Prosthetics and orthotics is a growing field, with a history dating back thousands of years. In primitive times, prostheses were made of metal, leather, and wood; the first known prosthesis dates back to 300 BC.1 Orthotic devices date back to as early as Egypt’s Fifth Dynasty (2730–2625 BC), when a device to stabilize the knee joint was created. Prosthetics and orthotics gained prominence in the United States during the Civil War, when the need for artificial limbs increased dramatically and artificial limb companies were formed.2 Current prostheses and orthoses have changed significantly since these early beginnings because of advancements in both knowledge and technology. Along with shifts in the prosthetic and orthotic (P&O) devices themselves, the field itself has also shifted, with increasing requirements for schooling and certification now required to become a professional in either field. Research and publications related to prosthetics and orthotics have evolved along with the profession itself.
Previous research has focused on the quality and the quantity of articles related to rehabilitation and assistive technology including prostheses and orthoses. A bibliometric analysis by Ryan et al. investigated trends in published articles on assistive technology, including artificial limbs and orthotic devices, during the past 40 years.3 This article investigated the journals publishing significant numbers of articles related to assistive technology and the quality of these journals. Other bibliometric analyses have focused on the quality of research and the journal impact factor of journals related to physical medicine and rehabilitation.4–6 However, there has not been a bibliometric analysis performed specifically addressing the quality and the quantity of research in the field of prosthetics and orthotics.
The Journal of Prosthetics and Orthotics (JPO) is published by the American Academy of Orthotists and Prosthetists (the Academy). The Academy was founded in November 1970 to further the scientific and educational attainments of the profession. In 1978, the Academy took over the publishing of the Newsletter of Amputee Clinics and renamed this publication Newsletter: Prosthetics and Orthotics Clinic. In 1980, the Newsletter became a professional journal titled Clinical Prosthetics & Orthotics. This new publication began attracting an increasing amount of technical and clinical articles written by orthotics and prosthetics practitioners. In 1988, the Academy’s journal, Clinical Prosthetics & Orthotics, was merged with the journal of the American Orthotic & Prosthetic Association (AOPA), Orthotics & Prosthetics Journal; this new publication was called the JPO. Although the JPO remained a quarterly publication, the size of the journal grew substantially, and more emphasis was placed on research-based articles. The JPO also began implementing a peer-review process for reviewing manuscripts.
The purpose of this bibliometric analysis was to describe trends in publication of the JPO during the past 3 years. We focused on publication patterns in the areas of article type, demographics of subjects, study details, component/intervention type, funding, and author training. The results of this analysis reflect the current JPO publishing trends.
All articles published in the JPO in 2009, 2010, and 2011 were included in the review. Editorials written by the journal editor, letters to the editor, special issues, and supplements were excluded. Two authors (R.M.M. and A.L.L.) conducted the review, and a third author (M.J.H.) would review in the event of a disagreement. When available, articles were obtained on JPO’s website; otherwise, hard copies were used.
Six main domains were chosen for our coding system: 1) article listing description, 2) article type, 3) subject demographics, 4) study details, 5) funding, and 6) author training. The domains were further broken down into 19 subcategories, then reliability testing was conducted.
For reliability testing, the authors agreed a priori that two sample articles would be analyzed per reliability assessment round, and the process would be repeated until reliability reached greater than 90%. At that point, articles would be analyzed independently. After reliability testing, two additional subcategories were identified, totaling 21 subcategories (Table 1).
Finally, Google Scholar (http://scholar.google.com) was used to count the number of times the articles in this bibliometric analysis were referenced in other articles.
Descriptive statistics such as mean, median, and interquartile ranges were used in this analysis. Interquartile range was calculated when determining the age range of subjects participating in individual studies. Microsoft Excel 2010 (Microsoft Office Excel 2010 version 14.0, Redmond, WA, USA) was used to calculate interquartile range.
PRELIMINARY RELIABILITY SCREENING
Initial preliminary reliability screening revealed 81% agreement between the two raters. Sources for threats to reliability were discussed among the author team, and the reliability assessment was repeated four additional times. On the fifth reliability assessment, at which point 10 articles had been reviewed, 96% reliability was achieved. As a result of the reliability screening process, revisions were made to category descriptors, and two additional categories were added, resulting in the 21 total previously described.
