PDF
Citation
Print
INTRODUCTION
Upper limb amputation, defined as the loss of the hand and arm responsible for strength, sensation, and fine motor function, profoundly affects an individual’s quality of life, independence, and social participation [1]. Compared with lower limb amputation, the loss of an upper limb presents greater challenges for functional recovery and reintegration, given the role of the upper limb in activities of daily living and the complexity of prosthetic and rehabilitative interventions [2].
Epidemiological data on the prevalence and incidence of upper limb amputation are highly heterogeneous across countries, and many studies have failed to distinguish between upper and lower limb amputations, limiting the accuracy of global comparisons and the development of targeted policies. In a national estimate using the US Nationwide Inpatient Sample and census data, approximately 1.6 million individuals were living with limb loss in 2005, of whom 541,000 had upper limb loss and approximately 41,000 had major upper limb amputation, excluding finger amputations [3]. In South Korea, the 2020 National Survey of Disabled Persons reported approximately 125,457 individuals with upper limb amputation, with the majority having amputations below the wrist. Notably, the number of individuals with amputations above the elbow more than doubled compared with 2011 [4].
The anatomical level of upper limb amputation is a key determinant of functional prognosis. Most studies have classified upper limb amputation into major (above or below the elbow) and minor (below the wrist). Although the most common levels vary by country, below-elbow amputation, excluding finger loss, has been consistently reported as the predominant level in both developed and developing countries [4-13].
The causes of upper limb amputation are diverse and can be broadly categorized into traumatic, congenital, and medical etiologies. Traumatic amputations are primarily caused by industrial accidents, traffic collisions, and combat injuries. These traumatic events are particularly prevalent among workers because of their occupational environment [13]. Congenital limb deficiency resulting from malformations during embryonic development accounts for a substantial proportion of upper limb impairment in pediatric populations [14]. However, because the scope of this study was limited to acquired upper limb amputation, congenital limb deficiency was not included in this review. Additionally, nontraumatic amputations caused by medical conditions, such as diabetes, vascular disease, infection, and malignancy, have been reported, especially among individuals with disabilities [4].
To date, much of the literature on upper limb amputation has been based on local data, and few studies have clearly distinguished upper limb from lower limb amputation in epidemiological analyses. Consequently, a comprehensive understanding of the global patterns of upper limb amputation, including prevalence, incidence, amputation level, and etiology, remains limited. This systematic review aimed to address this gap by synthesizing population-level data from multiple countries, thereby providing a foundation for resource allocation and policy development in rehabilitation and disability services.
METHODS AND REVIEW PROCESS
This review followed the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines.
Search strategy
A highly sensitive search strategy was developed as shown in Fig. 1 and Appendices 1-4. The final search terms and subject headings (expanded where appropriate) were systematically applied to four databases: PubMed, Embase, the Cochrane Library, and Regional Information Sharing Systems. The searches were conducted in March 2023 and repeated in May 2023. The search was restricted to the articles published in English. All retrieved records were imported into EndNote™ (Clarivate) for collation and the removal of duplicates.
Screening, data extraction, and risk of bias assessments were independently performed by two reviewers who were blinded to each other’s evaluations at all stages. Discrepancies were resolved through discussions with a third reviewer. Within each keyword group, search terms were combined using the Boolean operator “OR,” and the keyword groups themselves were linked using the Boolean operator “AND.” In PubMed, Medical Subject Headings were used whenever applicable, and truncation was applied using the asterisk (*) symbol.
Eligibility criteria
The following inclusion criteria were applied to identify studies relevant to the objectives of this review:
1. Studies focusing on acquired major upper limb amputations were included. Studies focusing on congenital limb deficiency were excluded in this review. Studies that did not differentiate between upper and lower limb amputations or between major and minor upper limb amputations were excluded. Studies that exclusively reported minor upper limb amputations were also excluded.
2. Observational studies were included; randomized controlled trials were excluded.
3. Review articles were eligible; protocols and clinical guidelines were excluded.
4. Studies with sample size of at least 100 participants were included. Case reports, case vignettes, and qualitative studies, such as focus group interviews, were excluded.
