Prosthesis receipt is associated with improved participation and pain interference following upper limb amputation

Abstract

Purpose

Following upper limb amputation (ULA), resuming everyday activities is a primary goal for patients. In many cases, a prosthesis can assist patients by increasing physical function and overall well-being. This study aimed to investigate the relationship between prosthesis receipt and two domains – Ability to Participate in Social Roles and Activities (APSRA) and Pain Interference (PI).

Materials and methods

A retrospective cross-sectional review of outcomes collected across private prosthetic clinics in the United States was performed. Individuals were included if they had an outcome collected before prosthesis receipt and within twelve months after prosthesis receipt. There were no exclusions based on level of amputation, etiology, or device type.

Results

Univariate model results demonstrated a significant improvement in average APSRA (Baseline: 41.5±8.2 and Follow-up: 46.7±9.1, p<0.001). Additionally, PI significantly improved (Baseline: 58.5±8.5 and Follow-up: 54.9±9.3, p<0.001). These significant differences persisted when accounting for confounders such as age, hours worn, sex, amputation level, time since amputation, and time from first prosthesis delivery to follow-up.

Conclusions

Prosthesis receipt improved individuals’ ability to increase their activity and participation. This coincided with improvements in pain interference. Results from this study can aid clinical decision-making for clinicians and patients as they explore interventions following ULA.

Introduction

An upper limb amputation (ULA) can affect many aspects of one’s health including participation in activities of daily living (ADLs), social roles, and working roles.^1,2 For those with ULA, participation in ADLs can be difficult, especially tasks that require more fine motor movements of the terminal device. ³ Individuals who do not use a prosthesis not only reported more difficulty with activities, but also required additional assistance with ADLs compared to those who used a body-powered prosthesis. ⁴ Individuals with ULA also report concerns about pain, over-use injuries of their unaffected side, psychological and emotional distress, and restriction with participation in social roles and activities. ⁵ Social participation is crucial to overall health. As in the general population, for individuals with disabilities, decreased participation in social activities was found to be correlated with greater mortality and decreased quality of life. ⁶ Returning to work after an amputation can also be a challenge, as most people change jobs after their amputation, modify their work responsibilities, or become unemployed.^3,7 For persons with ULA, the use of a prosthesis has been found to improve social participation, quality of life, and physical function.^8–10 Although there is compelling evidence of how upper limb prosthesis use can positively impact an individual’s participation, these studies do not explore the impact of a prosthesis in the early phase of rehabilitation, shortly after receiving their first prosthesis.

Conversely, pain levels negatively affect one’s well-being, including limiting one’s ability to participate in work and social activities.^11–13 Pain is a significant factor for individuals’ post-amputation, as a majority of those with ULA experience phantom pain and residual limb pain.^14–16 One study reported that out of 104 respondents with ULA, 90% reported pain in the first six months following amputation. ¹⁵ Use of an upper extremity prosthesis may help alleviate pain through increased use of the residual limb or activation of a myoelectric prosthesis^17,18; however, the literature is mixed with some studies reporting decreased pain while others report increased pain.^17–19 Pain can influence participation; a previous cross-sectional retrospective analysis of patients with ULA found that pain interference and social participation are strong predictors of general well-being. ²⁰

While there are benefits to using an upper extremity prosthesis, the usage of these devices varies. Østlie et al. ¹⁰ found that 90.3% of individuals with ULA reported their most worn prosthesis as useful. Most individuals stated they used their prosthesis in approximately half of daily ADLs. Those that wear their prosthesis, regardless of control type (i.e., both body-powered and electric), reported their prosthesis was well-used for social activities and at work or school for upwards of 8 hours a day. ²¹ However, some literature is mixed where amongst a sample of 307 males with an ULA, most said they use their prosthesis for aesthetic restoration, and an eighth of these participants were not satisfied with their prosthesis function. ³ In the same study, only 23.5% were either somewhat or highly satisfied with ADLs they completed independently. Additionally, there is evidence of some discontinuation of prosthesis use amongst those with ULA. Over a 25-year period, about 20% of upper limb prosthesis users discontinued the use of their device. ²¹ In a more recent study, out of 244 individuals with ULA, 13.4% discontinued use of their prosthesis. ¹⁰ The primary reasons for prosthesis discontinuation are decreased comfort and dissatisfaction with function. ²²

Understanding the intersection of prosthesis use, pain, and participation is needed. Comprehending the longitudinal impact of receiving a first prosthesis, specifically in terms of participation in social roles and pain, is vital for both clinicians and patients to ensure the best clinical outcome and general well-being as patients start their prosthetic care journey. Therefore, the purpose of this study is to determine if there are significant improvements in self-reported social participation (measured by the Patient Reported Outcomes Measurement Information System – Ability to Participate in Social Roles and Activities: PROMIS-APSRA) and pain interference (measured by the Patient Reported Outcomes Measurement Information System – Pain Interference: PROMIS PI) from before to after first prosthesis receipt for those with ULA.

