With the increased emphasis on demonstrating the effectiveness of O&P care, establishing evidence is a vital step toward providing third-party payers with clinically based justification for reimbursement decisions. Research results, however, are not only for payers; they are also used by O&P providers to deliver evidence-based care—the best solution for patients’ needs, chosen from a changing multitude of options. Research is also being conducted to reach a more holistic understanding of the patients themselves, helping to increase the value of patient-centered care.
Research also moves technology forward. Once O&P devices meet their initial goals, such as allowing patients to walk or to grasp, continued research on those devices helps users walk more safely, grasp more gently, and so much more. Research continues to build upon prior research, either indirectly—as in work being done by Kenton Kaufman, PhD, PE, that uses broad medical data for O&P purposes, or research into residual limb health by Stefania Fatone, PhD, BPO (Hons), that relies on innovations in soft electronics—or directly, such as a study conducted by W. Lee Childers, PhD, MSPO, CP, to establish the effectiveness of a popular prosthetic feature. In acknowledgement of future iterations of a study or method, researchers often end published articles describing any limitations of their processes and potential next steps to further the conclusions.
Practitioners can conduct small-scale investigations based on their own clinical experiences, but in most cases, their results can’t be tested sufficiently or shared widely enough to stimulate change. Those shortcomings can be overcome by the benefits that academic collaborations and funding sources deliver as research partners. Expenses for the laboratory, the equipment, and the personnel who are trained in engineering, statistics, data analysis, computer programming, and biomedicine require resources that are best delivered at the university or association level, including through grant funding by vested parties.
(Editor’s note: To read more about research being conducted into the economics of O&P interventions, including the effects of employment and assistance programs, that is being supported by the American Academy of Orthotists and Prosthetists (the Academy) and the American Orthotic & Prosthetic Association (AOPA), read “Economic Impact of O&P Interventions: Research and O&P Organizations Lead the Way” in the February issue of The O&P EDGE.)
For this article, The O&P EDGE contacted a variety of researchers and institutions across the United States to gather information on current studies being conducted at top universities, governmental agencies, and research facilities by authorities in the field. Through their work, the researchers answer questions about why, how, and what if, for the benefit of the O&P profession and its patients.
While each segment in this article focuses on just one researcher involved with the project, it should be noted that they are all supported by collaborators, colleagues, co-investigators, students, and others working toward the same goal. Each selection addresses questions relevant to practitioners: What was the goal of this research, how was it conducted, what are the results, and where will this lead?
W. Lee Childers, PhD, MSPO, CP,
Georgia Institute of Technology, Atlanta
“The Effect of Multiaxial Prosthetic Stiffness on Angular Momentum in People With Transtibial Amputation Walking Over Uneven Terrain,” W. Lee Childers, Ryan D. Funderburk, Adam T. Smith, C. Jake Davidson
Publication of the research results is in process.
Negotiating uneven ground can be challenging for people who use lower-limb prostheses to walk, so researchers spend time searching for solutions that will allow greater stability in these situations. Manufacturers of prosthetic feet have contributed to a solution by adding multiaxial features that better reproduce the behavior of human ankles, which can stiffen as the terrain warrants. However, Childers found that there was a lack of evidence evaluating the prosthetic ankle stiffness as it relates to the user’s dynamic balance and gait over uneven terrain. Thus, his continuing research focuses on defining the effect of multiaxial stiffness on gait stability among people with unilateral transtibial amputations.
Eleven patients with unilateral transtibial amputations completed the study. To gather data, the research team asked the participants to walk a 7.3m x 0.8m destabilizing platform at a speed set by a metronome of 108 beats per minute. They each repeated the terrain test ten times. To create the varying terrain, small blocks were fixed to the platform; a rubber mat covered the blocks to prevent the participants from getting specific visual feedback about the path. The researchers used a harness system to maintain the participants’ safety while the tests were conducted. A full-body marker and an eight-camera motion capture system recorded limb kinematics.
Each participant was evaluated while using his or her regular prosthesis and an Endolite Multiflex foot with four ankle unit stiffnesses: soft, medium, firm, and locked out. The Multiflex foot was chosen for this study because it consists of a ball-and-socket joint and stiffness can be controlled by interchanging rubber isolators between the joint’s shank and foot sections. It also complies with billing code L-5986 (Multi-axial rotation unit), for which Childers’ research will provide evidence of the value and usefulness of multiaxial features in gait stability.