5. Articles written in the English language were included.
6. For studies addressing amputation levels, inclusion of major upper limb amputations was required.
7. For studies addressing descriptive epidemiology, reporting of incidence or prevalence rates was required.
8. For studies addressing the etiology, a description of the underlying causes of amputation was required.
Screening and data extraction
All records identified through the database searches were imported into EndNote 20 reference management software, and duplicates were systematically removed. The screening was conducted in two stages. First, two reviewers (Choi Y and Kim E) independently screened the titles and abstracts of all retrieved studies to assess their eligibility based on the predefined inclusion criteria. Studies that met the initial criteria were subjected to full-text review conducted independently by the same two reviewers. The reasons for exclusion were documented at the full-text stage for transparency. Any discrepancies or disagreements during either stage of the screening process were resolved through discussion with a third reviewer (Oh BM).
Data were independently extracted by two reviewers using a standardized form. The extracted data included general information (first author, year of publication, continent, and country), study characteristics (year[s] of data collection and total sample size), and detailed outcome measures. These outcome measures included the number of upper limb amputation cases, laterality (unilateral or bilateral, left or right), level of amputation, and etiology of amputation. In addition, epidemiological metrics, such as prevalence or incidence rates, were collected when reported. In cases where the incidence rates were not explicitly reported in the original articles, the incidence was calculated by the author (Choi Y) using the number of upper limb amputation cases as the numerator and the reported population size as the denominator. All extracted data were cross-checked for consistency, and any disagreements were resolved through consultation with the third reviewer.
Quality assessment
The risk of bias was assessed using a slightly modified version of the tool developed by Hoy et al. [15], which comprised ten items: four addressing external validity and six addressing internal validity. Each item was scored dichotomously (0 or 1), yielding a maximum total score of 10. Studies with scores ≥8 were considered to have a low risk of bias, scores of 6–7 as moderate risk, and scores ≤5 as high risk. Quality appraisal was performed only for descriptive purposes and was not used as a criterion for study exclusion.
Synthesis
Given the anticipated heterogeneity in the study populations, settings, and methods of reporting the incidence and prevalence of upper limb amputation, quantitative synthesis was considered inappropriate. Therefore, narrative synthesis was performed. The findings were summarized descriptively, with attention to variations in epidemiological estimates, amputation levels, causes, and study designs. The patterns and differences across studies were explored to provide a comprehensive overview of the available evidence.
RESULTS
Quality assessment
Of the 19 studies included in the review, 5 studies were assessed as having a low risk of bias, 7 studies a moderate risk, and 7 studies a high risk (Table 1). Common sources of bias identified across studies included the use of a consistent mode of data collection for all participants (n=19), the use of a sampling frame that closely represented the target population (n=18), the application of an acceptable case definition (n=16), and a low likelihood of nonresponse bias (n=15).
Study characteristics
The characteristics of the included studies are summarized in Table 2. When classified according to continent, 7 studies were conducted in North America [3,5-8,16,17], 3 in Europe [9,10,18], 9 in Asia [4,11-13,19-23], According to country, the largest number of studies were from the United States (n=7), followed by Iran (n=3), South Korea, and Taiwan (n=2 each), and one study each from Sweden, Norway, France, Saudi Arabia, and India.
The oldest included study was conducted by al-Turaiki and al-Falahi [11], published in 1993 and used data collected between 1977 and 1990. The most recent studies were published in 2022 by Dhillon et al. [21], Bok and Song [4], and Huang et al. [23] using data collected between 2018 and 2019, 2011 and 2020, and 1996 and 2013.
The study with the largest sample size was that of Bok and Song [4], which included 2,622,950 individuals as of 2020. In contrast, the smallest study was by Allami et al. [20], with 103 participants. The most common data sources were institution-based datasets (n=10), such as those from hospitals, rehabilitation centers, trauma centers, and prosthetic companies.
Characteristics of upper limb amputations
The characteristics of the upper limb amputations reported in the included studies are summarized in Table 2.
