Defining Success at Higher Levels: What Constitutes a Successful Outcome in Upper-Limb Prosthetics?

Home > Articles > Defining Success at Higher Levels: What Constitutes a Successful Outcome in Upper-Limb Prosthetics?
By Phil Stevens, MEd, CPO, FAAOP
silhouette of people at conference table

Several years ago, I received a communication from an occupational therapist asking me to participate in a survey-based study. I didn't realize it at the time, but I was being snowballed into the research effort, as enrolled participants were asked to solicit the involvement of other colleagues with interest and experience in upper-limb prosthetics. The premise was a basic one that resonated well with my clinical encounters. Individuals who experience upper-limb amputations are generally different from their peers with lower-limb amputations. They are usually younger at the time of the amputation and experience a single, traumatic event in contrast to a long-term, debilitating disease process. Their amputations are harder to hide from the world around them than those of people with lower-limb amputations, and the attempts at prosthetic restoration are more challenging and often less successful. In contrast to the fairly predictable, cyclical patterns of use associated with lower-limb prostheses, upper-limb tasks require greater dexterity and adaptability. While some individuals with unilateral upper-limb loss acclimate to the limited functionality offered by prostheses, many do not, choosing unilateral upper-limb function instead. Finally, due in large part to their small numbers, research on upper-limb prosthetic users is scarce, often focusing on the devices rather than on the people who wear them.

Within these considerations of the challenges and limitations of upper-limb prosthetic care, the intent of the study was to better define what constituted a successful outcome in this population with respect to prosthetic use, activities and participation, and the user's self-image.1 The survey method was based on the Delphi technique in which a group consensus is pursued through a series of anonymous surveys among a panel of experts.

The Panel

The panel in this study was composed of two main groups. The first was rehabilitation professionals, described as either "researchers who had authored an article on upper-limb amputation, prosthetics, or rehabilitation published in the last ten years," or as professionals "working as part of a rehabilitation team concerned with individuals with upper-limb amputation for at least three of the past ten years." These professionals were initially identified through the published literature and conference attendance, and were then encouraged to pass the survey invitation to other eligible professionals-the snowballing that brought the survey to my attention.

The other group vital to the success of the panel was individuals with upper-limb amputations. That cohort was recruited through international support groups, private prosthetic clinics, and attendance at an annual conference for people with amputations, with a similar request to snowball other eligible participants into the research effort. The eligibility requirements of the second group included being at least 18 years of age, being at least one year post major upper-limb amputation, and having the ability to read and understand English.

The initial panel consisted of 73 respondents, inclusive of 47 rehabilitation professionals, 22 individuals with upper-limb amputations, and four individuals who met the inclusion criteria of both groups. A modest level of attrition, which will be described, occurred through the survey process.

The Procedure

There are many variations of the Delphi technique. For this study, in the first round, the panel was sent a series of statements regarding prosthesis use, activities and participation, and self-image relating to the attainment of a successful upper-limb prosthetic management outcome. Participants were asked to rate each of the statements on a Likert scale of agreement, where 5=strongly agree, 4=agree, 3=neutral, 2=disagree, and 1=strongly disagree. Consensus was defined prior to administering the survey as a standard deviation of responses of less than one, suggesting small variation in subject responses. If the item had a mean score of 4 or higher with a standard deviation of less than one, it was accepted as part of a successful outcome. If the item had a mean score below 4 with a standard deviation of less than one, it was considered that the consensus position of the group was that the item was unimportant to a successful outcome. If the standard deviation for a given item was greater than one, then wider variations in responses suggested disagreement between panel participants. As part of the first round of the survey, respondents were asked to also list any additional statements that they felt were important to a successful outcome but were not included in the initial survey.

Items that were accepted in round one as important to a successful outcome were dropped from subsequent rounds of the survey. Items that attained a consensus ranking of unimportant were presented again in round two, as were all those items that failed to reach agreement and any new statements that were suggested by a member of the panel in the first round of the survey. In the third and final round of the survey, study participants were asked once more to rank their levels of agreement with those items that had failed to obtain a consensus position in the previous surveys.

As with any iterative survey, some attrition occurred through each round of administration. Thus, 73 responses were collected during the first round, 58 during the second, and 53 during the third and final round. The 53 respondents who contributed throughout the three rounds of the survey included individuals from the United States, the United Kingdom, Canada, Australia, Japan, and the Netherlands. Professionals were predominantly occupational therapists and prosthetists, but also included engineers and psychologists, along with a physiotherapist, a surgeon, a social worker, and a few other allied healthcare professionals. These individuals had, on average, 13 years of experience.

Of the 13 participants with limb loss who responded to all three surveys, the average age was 55 with an average of 30 years since their limb loss. The etiologies of limb loss included trauma (seven), cancer (three), congenital absence (two), and infection (one). Amputation levels were also diverse, including transradial (seven), transhumeral (four), wrist disarticulation (one), and shoulder disarticulation (one).