The authors independently reviewed the remaining 85 articles and entered the information into identical Microsoft Excel spreadsheets. A total of 95 articles published from 2009 through 2011 in the JPO were ultimately included in this review. A total of 31 articles were published in 2009; 35 in 2010; and 29 in 2011.
In terms of article type, 33 (34.7%) were case study/reports/series, 32 (33.7%) were prospective studies, 9 (9.5%) were technical notes, 8 (8.4%) were editorials (but not titled “Editorial” and thus were included), 7 (7.4%) were literature reviews, and 6 (6.3%) were retrospective studies.
For article type by discipline, the prosthetic breakdown included 24 case studies, 22 prospective studies, 6 technical notes, 5 editorials, 5 retrospective studies, and 3 literature reviews. Orthotics articles included 10 prospective studies, 9 case studies, 3 technical notes, 3 literature reviews, and 1 retrospective study. Three editorials and 1 literature review discussed both prosthetics and orthotics. Figure 1 summarizes the articles by research design and discipline.
DISCIPLINE AND LEVEL
Of the 95 articles published during the aforementioned period, 65 (68.4%) were related to prosthetics, 26 (27.4%) were related to orthotics, and 4 (4.2%) were related to both prosthetics and orthotics.
In 2009, a total of 23 articles (74.2%) were prosthetics related and 8 articles (25.8%) were orthotics related. In 2010, a total of 25 articles (71.4%) were prosthetics related, 8 articles (22.9%) were orthotics related, and 2 articles (5.7%) were both prosthetics and orthotics related. In 2011, a total of 17 articles (58.6%) were prosthetics related, 10 articles (34.5%) were orthotics related, and 2 articles (6.9%) were both prosthetics and orthotics related.
The prosthetics-related articles included the following amputation levels: 15 transtibial, 12 knee disarticulation/transfemoral, 11 multiple lower limb, 10 nonanatomic, 4 transradial, 3 multiple upper and lower limb, 3 wrist disarticulation/partial hand, 2 ankle disarticulation/partial foot, 2 multiple upper limb, 2 elbow disarticulation/transhumeral, and 1 hip disarticulation.
The orthotics-related articles included 11 ankle-foot orthosis (AFO); 4 shoe related; and 1 of each of the following: knee-ankle-foot orthosis (KAFO), 1 Norsk Funktion Walking Orthosis, and 1 regarding manual muscle testing to assist in guiding orthosis use. The upper-limb orthotics articles included one forearm rotation elbow orthosis, one dynamic movement orthosis, one dynamic elastomeric fabric orthosis, one compression brace, one functional electrical stimulation (FES), and one nonanatomic. One article each for cranial remolding and scoliosis bracing was also included with upper-limb orthotics. The four articles involving both orthotics and prosthetics were all classified into the nonanatomic category.
SUBJECT DEMOGRAPHICS: AGE MEAN, AGE RANGE, AND SEX
A total of 38 prosthetics studies reported age mean, which ranged from 2 weeks to 86 years, and age ranges from 2 weeks to 95 years. A total of 15 orthotics studies reported age means for subjects, which ranged from 7 months to 74 years; 14 orthotics studies reported an age range, from 7 months to 79 years. The median age for the studies reporting age mean was 40 years (interquartile range, 16.4–53.1 years). The median age of the prosthetics studies was 33 years (interquartile range, 21–47.5 years), whereas the orthotics studies reported a median age of 15.4 years (interquartile range, 10.8–40.4 years).
Of the 67 studies involving human subjects, the number of subjects ranged from 1 to 368, with a median of 5 (interquartile range, 1–16). Table 2 summarizes the sample size per study.
The articles were reviewed for patterns in subject sex. Cumulatively, 1,284 male subjects and 454 female subjects participated in the studies. In prosthetics, 1,173 males and 357 females participated in the studies, whereas in orthotics, 111 males and 97 females participated.
The results for the types of etiologies related to the prosthetics and orthotics articles are summarized in Figures 2 and 3, respectively. Trauma and diabetes/peripheral vascular disease (PVD)/vascular-related amputations composed the two largest etiologies in the prosthetics-related studies, appearing in 27% and 19% of the studies, respectively. Medically induced amputation was the lowest occurring, appearing in only 1% of the prosthetics-related articles. Etiologies in the orthotics-related articles appeared more consistently, with each etiology occurring in 3% to 10% of the articles. Most of the orthotics-related articles were classified as nonhuman subjects (28%). Relative to our coding system (Table 1), this would include testing of orthotic materials or models, and literature reviews or any article was rated as nonanatomic under type of orthosis and rated as zero under number of subjects.