Materials and methods

Study design and population

A retrospective analysis of longitudinal outcomes collected between 2018 and 2023 across private Orthotic and Prosthetic (O&P) companies across the United States was conducted. Participating clinics serve patients across a wide geographic range across the United States and from a diverse range of insurance coverage, representing a range of socioeconomic backgrounds. All outcomes were collected during routine clinical care. This database review was approved by the Sterling Institutional Review Board (protocol #2786) and designated to qualify for exempt status. Study results are presented according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines (see Supplementary Checklist).

Patient data was included in the final sample if patients were prescribed a prosthesis and had two outcomes: one prior to prosthesis receipt (baseline) and one between two weeks and twelve months of prosthesis receipt (follow-up). For individuals with more than one follow-up outcome within 12 months, the outcome closest to prosthesis delivery was used. There were no exclusions based on the level of ULA, device type, or etiology. Amputation levels were stratified into four functional levels: Elbow Disarticulation/Transhumeral, Wrist Disarticulation/Transradial, Partial Hand/Partial Finger (PHPF), and Bilateral.

Outcome measures

The Patient Reported Outcome Measurement Information System (PROMIS) outcome measures ²³ were used to measure both social participation and pain interference. PROMIS measures have demonstrated strong psychometric properties across diverse clinical populations supporting their use in research and clinical practice. ²³ Both the PROMIS APSRA and PI outcome measures have been previously used and validated for rehabilitation tracking of musculoskeletal patient populations, making them appropriate for this study.^24,25 The 4-item PROMIS v2.0 APSRA 4a and a single question from the PROMIS v1.1 Pain Interference (item PAININ9) were used to measure participation and pain, respectively. Social participation from the APSRA is defined as the perceived ability one has to engage with their family, friends, and other activities that are meaningful. Pain Interference is defined as the degree to which pain disrupts a person’s ability to perform daily activities and participate in day-to-day life. Interference is distinct from the presence or intensity of the pain.

The PROMIS APSRA 4a evaluates a patient’s ability to perform family duties, leisure activities, usual work, and activities with friends. The APSRA was scored using HealthMeasures.net scoring services, which uses response patterns scoring to account for item difficulty and missing data. The resulting T-score describes a patient’s abilities where 50 represents the average APSRA score for the general population and every 10 points corresponds to one standard deviation. Individuals with a higher APSRA score report higher ability to participate in social roles and activities.

The PAININ9 item, “How much did pain interfere with your day-to-day activities?” from the PROMIS PI item bank was similarly scored using HealthMeasures.net scoring service. The resulting score was converted to a T-score with a higher score corresponding to worse pain interference.

Statistical analysis

Descriptive statistics (frequency, mean, standard deviation) were used to describe patient demographics. The data was first checked for normal distribution. A paired t-test was used to compare APSRA and PI baseline vs. follow-up scores. Two multivariate linear mixed effect models were used for APSRA and PI to control for any confounding effects such as age, hours worn, sex, amputation level, time since amputation, and time from first prosthesis delivery to follow-up.

Results

The final sample size included longitudinal data from 124 individuals with ULA (Table 1). The sample was predominantly male (73.4%) and had their amputation due to injury/trauma (62.9%). Level of amputation was mixed with 14.5% Elbow Disarticulation/Transhumeral, 22.6% Wrist Disarticulation/Transradial, 57.3% PHPF, and 5.6% Bilateral. Among patients that reported prosthesis type (n=62), 77.4% (n=48) reported wearing a body-powered prosthesis and 22.6% (n=14) reported wearing an electronic/myoelectric device.

Table 1.

Clinical characteristics at the time of assessment (n=124).

