Participants were recruited from cohorts of stroke survivors attending a rehabilitation intervention at a cocreation center for accessible rehabilitation technology [21] between September 2021 and April 2023. The inclusion criteria for this study have been described in detail previously [21]; briefly, participants had to have had a stroke within the last 12 months that resulted in mobility problems, be aged over 18 years, be well enough to engage in light to moderate exercise, and be able to attend the rehabilitation program at least twice a week. A range of outcome measures were taken before and after the program, including the ARAT. An overview of the full rehabilitation program is available in our previously published report [21].

Out of a total of 36 participants who agreed to take part in the intervention, 7 (19%) were excluded from this study. Of the 7 excluded persons, 5 (71%) withdrew from the intervention (2/5, 40% withdrew before the commencement and 3/5, 60% withdrew due to ill health or unwillingness to continue), and the other 2 (29%) of the 7 were excluded as a result of incomplete data.

The upper limb intervention involved activities designed to improve the upper limb functions of participants, delivered completely through the use of technology and therapy devices that either stimulated or promoted repetitive and intensive movement training. The upper limb and hand rehabilitation technologies available to the participants in this study are shown in Table 1. The participants spent at least 50-60 minutes of each of the 2-hour sessions engaging with these devices.

Figure 2 is a representation of the upper limb rehabilitation program used in the rehabilitation gym. The activities were divided into 2 categories. The first part aimed at priming the brain to prepare it for plastic response [29]. Priming focused on sensory stimulation including mirror therapy and electrical, thermal, and vibrational stimulation. These priming activities comprised the first 15-20 minutes of each rehabilitation session. This second part, that is, the “active training,” aimed to engage the participants in high-intensity motor tasks such as object grip and release, object manipulation, and reach to grasp, designed to improve range of motion, strength, and control. The participants were not limited in terms of the number of devices they could use.

Participants were given a 1-day initial appointment with a therapist at the rehabilitation gym before the commencement of the 8-week rehabilitation intervention. During this appointment, demographic data including stroke history were collected, along with a range of baseline assessments for mobility, communication, and cognition, including the ARAT [30]. The ARAT was used to categorize the participants into 3 different hand function groups: poor (scores of 0-9), moderate (scores of 10-56), and good (score of 57) [30].

During the preintervention visit, participants were allowed to communicate their rehabilitation goals and interact with the upper limb devices to understand how they are set up and operated. The rehabilitation goals of the participants were summarized based on their different hand functions to help understand the needs of stroke survivors who fall under each of the different hand functions, particularly the hand rehabilitation goals of those with poor hand function.

Following the goal setting and initial interaction with the devices, a rehabilitation program was drawn up. The rehabilitation program was individually tailored by a physiotherapist using the rehabilitation goals of the participants. The program however only acted as a guide, as participants had the freedom to interact with any of the devices. The ability of the participants to use each rehabilitation device was observed and recorded. At the end of the intervention, all the observations from participants with poor hand function were gathered and studied to see how they interacted with the hand-based rehabilitation devices. Three of the upper limb devices—the GripAble, NeuroBall, and Semi-Circular Peg Board (see Figure 1)—were selected for observation in this study. The reason for selecting these devices is because these 3 devices were the only devices listed under the “active training” category (see Figure 2) at the time of the study that were used to primarily train motor activities in the hand (involving the wrist and fingers) in addition to training other parts of the upper limb.

The simple percentage method was used to estimate the percentage of stroke survivors who fall into each category of hand function. A 1-way ANOVA was carried out using Minitab statistical software (Minitab LLC), with the Dunnett multiple comparison method used to compare the ages of the group with poor hand function to those with moderate and good hand function.

This study was approved by the University of Strathclyde ethics committee (approval UEC 20/08). The participants provided written informed consent before the study, and their participation was voluntary (no compensation was provided). All identifiable data were pseudoanonymized and replaced with a code.

Observations from 29 participants were included in this study. Their average age was 59.10 (SD 13.62) years with an average of 3.140 (SD 2.31) years after stroke. Of the 29 participants, 17 (59%) were hemiplegic on the left side of their body, whereas the remaining 12 (41%) were hemiplegic on the right side of their body (Table 2).

Of the 29 participants, 10 (34%) scored between 0 and 9 on the ARAT and were grouped as having poor hand function, 17 (59%) scored between 10 and 56 on the ARAT and were grouped as having moderate hand function, and 2 (7%) scored 57 on the ARAT and were grouped as having a good hand function. There was no statistical difference in age between the poor hand function group and both the moderate hand function (P=.06) and the good hand function (P=.37) groups. Similarly, there was equally no difference in the years after a stroke between the poor hand function group and both the moderate hand function (P=.09), and the good hand function (P=.99) groups. There was also no observed difference in the hemiplegic side of those with poor hand function (left: 5/10, 50%; right: 5/10, 50%).

Table 3 shows a summary of the rehabilitation goals of stroke survivors based on their different hand functions. Participants with poor hand function stated goals that were more toward gaining movements in different parts of their upper limb, as well as improving the ability to carry out active movements that will enable them to grasp and release objects. However, stroke survivors with moderate and good hand function had goals that were focused on how to improve grip strength, fine motor movements, release time, as well as purposeful movement of the upper limb (see Table 3). Those with poor hand function who recorded a score greater than 0 on the ARAT equally communicated the need to improve grip strength.