The articles were also evaluated to determine whether the research undertaken was supported by an extramural funding source. If an international funding source sponsored the research in the home country, this was classified as national because the funding remained domestic to that country. Of 65 prosthetics articles, 25 (39%) reported funding support. Of these, 19 received national or international funding (13 US, 6 international), 3 reported foundation support (2 US, 1 international), 2 had US state support, and 1 had US industry support. Nine (35%) of 26 orthotics articles reported funding support. Three received national support (two US, one international); three, foundation support (two US, one international); two, US state support; and one, international industry support. One of the dual prosthetics/orthotics publications was funded by a grant from the American Academy of Orthotists and Prosthetists, with flow-down funds from the US Department of Education.
An author count was conducted to determine the number of authors per publication. In addition, the types of degrees possessed by listed authors were counted. The results of the author count and types of degrees are summarized in Figures 4 and 5, respectively. Three-author collaboration was the most prevalent in the prosthetics-related articles (23%). In orthotics, two- and three-author collaborations each occurred in 27% of the articles. Articles involving both prosthetics and orthotics most frequently had single authorship, occurring in 50% of the articles. With regard to authors’ degrees, clinical credentials were the most prevalent (36%), followed by academic doctoral degrees (28%). There were 21 authors (5%) with engineering degrees in the total degree count. Nineteen of the 21 had authorship in the prosthetics-related articles.
The most cited article was referenced 13 times. Twelve articles were referenced twice, and 24 articles in this bibliometric analysis were referenced once. The seven articles referenced five or more times were prosthetics related. The most referenced orthotics article was referenced three times. A summary of the articles that were referenced three or more times is included in Table 3.
We hypothesized that more articles appearing in the JPO during the period analyzed would focus on prosthetics. We further hypothesized that most studies would be unfunded and that a greater number of those that were funded would be prosthetics related.
Prosthetics and orthotics had equal numbers of literature reviews; however, prosthetics led orthotics in all other article types. Prospective and case studies were the leading article type in both disciplines. The large number of case studies decreased the number of subjects in the studies. Case series and studies frequently use a small number of subjects, as evident in the results, in which 37 studies (29 prosthetics, 8 orthotics) had one to five subjects. To influence policy and reimbursement decisions and to make clinical practice guidelines, study designs would likely need to include more subjects so that generalizable conclusions can be drawn.
Our review confirms that more than half of the articles (68%) in this 3-year review contained prosthetics as the main topic. Such a trend may point to greater research interest in the field, with impetus toward publishing research results. In a year-by-year analysis, however, whereas the orthotics-related articles increased from 8 in both 2009 and 2010 to 10 in 2011, the number of prosthetics-related articles increased from 2009 to 2010, then dropped in 2011. There could be many reasons for this. For instance, it can be speculated that perhaps a number of articles submitted to the JPO did not pass peer review or simply that more articles on orthotic topics were submitted. More publication years would need to be evaluated to speak to potential trends.
Subject demographics revealed that the median age of subjects in the orthotics-related studies was lower (15 years) than in prosthetics (33 years). With so many congenital and childhood pathologies that require orthotic intervention such as cerebral palsy, muscular dystrophy, plagiocephaly, hip dysplasia, and clubfoot, this is not surprising.9 Recent epidemiologic data show that the great majority of amputees is in fact older (highest level: >85 years),10 so it is not surprising that the prosthetic study subjects are older. What is surprising is that the JPO’s 2009–2011 subject mean age is not greater. However, a recent literature review of microprocessor knees shows a comparable age disparity in prosthetic research in which the mean ages of transfemoral amputees were in the fourth decade of life.11 Again, this does not accurately reflect societal demographics and emphasizes the fact that younger amputee subjects are participating in prosthetic studies at greater levels than their senior counterparts.