“The main focus of this work was to justify that it is a good thing for prosthetic feet to have multiaxial function,” Childers says, because if it can prevent falls among its users, its value is demonstrated to the payers.
The four stiffness conditions and the habituated prostheses were tested in random order. Calculations determined each participant’s center of mass and angular momentum, and minimum toe clearance of the affected limb during swing was used as an indirect measure of gait stability. An increase in angular momentum indicates that a person is having more difficulty remaining upright and therefore stability is decreased. Because people tend to lift their feet higher when they are unsure of their environment, toe clearance is a useful stability indicator, Childers says. The locked ankle significantly increased angular momentum when compared with the compliant ankles. Toe clearance was significantly smaller when comparing the normal, soft, and medium ankle stiffnesses to the locked condition.
The research team concludes that the multiaxial features increase stability on uneven surfaces, and while each person had an ankle stiffness that maximized his or her gait stability, the stiffness did not correlate with the manufacturer recommendations for ankle stiffness selection. Childers and colleagues note that further research should repeat the tests with more participants before definitive recommendations can be made to practitioners about ankle stiffness.
The next step toward providing a clinical application for practitioners to use when fitting devices, which Childers has already begun, involves comparing the levels of ankle stiffness with the results of the stability tests to establish the ideal ankle stiffness for each of the participants. Childers will then compare those results to the manufacturer’s recommendation to further substantiate an ideal method of correlating ankle stiffness and dynamic balance and gait.
“What I hope to find is that there is a stiffness value divided by [the user’s] weight that is constant for everybody. When that is the case, that’s something manufacturers can use, too,” he says. “Now that we know that multiaxial units are good, then how do we design the best one? How do we figure out what is optimal? That’s what we’re working on right now.”
For more information about Childers’
Stefania Fatone, PhD, BPO (Hons), Northwestern University, Chicago
“Ability of Epidermal Sensors to Measure Lower Limb Temperature During Activity With a Prosthesis Stimulator,” Stefania Fatone, John Rogers, Todd Coleman, Yonggang Huang
Results of the ongoing research will be presented at the 2017 Academy Annual Meeting.
Much of Fatone’s research focuses on maintaining the health of prosthetic users’ residual limbs, which are often damaged by the effects of the socket on skin and soft tissue. She and her colleagues are evaluating a wireless, noninvasive sensor that adheres to the skin and collects shareable data about the temperature and pressure at the socket-limb interface. The epidermal sensors are easy to use, inexpensive, and stretchable. The work, in collaboration between three universities, requires developing the temperature, pressure, and strain sensors, and hydration and blood flow sensors that operate inside a prosthetic socket. About to begin year three of a four-year grant, the team has finalized a prototype that is reliable and user-friendly. Computation modeling is also being conducted to facilitate processing the sensor data to make it useful in a clinical setting.
Courtesy of Shutterstock Inc..
As an initial step, the research team led two focus groups, one with seven certified prosthetists and one with seven lower-limb prosthesis users, to gather information about problems they had encountered with prosthetic sockets, how a residual limb monitoring system could be used in clinical practice, and how the system could be configured. The prosthetists came from a variety of practice settings and had four to 33 years of clinical experience. The users had diverse levels of amputations and etiologies, and their experience using lowerlimb prostheses ranged from less than one year to 40 years. Perhaps unsurprisingly, skin breakdown, sweating, and volume fluctuation were problems listed by both groups, and in-socket temperature was a measurement priority.
The second stage of the project involved using a prosthesis simulator, which was fabricated by a certified prosthetist and fitted to the left leg of an able-bodied participant. Data was collected simultaneously from the participant’s legs by eight thermocouples and eight epidermal temperature sensors placed on top of the thermocouples on four sites on each leg (the tibial tubercle, fibular head, distal tibia, and medial gastrocnemius). The team used the data from the thermocouples, previously used to collect in-socket data by other researchers, to validate the accuracy of the temperature readings from the epidermal sensors. The temperature reading from both sensor types corresponded reasonably well, says Fatone. “We also anticipate that our epidermal sensors will be more durable than thermocouples for longer term use within a prosthesis,” she adds. “Our experience with using thermocouples is that they break more regularly than the epidermal sensors.”