Patient demographics	Mean ± SD
Age (yrs)	43.4 ± 15.0
Height (m)	1.8 ± 0.2
Weight (kg)	85.9 ± 20.3
Time since amputation (months): all	34.4 ± 103.0
Time from first prosthesis to Assessment 2 (months): all	3.4 ± 3.2
Hours worn Daily	6.2 ± 4.9
​	n (%)
Sex
Female	33 (26.6)
Male	91 (73.4)
Amputation Level
Elbow Disarticulation/Transhumeral	18 (14.5)
Wrist Disarticulation/Transradial	28 (22.6)
Partial hand/Partial Finger	71 (57.3)
Bilateral	7 (5.6)
Cause of Amputation
Vascular	5 (4.0)
Injury/Trauma	78 (62.9)
Infection	13 (10.5)
Congenital	3 (2.4)
Cancer/Tumor	5 (4.0)
Other/Not Reported	20 (16.1)
Device Type
Body Powered	48 (38.7)
Electronic Arm	14 (11.3)
Not Reported	62 (50.0)
Payer
Medicare	8 (6.5)
Medicaid	22 (17.7)
Private	33 (26.6)
Veteran Administration	5 (4.0)
Worker’s Compensation	52 (41.9)
Unreported	4 (3.2)

Comparing baseline to follow-up, the average APSRA score improved from 41.5±8.2 to 46.7±9.1 (Figure 1) and the average Pain Interference improved from 58.5±8.5 to 54.9±9.3 (Figure 2). Time since amputation and amputation level were significant predictors of improved APSRA scores. The time since amputation was on average 3 years with a large standard deviation (103.0 months), reflecting the inclusion of both individuals that recently experienced an amputation and those with long-standing amputations. When included as a covariate in the linear mixed effects model, longer time since amputation was associated with greater improvements in self-reported social participation scores (Table 2). With respect to level of amputation, those with PHPF amputation compared to Elbow Disarticulation/Transhumeral had a greater effect on improved APSRA scores (Table 2). Individually comparing the four amputation level groups, all groups reported improved APSRA T-scores, with the PHPF and Bilateral groups reporting the biggest improvements (Table 3). In the case of PI, where changes were reported in response to the PAININ9 item, the changes tended to be from “quite a bit” and “somewhat” to “a little bit” and “not at all” (Table 4).

Figure 1.

Ability to Participate in Social Roles and Activities (APSRA) at baseline and follow-up. Participation significantly improved at follow-up (p<0.001). Error bars represent the standard error.

Figure 2.

Pain interference scores at baseline and follow-up. Pain interference significantly improved at follow-up (p<0.001). Error bars represent the standard error.

Table 2.

Linear mixed effect model for APSRA and Pain outcomes (n=124).

​	APSRA T-score	Pain T-score
	Estimate [95 %CI]	Estimate [95 %CI]
First Prosthesis Intervention
Before	ref	ref
After	5.22 [3.81, 6.63]*	-4.04 [-5.58, -2.50]*
Hours Worn	0.27 [-0.01, 0.56]	-0.23 [-0.54, 0.07]
Age (years)	-0.07 [-0.16, 0.01]	0.06 [-0.03, 0.16]
Sex
Female	ref	ref
Male	-0.90 [-3.92, 2.11]	-0.75 [-4.22, 2.72]
Time from first prosthesis to assessment 2 (months)	0.03 [-0.38, 0.43]	0.06 [-0.39, 0.51]
Time since amputation (months)	0.01 [0.00, 0.03]*	-0.01 [-0.02, 0.01]
Amputation level
Elbow Disarticulation/Transhumeral	ref	ref
Wrist Disarticulation/Transradial	-0.64 [-5.06, 3.77]	1.06 [-3.99, 6.12]
Partial hand/Partial Finger	3.89 [0.15, 7.64]*	-2.54 [-6.81, 1.72]
Bilateral	0.64 [-5.71, 6.99]	2.00 [-4.89, 8.89]

*- significant at p<0.05.

Table 3.

Average APSRA T-scores by amputation level.

Amputation level	APSRA T-score: Baseline	APSRA T-score: Follow-up
Elbow Disarticulation/Transhumeral (n=18)	40.6 ± 9.7	44.0 ± 11.8
Wrist Disarticulation/Transradial (n=28)	41.4 ± 7.8	44.2 ± 8.3
Partial hand/Partial Finger (n=71)	43.0 ± 7.9	49.0 ± 8.2
Bilateral (n=7)	37.4 ± 8.5	48.4 ± 9.4

Table 4.

Changes in the prevalence of Pain Interference PAININ9 item response options and PI T-score by amputation level from baseline to follow-up. A positive percent change in the “not at all” or “a little bit” response options indicates an improvement (less pain interference).