Etiology analysis revealed a more even distribution among the orthotics-related articles, whereas trauma was the leading etiology in the prosthetics-related articles. The orthotic articles are more representative, whereas the prosthetic articles are again not representative of societal demographics. Epidemiologic data clearly indicate that the greatest number of amputees is of dysvascular etiology,10 yet traumatic amputees represent the most prevalent etiology in published research in the JPO. Although this may not be representative of societal amputation demographics at large, this may be a highly appropriate research choice in many cases such as in the case of high-intensity athletics-related interventions. However, to affect reimbursement policy or clinical practice for most amputees, research subjects’ demographics should more closely match societal demographics.
Subject sex was also more evenly distributed in the orthotics-related articles. In the prosthetics articles reviewed, the males outnumbered the females by a greater than 3:1 ratio. Evidence dictates that more males are amputees; therefore, higher numbers of males are more likely to be recruited. For instance, Dillingham et al.10 found no notable difference in dysvascular and cancer-related etiologies by sex. However, they found that males were 4.94 times more likely to have a trauma-related amputation than women. Obvious issues with lower female participation include potential differences in use, preference for cosmetic finishing, differences in prognostic and satisfaction expectations with devices, and more.
A comparable number of prosthetics (38%) and orthotics (35%) publications during this period received extramural funding support. The total number of publications with extramural funding support comprised only 36% of the 95 articles reviewed, leaving the majority (64%) unfunded. The fact that such a high number of publications were unfunded makes it attractive to speculate that this is one potential contributor for the low subject number per article; that is, more rigorous study designs with extramural funding would likely have greater sample sizes and prospective designs. Where the present ratio of prospective studies to case studies in the JPO from 2009 to 2011 is close to 1:1, if more studies were extramurally funded, the proportion of prospective studies may likely increase. Similarly, if extramural funding levels increased, boosting the ability to increase sample sizes and the ratio of prospective to case studies, then such a body of knowledge would be better positioned to influence clinical and extramural funding decisions.
A survey of prosthetics and orthotics professionals12 revealed that the JPO is the most read peer-reviewed periodical by P&O practitioners, indicating the journal’s content utility among the clinical audience. This may be in part due to the very high contribution from P&O clinicians as evidenced by the fact that the most prevalent contributors were clinicians (Figure 5). Prosthetic and orthotic practice, however, is multidisciplinary, as evidenced by the multiple types of clinicians and nonclinicians (i.e., engineers). Prosthetic and orthotic science is also greatly supported by individuals with academic doctoral preparation, which was the second highest contribution to the prosthetic and orthotic articles.
A journal’s impact factor measures the number of times that articles published in a specific period are referenced in other articles at a later period.13 A number of JPO articles were cited in other periodicals. This suggests that a few key articles from JPO’s recent years are in fact making an impact in the science reported in other journals. Of the seven articles that were referenced five or more times, three were case studies, two were prospective studies, and two were editorials. The JPO is a highly specialized journal, which is one potential reason for its presently limited number of outside referencing. In addition, we evaluated the three most recent years, and it could take additional time to build a substantial reference count. More years of JPO publishing would need to be evaluated to determine whether continued outside referencing would occur with more time. Coronado et al.6 used HistCite (Thomson Reuters, New York, NY, USA) in their bibliometric analysis. However, HistCite does not include JPO in its database; therefore, it was not possible to use for comparison with our results. Alternatively, Google Scholar was used.
Future research should look at more years of JPO and investigate whether different trends appear in a longer period such as continued growth of referencing in other journals and the slight increase in orthotic articles observed here. The creation of Google Scholar has increased the interest in and access to the JPO. However, the fact that the JPO is not indexed in MEDLINE limits some amount of access to it and, more importantly, the ability to use tools such as HistCite and Web of Knowledge (Thomson Reuters, New York, NY, USA) to determine the scope of impact. Application to the National Library of Medicine for inclusion in the MEDLINE database is in process. In addition, it would be useful to compare the results of bibliometric analyses of the JPO with those of other highly specialized journals to determine whether publication trends are similar in other journals with common topical interests. Such a comparison could be useful in soliciting specific content to accurately reflect societal demographics.
In conclusion, this 3-year bibliometric analysis of the JPO reveals that a greater number of prosthetic articles than orthotic articles are published. The number of case studies to prospective research studies is approximately equal. By age, subjects are not highly representative of the demographics of device users according to epidemiologic data. A similar issue is present regarding amputation etiology, in which more subjects with traumatic amputation participate in research published in the JPO. Regarding financial research support, it appears that most P&O research published in the JPO is not extramurally funded.