. Execution of the test and the tasks from the
SHAP with the prosthesis simulator does not have any risks.