Temperatures were collected during five minutes and 25 minutes of seated rest with the limb bare; five and ten minutes of seated rest with the simulator donned; 30 minutes of treadmill walking at a constant speed with and without the simulator; 30 minutes of seated rest with the simulator donned; and while donning and doffing the simulator.
The temperatures of the contralateral limb were generally constant throughout testing while temperature within the simulator increased 5-7 degrees Celsius once it was donned. Donning the gel liner caused a temperature drop, while doffing the simulator caused a nearly instant temperature drop. The epidermal sensors were durable to socket conditions and did not cause any adverse skin problems.
“Our next steps are to collect temperature data from amputee subjects wearing different socket conditions to see how temperature responds, and build algorithms into the software to interpret that data in a clinically meaningful way,” Fatone says. “We are also developing a pressure sensor using the same epidermal electronics and hope to begin testing of that sensor soon.”
For more information about Fatone’s research projects, visit www.scholars.northwestern.edu/en/persons/stefania-fatone
Mark Geil, PhD, Georgia State University, Atlanta
“Expectation and Confirmation Bias in Orthosis Use and Perception,” Mark D. Geil, Brittany Balsamo
The project is currently undergoing data analysis.
It can be difficult for practitioners to differentiate between the functionality of an O&P device and the user’s comfort level with it. If the user doesn’t like his or her choice of device, for whatever reason, it may be relegated to disuse. Understanding and beginning to quantify elements of the user’s perception is the focus of Geil’s research. The study seeks to answer the question, “Do people prefer new, computerized O&P components just because they are ‘new’ and ‘computerized?’”
The study began as a master’s thesis in biomechanics by one of Geil’s graduate students, Brittany Balsamo, MS. It includes two traits of basic human psychology, expectation bias—a person’s feeling or belief about how successful something will be—and confirmation bias—a person’s tendency to seek out or create new evidence in ways that validate his or her expectations—to help establish how personal predisposition may affect patients’ feelings about their prosthetic limbs.
In O&P, this translates to the possibility that a patient will seek out ways to validate his or her expectations that one device will perform better than another because it’s newer or more technologically advanced. “I’m curious about being able to tease out the change in perception and performance due to the increased function of the device itself and the change that’s due to psychology—due to the user’s own expectations that the device will perform better,” Geil says.
Tinxi / Shutterstock Inc..
The challenge for practitioners is that these human biases can make it difficult to receive accurate feedback from their patients.
The tests involved two identical off-the-shelf knee orthoses. The research team added an LED, a switch, and a mini-USB port on one of the orthoses and led the 18 participants to believe it was a new, computerized prototype that could dynamically change the joint’s stiffness, before testing their gait performance as they wore each brace. The healthy, young adult participants completed initial questionnaires about their expectations, including which they thought would perform better; after gait testing, they were surveyed about which brace they thought performed better, and which brace they preferred.
While the research team is still completing the final data analysis, they have found that 61 percent of the participants expressed a preference for the “computerized” brace before they saw it or used it, based on the description they were given and a mock manufacturer’s flyer. Following the walking trials, the number of participants who preferred the “computerized” brace increased to 83 percent.
The users’ perception that one brace worked better did not change the way they walked during the gait analysis. There was no difference in temporo-spatial parameters like speed and step length, in kinematic measures such as knee flexion angle, or in kinetic measures like ground reaction force.
As with many studies, there are limitations that must be considered when interpreting the results for clinical practice. In particular, the cohort was able-bodied and likely walked with an individually optimized gait pattern. “We’ve considered how the research might work in a population recovering from injury, for example,” says Geil. “If they actually need a brace to improve their gait, we might then see differences in their gait patterns if they expect one brace to perform better.”
Geil says that a future project may test for these types of biases with a conventional prosthetic component and a mock-up of a microprocessor=controlled component. He and his team are preparing grant proposals to fund the next steps in the research.
For more information about Geil’s research projects, visit education.gsu.edu/profile/mark-geil.
