American Academy of Orthotists & Prosthetists
Professionals Advancing Care Through Knowledge
Getting Started in Prosthetic-Orthotic Research
Thomas M. Gavin, CO
Avinash G. Patwardhan, PhD
ABSTRACT
The authors discuss getting started in O&P research and provide examples for organizing research projects, including how to identify the basic problem to forming the hypothesis, testing the hypothesis and drawing statistically or clinically significant conclusions.
Several common early errors in getting started with a project design are briefly discussed. Being concise, clear and significant are shown to be some of the most important factors in formulating a project. Constraints of time, funding and proper equipment to carry out research investigations should be considered but can be overcome and should not discourage the new researcher. Solving these logistical constraints is also discussed.
The impact of O&P research is paramount to the future of our field with the waning physician interest in carrying out O&P projects.
Introduction
Research is defined as a quest or investigation to reveal truth or to solve a problem utilizing scientific method. Research can be as simple as trying several different rivets to decide which is most suitable for a given application or as complex as developing a skeletal attachment for a prosthesis.
Correctly investigating an abstract theory requires much time. We may be inspired by seeing a repeating clinical problem or by realizing that a clinical method we are required to perform was founded intuitively and based on unproven assumptions.
However, it may require 10-15 hours per week for years to investigate a given problem. Clinicians sometimes abandon research because they were unaware of the time requirement before they started. Motivation and a good attitude are essential.
Frequently, ambitions are much greater than accessible funding and equipment for the beginning researcher. To start out, it is best to conduct studies that don't require significant funding and can be done in the lab or at home. Familiar surroundings will yield a certain comfort. This is the best low budget method to begin establishing a research track record. Clinical studies of patient acceptance or outcome of a given method represent a major portion of published research in O&P.
Researchers must have expert clinical understanding of the chosen topic. Prosthetists/orthotists who spend 90 percent of their clinical time in prosthetics may not be ideal candidates to conduct orthotic studies. Be sure to choose a topic relevant to your clinical practice. Conduct a search of the relevant literature to obtain a thorough understanding of not only the chosen topic, but of previous studies in this area Do not repeat conclusive studies.
Most departments in medical universities have full-time basic science re searchers and statisticians who actively pursue clinicians to become involve with or conceptualize research projects. Basic scientists are problem solvers, and statisticians are analysts, both in need of problems to solve and analyze. Clearly, the clinician can present problems, analyze the clinical significance of the problem, conceptualize the study and analyze the clinical pertinence of the results.
Identifying a Problem
A problem is the basis for any project. Solving a problem also may answer a perplexing question. Common problems are clinical inconsistencies. During the daily routine in O&P we see much repetition and should be able to separate consistent from inconsistent outcomes. For example, we may fit 100 ankle-foot orthoses in a year and see a consistency in improved gait. However, we may see all these patients for a one year follow-up and notice that knee extensor tone and power improved in some patients, remained the same in some and decreased in others for no apparent reason. This is a clinical inconsistency and would serve well as a research problem.
Another common problem is patient acceptance. Cosmesis, weight and heat are common complaints about prostheses and orthoses. Development of newer devices that address these problems frequently arises from lack of patient acceptance. The shorter thoraco lumbosacral orthoses (TLSO) to replace the Milwaukee brace in the treatment of scoliosis, the postoperative TLSO to replace the Risser cast and the endoskeletal prosthesis are direct results of this problem.
One of the most popular questions to investigate is which device will function best for a given condition when several orthoses or prostheses may be selected (comparative analysis). For example, a multitude of knee orthoses exists to treat the injured anterior cruciate ligament (ACL). All claim to be adequate in relieviing strain on the ACL and other associated ligaments; however, studies have shown some of the devices provide only marginal function while others yield better results.
Answering the question why is invaluable to better understand research methods. Why do we place the thoracic pad of a Milwaukee brace at and inferior to the apex of a curve? Why do tonereducing ankle-foot orthoses reduce tone?
Forming the Hypothesis
A hypothesis is an educated guess. Before forming the hypothesis, answer some preliminary questions.