North America
In a study conducted in the United States and published by Dillingham et al. in 1998 [16], 6,069 trauma-related limb amputation cases were identified in the Maryland Health Services Cost Review Commission database between 1979 and 1993. Among these, 211 cases were classified as major upper limb amputations and 3,897 minor amputations. The level of major amputation included 95 cases above the elbow and 93 cases below the elbow.
McCall and Horwitz (2006) [17] analyzed Workers’ Compensation claims data from 1994 to 2003 and identified 2,297 amputation claims, of which 2,295 involved the upper extremity. Among individuals with upper limb amputation, most underwent minor amputations. Major amputations were less common and included lower arm (n=7), upper arm (n=4), wrist (n=4), and combined wrist and hand amputations (n=1).
According to Ziegler-Graham et al. (2008) [3], approximately 1.6 million individuals were living with limb loss in the United States in 2005, based on 1988–1999 data from the Nationwide Inpatient Sample of the Healthcare Cost and Utilization Project and annual census estimates. In 2005, approximately 41,000 were living with a major upper limb amputation and 500,000 with a minor upper limb amputation.
In a study by Barmparas et al. (2010) [5], 8,910 patients with limb amputation were identified in the National Trauma Data Bank from 2000 to 2004, of whom 782 had upper extremity amputations. Among the 302 individuals with major upper limb amputation, 183 had amputations below the elbow, and 119 had amputations at or above the elbow.
Krueger et al. (2012) [6] analyzed data from the Military Amputation Database and identified 1,221 patients with 1,634 amputation events between 2001 and 2011. Among the 228 major upper limb amputations, the most common type was transradial (n=110), followed by transhumeral (n=81).
Tennent et al. (2014) [7] reviewed records from the US military trauma registry and identified 1,315 service members with limb amputations, among whom 173 had upper limb amputations. The most common major amputation level was transradial (n=82), followed by transhumeral (n=59).
In a pediatric population study, Vakhshori et al. (2019) [8] identified 6,130 children with upper limb amputations using the Kids’ Inpatient Database, a national all-payer database of the Healthcare Cost and Utilization Project. Excluding finger amputations, 298 cases involved below-elbow amputations and 142 involved above-elbow amputations.
Europe
In a study from Sweden published by Asplund et al. in 2009 [9], 4,202 traumatic amputation cases were identified in the Swedish Hospital Discharge Register between 1998 and 2006. Of these, 825 involved upper limb amputations. When excluding amputations of the hand or fingers, the number of major upper limb amputations included 34 cases between the shoulder and elbow and 72 cases between the elbow and wrist.
Østlie K et al. (2011) [10] reported on 416 individuals with upper limb amputations registered in the databases of two upper limb prosthetics companies and three of the largest hospitals in Norway from October 2006 to May 2008. Among them, 211 had amputations in the right arm, 181 in the left arm, and 21 in bilateral amputees. The most common major amputation level was the forearm (n=176), followed by the upper arm (n=99), through-wrist (n=73), elbow joint (n=24), and shoulder joint (n=23).
In a study conducted in France and published by Pomares et al. in 2018 [18], 2,247 upper limb amputation cases were identified among 55,238 patients treated for traumatic limb loss at the Nancy University Hospital. Of these, 1,715 were traumatic amputations and 532 were nontraumatic. Right arm amputations accounted for 521 cases and left arm amputations for 611 cases. A total of 1,132 patients underwent surgical amputation, with finger amputation being the most frequent. Among the major amputations excluding the fingers, four involved the forearm and another four involved the upper arm.
West Asia
In a study published in Saudi Arabia by al-Turaiki and al-Falahi in 1993 [11], 610 upper limb amputation cases were identified among 3,210 limb amputees who visited the Riyadh Medical Rehabilitation Centre between 1977 and 1990. Of them, 17 were bilateral amputees. The most common levels of major amputations were transradial (n=243) and transhumeral (n=152).
Mousavi et al. (2012) [19] analyzed data from the archives of educational hospitals in Kerman, Iran, and identified 624 limb amputation cases between 2003 and 2011. Among these, 144 involved upper limb amputations, including 78 major (above the wrist) and 66 minor (below the wrist) cases. The most frequently reported sites of major amputation were the arm (n=39), forearm (n=20), and elbow (n=10).