The Results

By the third round, consensus had been obtained for 27 of the 30 items suggested as possible components of a successful outcome. Of these, 11 had a mean rating of 4 or higher and were accepted as part of a successful outcome. Sixteen of the items achieved consensus as not being required for a successful outcome, and only three items failed to ultimately obtain a consensus position. The 30 statements were organized within the three domains of prosthesis use, activities and participation, and self-image.

Prosthesis Use

Within the domain of prosthesis use, three items obtained consensus inclusion as part of a successful outcome. They were:

prosthesis success

By contrast, a number of statements indicating criteria for a successful outcome were ultimately rejected by the group. These included:

  • when a prosthesis is used for a person's preamputation job or activities
  • when a person is satisfied with the cosmetic appearance of the prosthesis
  • when a person uses the prosthesis for both functional and cosmetic purposes
  • when a person is content not to wear a prosthesis

Two additional statements also had mean values below 4, but because their standard deviations exceeded one, they could not be considered as having obtained a consensus position. These were:

  • when a person wears the prosthesis all day, every day
  • when a person feels the prosthesis is part of him- or herself

Taken collectively, these statements, both those that were accepted and those that were ultimately rejected by the panel, begin to provide patients and practitioners alike with reasonable expectations of the outcomes that are generally attainable with upper-limb prostheses. The lofty goals of reaching a state where the patient chooses to wear his or her prosthesis all day, every day, or integrates the prosthesis into his or her perception of self, while certainly desirable, were not considered reasonable requirements of a successful outcome. Reengagement in all preamputation jobs or activities was also not considered a reasonable requirement for a successful outcome. However, likewise, the opposite extreme of complete rejection of the prosthesis was not endorsed. Rather, a middle ground of empowering an individual to wear a prosthesis when he or she chose to on an activity-specific basis was endorsed by the panel as a key component of a successful outcome.

Activities and Participation

Further insights were gained by reviewing the proposed components of a successful outcome with regard to activities and participation. Here, six statements were accepted, four were rejected, and one failed to reach agreement. The statements that achieved consensus acceptance included the following:

prosthesis success

Rejected statements included the following:

  • a person's achievement of tasks set by rehabilitation professionals
  • a person's ability to perform activities to the same standard as before limb absence
  • a person's ability to perform activities within the same time parameters as prior to limb loss

While statistical agreement was not reached, the mean score of the following statement was also below 4, suggesting rejection, though without consensus:

  • when a prosthetic user can perform an activity bilaterally to an equal standard as a two-handed person

The picture of a successful outcome is further refined when considering these statements. Matching preamputation abilities and functional speeds were not endorsed by the panel as reasonable requirements. However, attaining basic levels of functional independence, including performing self-care, completing activities of daily living, and returning to some type of employment, were accepted. The more subjective elements of performing to the best of one's ability and finding satisfaction in the performance were also accepted.


The final domain of consideration was that of self-image. Two statements attained consensus acceptance and five attained consensus rejection. The accepted statements were:

prosthesis success

Statements rejected by consensus included:

  • when a person is confident to show his or her residual limb in public
  • when a person does not feel he or she stands out
  • when a person does not mind looking at his or her residual limb

While the panel had rejected the notion that patients must be satisfied with the cosmetic appearance of their prostheses, they endorsed the position that they should not feel self-conscious when in public and should possess a positive body image. Considered against the rejected statements, the panel's consensus position of a successful outcome is one in which a person may feel a sense of standing out, but does not feel self-conscious about his or her appearance. Further, having a positive body image need not require a person to show his or her residual limb in public or prefer not to look at the residual limb.


As a member of the panel that participated in this study, I've had access to this data for some time and have found it to be very helpful when counseling new patients and educating other medical professionals. The O&P profession's ability to manage upper-limb absence is currently limited. We are simply unable to adequately replace the limb segments that have been lost. Individuals new to the realm of prosthetic rehabilitation, including patients, family members, and novice medical professionals, may not be aware of these limitations. As such, their expectations of upper-limb prosthetic rehabilitation may not be realistic or attainable.

By contrast, the consensus positions of the panel in this study are generally consistent with these limitations and provide all parties with some idea of what constitutes a set of reasonable expectations. If these expectations are established early in the experience and training of patients and professionals alike, prosthetic rehabilitation is more likely to be valued, pursued, and utilized, its inherent limitations notwithstanding.

Phil Stevens, MEd, CPO, FAAOP, is in clinical practice with Hanger Clinic, Salt Lake City. He can be reached at .


  1. Nimhurchadha, S., P. Gallagher, M. MacLachlan, and S. T. Wegener. 2013. Identifying successful outcomes and important factors to consider in upper limb amputation rehabilitation: An international web-based Delphi survey. Disability and Rehabilitation 35 (20):1726-33.