Is the chosen topic clinically relevant? The turnbuckle spinal cast of Hibbs was abandoned several decades ago when the Harrington rod became a standard, so investigating the distractive forces of the Hibbs cast is no longer relevant.
Has the problem or question been addressed previously? Is the method chosen to investigate the problem satisfactory? Is technology available for investigation adequate, and is the design of the research ethical and legal?
Are the variables too great to extrapolate a reliable answer? If the project appears to be feasible, conduct a thorough literature search and form a base of data.
Pertinent literature would include biomechanics and kinematics of the device or method variant; the pathology being treated, including the pathomechanics, neurology, physiology, and kinesiology; and all relevant clinical literature, including conceptual, theoretical and result studies. Once a wealth of knowledge has been obtained, the researcher must decide exactly what to investigate.
At this point many variables are apparent. Be clear and concise when stating the hypothesis, and do not include tangents or too many variables. Each variable has an exponential result in increasing the length of the study, and tangents are usually equivalent to doing a separate study.
For example, the hypothesis, "Willner's instrument of spinal stabilization is a good predictor of the outcome of orthotic treatment for chronic low back pain" is simple, concise and to the point. Willner's instrument is the device, chronic low back pain is the pathology and the outcome predictor is to be investigated.
An example of an overly ambitious hypothesis with variables and tangents would be "Willner's instrument of spinal stabilization set in extension for discogenic patients and flexion for spondylotlisthesis and stenosis patients will be a good predictor of the outcome of orthotic treatment and will assist in delaying and preventing lumbar fusion."
Immediately the statistical analysis is severely complicated. Instead of analyzing one group of patients to predict the desired outcome of treatment, as in the first hypothesis, we have now added several dimensions. We have outcome predictor to investigate, variables of flexion of one group of pathologies and extension in another, and a tangent that may be unprovable in delaying or preventing surgery. This clouds the primary point of investigation and leaves the researchers, statisticians and reviewers confused.
In the discussion aspect of the final paper one may expound on the variables and tangents as long as they are identified as postulations and conjecture. Discussions of this nature may motivate the reader to continue this research. The variables and tangents are not part of the hypothesis and therefore have not been investigated. Take the project one step at a time and be thorough when investigating the concise hypothesis.
Methods
Once a problem worthy of investigation is identified, is found, knowledge about the subject has been gathered and the hypothesis has been formed, researchers must decide how far to go. To make this decision the clinician-researcher must consider all limitations. To be journal-worthy and accurate, research must have a solid design, and results must be backed up with sound statistical analysis.
If these criteria are not met, even the most clinically significant study on the largest patient populations will not be credible and will likely be inaccurate. Following established criteria will help to accurately reveal clinical truths and enable those truths to be shared through publication.
In vivo studies are studies carried out in real situations. For example, if the study is designed to investigate skin pressures in ankle-foot orthoses and the measurements and data are derived from patients or normal human subjects, this is an in vivo study. If data are derived from a mathematical model or a surrogate, it is considered in vitro. There are advantages and limitations to both types of studies. In vivo studies are considered to be more accurate; however, all studies cannot be carried out on live human subjects.
In Vivo Studies
Retrospective studies are usually done with data derived from pre-existing conditions. Although these studies are objective, they usually lack reliability and accuracy because facts are gathered from patient records and memory instead of established research methods. These studies usually yield some relevant conclusions; however, the quality would have been improved greatly if the study were designed prior to fact gathering.
Prospective studies are carried out with the research design done prior to creating the conditions to be investigated. This allows the researcher to alter record keeping to derive desired data. Patient questionnaires may be prepared and additional measurements may be taken during routine patient management.
The researcher must be careful about the degree of objectivity practiced in this type of study. Although one may have high-quality research design and derive all data correctly, the results may be skewed by variables in patient management since the clinician is also the researcher and is aware of the hypothesis he has set out to prove.
The prospective study is the study of choice since no design conditions are based on existing records or memory for the collection of the data.