In a study conducted by Ghavami et al. (2014) [12] at Shahid Motahari Burn Hospital, 682 patients with electrical burn injuries were reviewed. Excluding finger and thumb injuries, the most common level of injury was the forearm, which has been reported in approximately 50–60 cases.
Allami et al. (2016) [20] analyzed data from the Veterans and Martyr Affairs Foundation and identified 103 Iranian veterans of the Iran–Iraq War with bilateral upper limb amputations, all of whom were included in the study. The most frequently affected levels were the fingers or wrist (n=108), elbow (n=82), and transhumeral or more proximal levels (n=16).
South Asia
In a study conducted in India and published by Dhillon et al. in 2022 [21], 17,445 trauma cases that presented to the trauma center at the Postgraduate Institute of Medical Education and Research were reviewed. Among these, 104 involved upper limb amputation. The most common levels of major amputation were above (n=44) and below the elbow (n=40), followed by shoulder disarticulation (n=13), elbow disarticulation (n=4), and wrist disarticulation (n=3).
East Asia
In a study conducted in Taiwan and published by Liang et al. in 2004 [13], 2,970 upper limb amputation claims were identified in the workers’ compensation database from 1999 to 2001. After excluding 10 patients involved in traffic accidents and another 10 with missing sex or age data, 2,950 cases were analyzed. Among them, 2,925 were unilateral and 25 were bilateral. Right-sided amputations were slightly more common (n=1,495) than left-sided amputations (n=1,480). Among the unilateral cases, the number of major amputations, excluding finger amputations, was greater for the below-elbow (total, n=78; male, n=59; female, n=19) than for the above-elbow (total, n=31; male, n=22; female, n=9).
Ro et al. (2019) [22] identified 49,535 cases of upper limb amputation in South Korea between 2004 and 2013 using a workers’ compensation database. The number of cases per year, including finger amputations, ranged from 4,316 in 2004 to 4,441 in 2013, with annual figures peaking at 5,565 in 2007.
Bok and Song (2022) [4] analyzed data from the National Survey of Disabled Persons in South Korea from 2011 to 2020. The numbers of upper limb amputees were 133,662 in 2011, 149,249 in 2014, 140,867 in 2017, and 125,457 in 2020. Across all survey years among upper and lower limb amputees, amputations below the wrist accounted for the majority (50%–70%), followed by the elbow and wrist (~7%), and above the elbow (3%–10%). Notably, the proportion of above-elbow amputations increased from 3.5% in 2011 (n=5,771) to 9.7% in 2020 (n=16,924).
In Taiwan, Huang et al. (2022) [23] used the Longitudinal Health Insurance Database, a nationally representative cohort randomly sampled from the National Health Research Database, and identified 1,188 patients with upper limb amputation. Among them, 104 underwent major amputations and 1,084 underwent minor amputations. According to age group, 966 patients were aged ≤64 years, 76 were aged 65–74, 37 were aged 75–84, and 5 were aged ≥85 years.
Incidence and prevalence of upper limb amputation
Table 3 presents the incidence and prevalence rates of upper limb amputation performed in various population groups in different regions. The reported incidence and prevalence widely varied depending on the study population, data source, and definition of major versus minor amputations.
In North America, Dillingham et al. (1998) [16] reported a gradual decline in the incidence of major upper limb amputations in Maryland from 1979 to 1993, with above-elbow amputations decreasing from 0.10 to 0.04 per 100,000 and below-elbow amputations from 0.22 to 0.08 per 100,000. Ziegler-Graham et al. (2008) [3] estimated that approximately 41,000 individuals in the United States were living with a major upper limb amputation in 2005. Trauma was the predominant etiology (n=34,000), followed by dysvascular diseases (n=5,000), and cancer (n=2,000).
In Europe, Østlie et al. (2011) [10] reported a prevalence of 11.6 per 100,000 in adults aged ≥18 years based on prosthetic service data and hospital data in Norway.