Amputation Level	PI response					PI T-score
	Not at all	A little bit	Somewhat	Quite a bit	Very much	Baseline	Follow-up
Elbow Disarticulation/Transhumeral (n=18)	2%	1%	-2%	-1%	0%	59.8 ± 9.9	57.3 ± 11.1
Wrist Disarticulation/Transradial (n=28)	2%	1%	-4%	1%	0%	60.0 ± 8.1	58.1 ± 9.8
Partial hand/Partial Finger (n=71)	12%	0%	-2%	-6%	-3%	58.2 ± 8.5	53.3 ± 8.8
Bilateral (n=7)	-1%	2%	1%	-1%	-1%	61.2 ± 9.0	59.6 ± 3.8

Discussion

The goal of this study was to explore the impact of first prosthesis receipt on social participation and pain interference during routine clinical care. Results demonstrated that individuals who received a prosthesis report significantly improved pain interference and social participation. Specifically, PI decreased by 4.0 points and APSRA scores increased by 5.2 points in the linear mixed effects model. At baseline, participants reported an average APSRA T-score of 41.5 which is almost one standard deviation below the general population mean, placing them in approximately the bottom 15^th percentile for this construct relative to the general population. After prosthesis receipt, patients are closer to the average of the general population (approximately 40th percentile). The increased APSRA score is reflective of potential increased engagement with the prosthesis to further participate in leisure activities, family duties, activities with friends, and responsibilities at work. This increased social participation is vital for maintaining an overall high quality of life^6,24,26 and community re-integration following a major life event such as amputation. These improvements remained significant after controlling for age, sex, hours worn, amputation level, and time since amputation.

While no minimal clinical important difference (MCID) has been established for the PROMIS APSRA or PI measures in individuals with upper limb amputation, a referenced threshold for minimal important change in PROMIS Pain Interference T-scores is 3.2 points. ²⁵ The observed improvements in both APSRA (+5.2 points) and PI (−4.0 points) exceed this threshold, suggesting these changes may be clinically meaningful in addition to being statistically significant. Further research is needed in this area to fully realize the clinical impact and clinical interpretation should consider the individual patient’s experience when interpreting these results.

Among these covariates, longer time since amputation and more distal amputation were the only significant predictors of social participation. While the study is cross-sectional and does not follow individuals longitudinally, the inclusion of time since amputation as a covariate provides insight into an important demographic pattern. Previous research has indicated that longer time since amputation is related to greater psychosocial adjustment. ²⁷ Therefore, patients that have a longer lived experience with their amputation may have less depressive symptoms or other psychological symptoms leading to improved self-reported social participation after receiving a prosthesis. ²⁷ In addition, patients with more proximal levels of amputation (transhumeral and shoulder disarticulation) reported a lower extent of prosthesis adoption (i.e., used their prosthesis less often) ²⁸ which may explain the greater increase in participation for patients with PHPF amputation. In addition, for patients with transhumeral or shoulder disarticulation, other factors such as the fewer componentry options available for these patients, weight of the device, and function may be limited and also explain the larger improvements in participation scores in the PHPF group.

Specifically, patients with PHPF amputation showed an additional 3.89-point increase in social participation scores compared to patients with Elbow Disarticulation/Transhumeral, based on the linear mixed model. From baseline to follow-up, the mean APSRA T-scores changed from 43.0±7.9 to 49.0±8.2 for individuals with PHPF amputation and from 40.6±9.7 to 44.0±11.8 for individuals with Elbow Disarticulation/Transhumeral amputation (Table 3). These findings are consistent with prior literature reporting that individuals with PHPF amputation experience significant disability related to hand appearance and its impact on social engagement. ²⁹ Historically, prosthesis options for individuals with PHPF amputation included primarily passive, custom lifelike silicone prosthesis designs.^30,31 One older study reported that less than half of individuals with PHPF amputation were able to return to their previous employment after their amputation, possibly due to limited prosthetic function at the time. ³² However, with newer externally powered prostheses,^33–35 and more durable passive ratcheting and body powered designs, individuals may have increased function and adaptability. In one such study, when using a body-powered PHPF prosthesis, participants demonstrated more variable task performance compared to using a body-powered transradial prosthesis, facilitating better reaching and grasping. This variability allowed the user to adapt their behavior across various performance tasks, especially those beneficial for completing ADLs. ³⁶ The increased performance and functionality of a more distal amputation level, combined with new prosthetic choices, may impact overall social participation. Future research to further explore the experiences of PHPF prosthesis users is needed to better understand how their prosthesis is contributing to improved participation.