In Vitro Studies
The advantage of in vitro studies is that new ideas may be tried without risk to human life. An example of an in vitro study would be to investigate the effects of a new TLSO designed to treat scoliosis in the high thoracic spine. These orthoses are not and have never been advocated to treat these high curves because it has always been thought that the Milwaukee brace is the only orthosis designed to control scoliosis in that area of the spine. To speculate that an intuitively designed orthosis may work and to then test it on a patient is not only unethical but inherently dangerous and risks putting a patient through very risky surgery.
It would also be irresponsible if an experienced clinician researcher designed an orthosis that enables patients with this condition to be treated with a more acceptable, more cosmetic orthosis yet not do everything possible to see if it works. This is a case for an in vitro study.
A mathematical computer model of the scoliotic spine and the new orthosis may be constructed to analyze the new design for effectiveness. The model may also allow for the design to be made even more effective by exploring variants of the design. Once the modeling study is carried out, the results of this in vitro study, if acceptable, may be used as a basis to test the design on patients and to design an in vivo study. Cadaver and animal models also may be invaluable tools in laboratory studies.
The shortcomings of in vitro studies are that mathematical models are not capable of considering all the variables of the human body. Animal surrogates do not possess the same properties as humans, and cadavers no longer have functional muscle, live bone or functional organ systems. These studies only yield relative data and are useful in designing in vivo studies and exploring variables.
Statistical Analysis
In the modern world of medical research, statistical analysis is a must. Did the researcher have significant power in the results? Was the population of a statistically significant size? Was the percentile error low enough? Did the researcher carry out the correct pairings test? Are the results reliable? These are just some of the questions to address.
Most major research funding agencies like the National Institutes of Health and the Department of Veterans Affairs require proof of statistical design before funding a research proposal. Unless you have a solid academic background in statistics, it is best to either co-author with or consult a statistician before designing the project.
Overcoming Complexities
In the prosthetic-orthotic world it is difficult enough to practice and understand the daily technical and clinical complexities of our profession, and sometimes it appears overwhelming to consider designing and conducting studies that include statistical analysis. Most physicians and other health professionals have decided that the only way to be credible in research is through the multidisciplinary approach.
Usually research projects consist of investigators who are health professionals, sometimes several from multiple specialties, basic scientists and statisticians. Single author papers are usually strong and accurate in one or two areas but may lack statistical or design strength, making a feasible project lack credibility.
Begin with a Pilot Study
Once you feel your project is feasible, try it on a small scale. Don't worry about statistics at this point, just try to get a feel for it. Try five or six cases, do measurements, collect data and see where consistencies and inconsistencies are. This is commonly referred to as a pilot study and is essential to calculate the amount of time and resources the project will require.
Most funding agencies require results of a pilot study to be included with the grant proposal. Pilots are usually low-budget, simple ways of getting started and working out the "bugs." They also allow clinicians to make preliminary predictions.
Conclusion
Prosthetics and orthotics is our profession. We can no longer sit back while physicians and scientists do research for us. In the past, prosthetists/orthotists may have fit the prosthesis or orthosis, read the manuscript, made last author on the paper or given an acknowledgement. The time has come for us to be co-investigators and co-principal investigators. We need to formulate the study, lead and assist in all phases of gathering and analyzing data, and draw the conclusions.
Do not feel that conducting research is "over your head" or is too difficult. Research is like anything else. Once you do it a few times you will wonder why you did not begin sooner.
Thomas M. Gavin, CO, is director of clinical services at BioConcepts Inc., Burr Ridge, Ill., and a research orthotist with Orthopaedic Biomechanics Laboratory, Rehabilitation Research and Development Center, Department of Veterans Affairs Hospital, Hines, Ill.
Avinash G. Pawardhan, PhD, is a professor in the department of orthopaedic surgery, Loyola University Medical Center, Maywood, Ill., and is a director of the Orthopaedic Biomechanics Laboratory, Rehabilitation Research and Development Center, Department of Veterans Affairs Hospital, Hines, Ill.