In East Asia, Liang et al. (2004) [13] documented an overall incidence rate of 12.5 per 100,000 among Taiwanese workers, with higher rates in males (18.9 per 100,000) than females (5.1 per 100,000). Ro et al. (2019) [22] reported annual crude incidence rates ranging from 17.6 to 23.7 per 100,000 workers in Korea between 2004 and 2013. In the current review, these rates were calculated based on the number of amputations and workforce data reported in the original study. Bok and Song (2022) [4] analyzed nationally representative survey data and reported an increased prevalence of major upper limb amputations among individuals with disabilities from 6.9 per 1,000 in 2011 to 11.3 per 1,000 in 2020. Notably, the prevalence of above-elbow amputations increased from 2.2 in 2011 to 6.5 per 1,000 in 2020, whereas the prevalence of below-wrist amputations declined over the same period. In Taiwan, Huang et al. (2022) [23] observed a significant downward trend in the age- and sex-adjusted incidence of major upper limb amputation in older adults aged ≥65 years, from 0.54 per 100,000 in 1996 to 0.40 per 100,000 in 2013 (average annual percent change=-5.6, p-for-trend=0.005).
Etiologies of upper limb amputation
Table 4 summarizes the reported etiologies of upper limb amputation across studies from the eight countries. Trauma emerged as the predominant cause in all regions, with notable differences in the types of traumatic events based on geography and study population.
In the United States, Vakhshori et al. (2019) [8] analyzed 6,130 pediatric cases and found that traumatic upper limb amputations were predominantly caused by machinery-related accidents, including lawn mowers (n=729), caught between objects (n=640), and power tools (n=557), in the national inpatient database. Motor vehicular accidents (n=294) and firearm- or explosive-related injuries (n=142) were also notable contributors. In Norway, Østlie et al. (2011) [10] reported occupational accidents (n=91) as leading causes of upper limb amputations, followed by traffic incidents (n=31), other accidents (n=28), and explosion (n=19) among adults aged ≥18 years, based on data from two prostheses companies and three hospitals. Similarly, in France, Pomares et al. (2018) [18] identified saw injuries (n=371), crush injuries (n=232), and machinery-related trauma, such as lawn mowers (n=120), as common causes among adults aged ≥16 years in a hospital database.
In West Asia, al-Turaiki and al-Falahi (1993) [11] reported that 530 of 610 upper limb amputation cases in Saudi Arabia were trauma-related in a database from a rehabilitation center. Similarly, Mousavi et al. (2012) [19] found that trauma was the most frequent etiology in a hospital database. Of the 144 patients with upper limb amputations, approximately 66.7% (96 patients) underwent amputation due to trauma.
Liang et al. (2004) [13] analyzed 2,950 upper limb amputation cases in East Asia using Taiwan’s workers’ compensation database, and all cases were attributed to traumatic causes. Among these, 2,492 (84.5%) cases resulted from injuries involving powered machinery, such as lathes, presses, grinders, centrifuges, drills, roller mills, shears, and slicers.
Bok and Song (2022) [4], using data from the National Survey of Disabled Persons in South Korea, reported that trauma was the most common cause of upper limb amputation. However, they identified a substantial number of cases attributable to congenital anomalies and disease-related conditions. In 2020, among individuals with upper limb amputations, 113,515 were due to trauma, 909 were related to congenital limb deficiencies, and 7,816 were associated with nontraumatic medical conditions. These findings reflect both the predominance of trauma and the growing burden of chronic diseases in the etiology of upper limb amputations among the Korean population.
DISCUSSION
The key findings of this study are presented in Fig. 2. This study is the first systematic review to comprehensively examine the global epidemiology of upper limb amputation, including its prevalence, incidence, amputation levels, and underlying causes. A total of 19 studies were analyzed, revealing considerable heterogeneity in epidemiological characteristics depending on the country and data sources.
The prevalence of upper limb amputation significantly varied according to the characteristics of the study population. In 2011, the prevalence of major upper limb amputations in adults aged ≥18 years in Norway was 11.6 per 100,000 individuals [10]. In contrast, a substantially higher prevalence was observed in populations with disabilities. According to the Korean National Survey of Persons with Disabilities in 2011, the prevalence of major upper limb amputation was 693.7 per 100,000 individuals (6.9 per 1,000 individuals) with disabilities [4].