He (Helen) Huang, PhD, North Carolina State University, Raleigh/University of North Carolina at Chapel Hill Rehabilitation Engineering Center
“A New Powered Lower Limb Prosthesis Control Framework Based on Adaptive Dynamic Programming,” Yue Wen, Jennie Si, Xiang Gao, Stephanie Huang, He (Helen) Huang
Huang’s research addresses solutions that give people more control of their prosthetic devices, whether it’s providing neural control of lower-limb prostheses for adapting to various terrains, methods to improve myoelectric control of a prosthetic arm, or, as in this case, creating automatic tuning of a prosthetic leg. In this research topic, Huang and her colleagues set out to provide prosthesis users with more efficient and personalized walking control.
Powered lower-limb prostheses require tuning for stiffness and damping, for example, by a prosthetist in a clinic. This can be time and resource intensive for both parties, and hard to quantify because of the prosthetist’s personal skill in combination with unique user feedback. Huang and her colleagues used artificial intelligence (AI) to allow for automated tuning, making the powered prosthesis a more practical option.
In her previous approach, she designed an expert system, a branch of AI. The expert system encodes human knowledge, in this case a prosthetist’s, into a computer system as databases, rules, or policies. The initial step of the research involved testing the system with two able-bodied subjects and two people with transfemoral amputations. The team had the prosthetist and the expert system separately tune the prosthesis fitting.
After computer tuning, the researchers observed normative prosthetic knee kinematics and improved or slightly improved gait symmetry and step width with each subject. The expert system tuning required less time and no human intervention. However, tuning by the prosthetists reduced trunk sway, while the computer expert system sometimes led to slightly increased trunk motion. Building a more “knowledgeable” and effective computer expert system requires data collection from more prosthetists as they tune prostheses, which has not been available. Ideally, manufacturers of powered prosthetic devices could collect this clinical data, which would assist her in advancing the expert system, Huang says.
In the current phase of the research, Huang and her team, in collaboration with Jennie Si, PhD, Arizona State University, are addressing the challenges in the expert system using reinforcement learning–based control. This method uses AI to learn prosthesis control parameters and gait performance of individuals with amputations independent from prosthetists’ knowledge. The team has tested this algorithm via computer simulation and able-bodied human subject testing. The results show that reinforcement learning can tune the prosthesis control to achieve desired gait performance. The team will continue perfecting the idea to ensure the robustness and effectiveness of the algorithm for automatic prosthesis tuning. Huang and Si believe that AI is a powerful tool to make the advanced powered prosthesis smarter and more functional, and an AI-based tuning system can personalize prosthesis control toward desired gait performance and the user’s perceived walking stability and effort exertion.
For more information about
Huang’s research projects, visit
Kenton Kaufman, PhD, PE, Mayo Clinic, Rochester, Minnesota
“Risk Factors and Associated Costs of Secondary Health Conditions Among Adults with Transfemoral Amputations,” Kenton Kaufman, Hilal Maradit Kremers, Kurt M. Hoppe, Sue L. Visscher, Benjamin Mundell, Marianne T. Luetmer
The first report of the findings is expected later this year.
While the chances of being diagnosed with diabetes, hypertension, cardiovascular disease, osteoarthritis, and osteoporosis have been studied extensively in the general population, Kaufman’s research aims to determine the risk factors for those conditions specific to adults with transfemoral amputations, and the healthcare costs associated with those comorbidities. Kaufman’s initial focus is determining the likelihood that people with amputations will develop a secondary health condition and when that condition may develop. The research team will also look at healthcare costs associated with the secondary condition. Other studies quantifying the cost of amputation have been conducted, but they have not included analyses about how various comorbidities may influence the lifetime cost of amputation.
Prior research suggests that veterans with traumatic limb amputations have higher mortality rates from diabetes compared to veterans who sustained traumatic limb injuries resulting in disfigurement but not amputation, and, when compared to people without amputations, that having diabetes may lead to an increased mortality risk in people with dysvascular amputations; that people with traumatic amputations are about 1.5 times more likely to have osteoarthritis; and that people with lower-limb amputations are more likely to have osteoporosis.