When looking at pain interference, individuals’ PI T-score at baseline was on average 58.5. In this case, a single item from the PROMIS PI item bank was used to measure PI. As an example, when analyzing the percent change in response option for the PAININ9 item, among the 71 individuals with PHPF amputation, the number who reported a “very much” response decreased by 3% with similar reductions of 6% and 2% in the number of subjects reported “quite a bit” or “somewhat” respectively. By contrast, there was a 12% increase in the number of individuals who reported “not at all”. (Table 4). For this subgroup of patients, pain interference was meaningfully changed. Reduced PI could also translate to reduced reliance on pharmacological management, as this is one method to treat phantom limb pain for upper limb prosthesis users. ³⁷ Therefore, prosthesis provision may play an important role in improving overall patient quality of life by decreasing pain and increasing participation. Future research to further explore the types of pain, pain management, and other items from the PROMIS PI item bank are needed to better capture the patient experience.

The present study demonstrates the benefits of prosthesis receipt with respect to social participation and PI which builds on previous work that has identified increased physical function ³⁸ and overall well-being ²⁰ after first prosthesis receipt. These factors combined, further support provision of upper limb prostheses following ULA. A recent study on prosthesis discontinuation among individuals with traumatic ULA identified discomfort as the most common reason for stopping use. ³⁹ Additional factors, such as weight of the device and functional outcomes have also been identified as leading factors for discontinued use.^21,39–44 Future work that better elicits how patients balance the tradeoffs with upper limb prosthesis selection may be able to mitigate the influence from all these factors that lead to discontinued use and social participation.

This study only investigated a single timepoint after prosthesis receipt, so it is unknown how APSRA and PI scores will continue to evolve with time. These promising results in the first year following prosthesis receipt are encouraging for long-term improved health outcomes following ULA. Prosthetists can continue to measure these health domains during routine clinical care to ensure the device is well-fitting and ultimately helps their patient achieve their daily and long-term health goals. Longitudinal outcomes can later motivate future research to understand how pain and participation evolve over time. Specifically, understanding how social participation and pain change while transitioning from body-powered to myoelectric control, or vice versa, would be crucial for guiding future clinical decision-making.

While these findings offer important insights, several limitations should be considered. The outcomes assessed in this study were collected as part of routine clinical care. As such, some individuals may have returned to the clinic for a routine follow-up appointment without any problems, while some individuals may have returned because of a problem (e.g., broken componentry, poor socket fit due to residual limb changes, skin irritation). Therefore, the follow-up data is not guaranteed to reflect the highest possible performance after prosthesis receipt. In addition, all outcomes were collected under a single network of private clinics across the United States, which may limit generalizability to populations with different socioeconomic backgrounds or levels of healthcare access. The study cohort mostly had worker’s compensation, private, and Medicaid insurance, which reflects our cohort of mostly individuals with traumatic amputations, and future research is needed to understand prosthesis componentry access, usage patterns, and outcomes with respect to socioeconomic status. The cohort similarly limits generalizability to uninsured or underinsured individuals with upper limb amputation. This study also analyzed individuals receiving their first prosthesis within current clinical care system. As noted in Table 1, the time since amputation was close to three years with a large standard deviation. It is possible that some participants waited to receive their first prosthesis or switched providers after receiving a previous prosthesis, so this estimate may not accurately reflect the current time between amputation and prosthesis receipt. Additionally, this study primarily focused on individuals with unilateral ULA with a very small sample size for patients with bilateral amputations. This smaller sample size for this subgroup may limit the broader applicability of the results for these patients. The study was also not able to make any conclusions surrounding device type as only 50% of the sample had recorded the type of prosthesis they were prescribed (e.g., body-powered, myoelectric). As additional outcomes are collected, future research to expand our understanding of the experiences of patients across areas such as amputation levels, device types, and socioeconomic classes, is needed.

Conclusion

Results from this study indicate that individuals with ULA who received a prosthesis report improved participation in social roles and activities and improved pain interference within the first year of prosthesis receipt. Clinicians can use this knowledge to help in shared decision-making with patients when deciding on next steps in rehabilitation care and deciding whether to pursue a prosthesis. Future work can continue to expand this work to compare prosthesis control type (e.g., body-powered, myoelectric) and its influence on participation and pain interference.

ORCID iDs

Bretta L. Fylstra https://orcid.org/0000-0002-8167-0246

Dwiesha L. England https://orcid.org/0000-0002-4274-5356

Phillip M. Stevens https://orcid.org/0000-0003-3488-0707

Ethical considerations

This database review was approved by the Sterling Institutional Review Board (protocol #2786) and designated to qualify for exempt status.

Data Availability Statement

Data can be made available on reasonable request.*