Table 4 shows that 8 (80%) of the 10 participants with poor hand function could not interact with any of the 3 aforementioned devices to carry out active training. This value represents 28% (8/29) of the total population in this study. Only 2 (20%) of the 10 participants with poor hand function were able to engage with these devices; the ARAT score shows that these 2 participants had ARAT scores of 7 and 9, compared to the score of 0 that was recorded by the other 8 who were not able to engage with these devices.

This study was carried out to observe how stroke survivors with poor hand function interacted with hand rehabilitation devices such as the GripAble, NeuroBall, and Semi-Circular Peg Board. The findings show that stroke survivors whose poor hand function leads to an ARAT score of 0 cannot actively interact with hand rehabilitation devices.

About two-thirds (55%-75%) of persons who had a stroke sustain upper limb impairments [7]. The extent of the impairments varies from person to person (see Table 2). In some, it results in poor hand function, whereas others present moderate or good hand function. The level of hand function present after stroke subsequently influences the upper limb rehabilitation goals of the stroke survivor (see Table 3). Stroke survivors with moderate to good hand function, who are likely to possess some range of motion in the hand, can grip, grasp, or pinch [30,31] hand rehabilitation devices and so have upper limb rehabilitation goals aimed at strengthening the existing motor ability. These goals may be related to improving grip strength and endurance, the ability to release objects or release time, the existing range of upper limb movements, and finger dexterity and regaining the ability to manipulate small objects (see Table 3). However, those with poor hand function, especially those with an ARAT score of 0 who cannot grasp, grip, or pinch objects irrespective of the sizes [31], have upper limb rehabilitation goals that focus on recovering some movement in the joints (shoulder, elbow, wrist, and/or fingers; see Tables 3 and 4).

Muscle weakness and the appearance of muscle stiffness, tightness, or tone (evident by the presence of a clenched hand) were clinically examined as being responsible for the poor hand function of the participants in this study (see Figure 3). The appearance of clenched hands has been reported as a clinical feature of spasticity [32]; moreover, the presence of muscle stiffness, tightness, and tone have all been connected with spasticity [33,34]. Previous studies have reported both spasticity and muscle weakness as the 2 major motor impairments following a stroke [35,36]. The severity of these impairments led to difficulty in hand immobility in 80% of those with poor hand function (with an ARAT score of 0), and according to an earlier report [36], spasticity and muscle weakness can result in immobility.

The UK National Clinical Guideline for Stroke stipulates that stroke survivors should be considered for rehabilitation at any point after the stroke to potentially gain benefits [37]. However, an earlier study [38] that measured the accuracy of physical therapists’ early prediction of upper limb function reported that stroke survivors with ARAT scores more than 10 are those principally qualified to undergo rehabilitation exercises; this potentially excludes stroke survivors with poor hand function from taking part in hand rehabilitation. This study shows that not all stroke survivors with poor hand function should be considered ineligible to make use of hand rehabilitation devices, as those with some range of motion in their hand, as seen in participants with ARAT scores of 7 and 9 (see Table 4), can still benefit from hand rehabilitation devices and thus active hand rehabilitation.

Only participants who exhibited poor hand function with an ARAT score of 0 were not able to benefit from active hand rehabilitation using devices. Those in this category whose poor hand function was due to muscle weakness were unable to carry out any intended active movement on the hand rehabilitation devices (see Table 4), even when supported to place their hand on them. In contrast, those whose poor hand function was due to hand stiffness or tightness, in addition to their inability to carry out intended active movement, were also faced with the problem of accessibility, which made it difficult for them to fit the device.

A limitation of this study was the inability to assess these conditions (muscle weakness and muscle tone or tightness)—examined to be responsible for the poor hand function—using the relevant outcome measures, such as motricity index, grip strength or pinch strength (for muscle weakness), or the Modified Ashworth Scale (for spasticity) [39], to quantify their severity. However, their severity was such that the hand was not useful in carrying out any of the ARAT tasks [31], as indicated by an ARAT score of 0.

Improvement in technological advancement has led to the development of devices such as rehabilitation gloves (smart or robotic gloves) that can be useful in stretching the hands of stroke survivors with poor hand function without requiring their active participation [40,41]. However, only stroke survivors with low spasticity (who possess some range of active motion in the hand [42]) may be able to make use of these rehabilitation gloves [40]. This means those with considerable muscle stiffness resulting in difficulty in passive motion [42] are still unlikely to freely access these devices; thus, future design of rehabilitation devices for hand rehabilitation should consider the problem of device accessibility in people with poor hand function due to considerable muscle stiffness or tightness.

It is therefore concluded that not all stroke survivors with impairments in their hands can interact with the available hand rehabilitation technologies, as those with an ARAT score of 0 cannot actively interact with any hand rehabilitation device. Thus, the selection of devices for hand rehabilitation should first consider the hand function of the affected stroke survivor. Since muscle stiffness or tightness in the hand results in poor hand function that can impede access to hand rehabilitation devices, future design of devices for hand rehabilitation should consider the accessibility needs of those with poor hand function as a result of hand stiffness or tightness. A similar observational study involving more stroke survivors will help ascertain the percentage of stroke survivors who fall into the category of having poor hand function and is therefore recommended.