Kaufman expects the results to provide evidence about outcomes for people with limb loss and the need for high-quality prosthetic care. “The findings can also be used by key partners in medical research and policy, healthcare providers, health payers, federal health agencies, and patient advocates to compare patient care approaches, develop clinical practice guidelines, and ultimately, reduce disability due to limb loss,” he says.
Courtesy of Shutterstock Inc..
The research project relies upon the Rochester Epidemiology Project (REP), which has been supported by funds from the National Institutes of Health for over 50 years. The REP links and indexes the medical records of virtually all individuals who have resided in southeastern Minnesota since 1966 with the intent that it be used by researchers for detailed studies into disease etiologies and outcomes.
Kaufman and his team isolated the data of 163 adults with transfemoral amputations, and, for comparison, matched each case with records of ten people of the same age, gender, and comorbidity. The researchers are standardizing the costs for comparison and using an annual time scale to provide estimates of risks of an event occurring within a given year. Kaufman says that the results may demonstrate how limb loss, which typically results in prosthesis users developing an asymmetrical gait and reduced ability to be active, affects the likelihood of developing the secondary conditions. “When compared to matched control subjects without limb loss, the decrease in activity will increase the rate of bone loss on the amputated side, and the reduced symmetry should increase the [chance of] osteoarthritis on the nonamputated side,” Kaufman says.
After completion of this phase of the research, Kaufman plans to use the REP data to study the use of pain medication among people with amputations.
For more information about Kaufman’s research projects, visit www.mayo.edu/research/faculty/kaufman-kenton-r-ph-d/bio-00077935.
Joan Sanders, PhD, University of Washington, Seattle
“Diagnostic Assessment of Limb Fluid Volume Changes in People with Trans-tibial Amputation: Testing a Clinical Monitoring Tool,” Joan Sanders, Katheryn Allyn, Janna Friedly, Brian Hafner
The research will be presented at the 2017 Academy Annual Meeting.
Increases and decreases in the fluid volume of residual limbs over the course of a day are an inconvenient fact for people who use prosthetic limbs. Sanders is testing a clinical monitoring tool to assess the daily volume changes in the residual limbs of people with transtibial amputations to help develop each user’s unique fluid-volume profile. Recognizing those patterns and factors may help practitioners find individualized solutions that best meet each patient’s needs.
Sanders and her colleagues, in the early stages of this three-year study, have developed a custom bioimpedance analyzer to monitor fluid volume changes in the anterior and posterior regions of a residual limb. The analyzer is a portable version of a larger instrument Sanders had previously used to evaluate limb fluid volume in laboratory testing. It injects a small electric current between two electrodes and monitors the voltage from other pairs of electrodes on the residual limb, which can then be converted to fluid volume using a limb segment model—a computational model, or equation—that relates data from the sensor to limb fluid volume.
To collect the data, participants’ fluid volume was monitored during 25-minute test sessions that involved equal durations of standing, sitting, and walking. For a two-week period prior to the test sessions, each participant had been monitored with a custom sensor to observe the times and durations spent sitting, standing, and being active while wearing his or her prosthesis. A fluid volume profile was created from these results and presented to the practitioner for analysis and discussion. The discussion involved interpretation of the data with the research team, based on their experience using the limb fluid volume monitoring tool. “For example, if a person doffs for ten minutes then gains and retains much fluid volume, that would suggest that periodic doffing may be an effective accommodation strategy for them, as opposed to adding socks,” Sanders says. “The objective is to provide the practitioner with useful insight based on our experience as researchers. We communicate with them to determine if and how they use the information in clinical care, so that we can get an idea of how much the measurements are helping.”
The initial test results show that there are four standard fluid volume–profile categories, with varying changes or lack of change when people are sitting or standing, or when they are active, or in transitions such as moving from sitting to standing. The results may help predict the effectiveness of different accommodation strategies that people with amputations employ, such as sock addition or removal, periodic socket doffing, or a change to elevated vacuum or size-adjustable sockets. Sanders and her team plan to increase the number of participants to about 60 to increase the data available for interpretation.
“We are set up to travel to Pennsylvania to test a lot of people this year,” she says. “Where we go depends on practitioner interest. If there is a group of practitioners in a region eager to participate in this project, we would go there.”
For more information about Sanders’ research projects, visit depts.washington.edu/bioe/portfolio-items/sanders.