The incidence also differed according to the population characteristics. In 1993, the incidence rates of above- and below-elbow amputation in the general US population were 0.04 and 0.08 per 100,000, respectively [16]. Older adults showed a higher incidence than the general population; in Taiwan, the incidence rate in individuals aged ≥65 years was 0.54 per 100,000 in 1996 [23]. The incidence was markedly higher, with 12.5 per 100,000 workers in Taiwan (1999–2001) [13] and 19.1 per 100,000 workers in Korea (2004) based on worker compensation data [22].
The distribution of amputation levels also varied across countries. Below-elbow amputation was the most commonly reported major upper limb amputation level, with forearm amputations accounting for more than half of the cases in some studies [5,8,9,11,13,14]. As the amputation level directly influences functional recovery and prosthetic use, these findings highlight the need to consider level-specific distributions in national rehabilitation resource allocation and prosthetic device policies. In particular, above-elbow amputations present greater technical challenges and less favorable functional outcomes. However, data on this level remain scarce, underscoring a critical gap in research and development.
The causes of upper limb amputations or deficits were broadly categorized as traumatic, congenital, or medical. Trauma, including industrial accidents and machinery-related injuries, was the predominant cause of injury in most studies, particularly in those involving workers [13] and children [8]. However, a non-negligible proportion of cases were found to be caused by congenital anomalies or medical conditions, such as diabetes, vascular disease, and cancer. These patterns are particularly evident in studies that focus on individuals with disabilities [4] and children [14]. These findings indicate the need to incorporate nontraumatic etiologies into prevention and care strategies.
Epidemiological differences in upper limb amputation are likely associated with variations in healthcare systems and occupational safety standards based on the economic development status. Although countries with high rates of occupational injuries, especially low-income countries, are presumed to have a higher burden of upper limb amputation, the current literature includes significantly few studies from these regions. This gap limits the ability to design appropriate prevention policies and rehabilitation planning and calls for more active studies on upper limb amputation epidemiology in under-resourced settings.
Compared with lower limb amputation, studies on upper limb amputation and prosthetic development remain limited. Although lower limb prostheses have advanced through the standardization of gait mechanics, the upper limbs require intricate, high-precision motor control, making device development more difficult and commercialization more limited. In this context, quantifying the scale and characteristics of upper limb amputation can provide critical evidence for forecasting demand and guiding medical device innovation.
This study has several limitations. Although it aimed to provide a global overview of the epidemiology of limb amputation, data from Oceania, Africa, and South America were not included due to the absence of eligible publications meeting the inclusion criteria. This omission likely led to an underestimation of amputation cases in low- and middle-income countries, particularly those with high rates of trauma and occupational injuries, thereby limiting the global representativeness of our findings. In addition, definitions of amputation level and etiology were often inconsistent or unclear across studies, and some reports did not distinguish between upper and lower limb amputations or between major and minor amputations. The adoption of standardized reporting frameworks is therefore essential for future research. Moreover, most included studies did not report key clinical outcomes such as functional status, prosthesis use, or rehabilitation participation, hindering interpretation of the real-world impact of upper limb amputation. Finally, many studies were based on single-center or regional data, which limits the generalizability of the results.
Despite these limitations, this review provides valuable quantitative and qualitative insights into national patterns of upper limb amputation. These findings offer a foundation for the future planning of rehabilitation services and prosthetic distribution. As populations continue to age and chronic diseases become more prevalent, the relative importance of nontraumatic upper limb amputation is expected to increase. Addressing this shift requires increased attention and systematic data collection efforts.
CONCLUSION
This review synthesizes the current global evidence on the epidemiology of upper limb amputation, revealing marked heterogeneity in its prevalence, incidence, anatomical levels, and etiologies across populations and regions. The predominance of trauma-related cases, particularly among workers and children, highlights the need for improved injury prevention and workplace safety measures. Despite its impact, epidemiological data and research on upper limb amputation are limited, particularly in low-income settings. Expanding high-quality data collection and targeted rehabilitation strategies are essential to improve clinical care and prosthetic development.
XML Download