Rehabilitation engineering as the crow flies
J. Foort, R. Hannah *
S. Cousins *
Part I—Development Of The Biomechanics Clinic Team
This series of five papers, three of which are presented here, provides an overview of rehabilitation engineering from the development of a clinic team through to problem-solving techniques. The first paper discusses the development of the biomechanics clinic team, identifies the differences in engineering and medical approaches to patient problems and proposes guidelines for successful teamwork. The second paper demonstrates that the five parameters of motion, force, neural function, shape and tissue quality are basic to information needs for biomechanical problem-solving. The third paper describes a framework for problem-solving that has been successfully employed for both device and process development. The fourth paper delineates the criteria and constraints that determine clinical viability of "products" in rehabilitation engineering and the last paper of the series outlines a very productive dynamic group problem-solving technique. The papers are intended to enhance communications and demonstrate a more disciplined approach to rehabilitation engineering. The remaining two papers will be published in a forthcoming issue of Prosthetics and Orthotics International.
The biomechanics clinic is the prime arena for rehabilitation engineering and hence factors which influence development and define functions of team members need to be considered. We have undergone the process of developing such a team at the Canadian Arthritis and Rheumatism Society in Vancouver, detecting in the process a number of factors which lead us to recommendations for others who may wish to develop such a team.
It is the inclusion of engineers that creates both expectation and dismay among team members who are not really prepared for the dichotomy between the medical and engineering role.
The major differences are:
Medical people see the patient as a problem.
Engineers see the patient as an example of a problem.
The action period of engineers is long by medical time standards.
Medical people want an immediate, unique and complete solution to a particular patient's problem, while engineers want general solutions for populations. Once these differences are grasped, there is some hope that the team can work together toward real solutions. In the meanwhile, the pitfall for the engineers is that medical people may press for interim solutions to resolve this immediacy of need by having engineers adopt and modify existing solutions. Unless this serves as part of the educative process for the engineers that helps them to become familiar with existing solutions, it is a waste to have them do what prosthetists and orthotists are better trained to do. The engineering role is to define and solve problems. This is time consuming as it involves considerable personal education on the specific problem before anything useful can accrue.
The dichotomy eventually evaporates when good results are achieved through team action and when these results can be identified as a product of team action, that is, when a solution evolves which is clinically acceptable. For this to happen, there will be recognition that a given solution is transferable from the engineering to the regular treatment programme. New items can be transferred only when acceptable to the people who must use them.
Acceptability usually means that the solution fulfils such requirements as;
improving the function of patients
reducing work for the treatment staff
having a cost benefit
providing new insights.
Barriers to transfer need to be considered by the engineering people so that their solutions can be organized to offset these barriers. The barriers are;
new solutions make development of new skills necessary
they cost money
they make different demands on practitioner's time.
When these barriers are overcome by the benefits of either improved function, less work, lower costs, new insights or some worthwhile combination of these, then the new item will be accepted.
The solution and temptations to avoid
The team should be organized for mutual education. This can be achieved initially through involvement in modest practical research related to improving patient benefits through technology. Better use of what exists, trying new things (evaluation programme), literature reviews, using engineers in an advisory capacity at first, or using solutions they have already established competence in, are ways to initiate development. We have attempted to use and demonstrate positive problem-solving tech-niques, where negative feelings are reserved while team members strive to be supportive or to offer even better ideas as a solution is evolving. Members of the team will soon develop enough familiarity and confidence to cross each others professional boundaries under the influence of such positive feedback thus engendering the prospect for the overlapping of roles which exists in a mature team.
There is a temptation to get a "big gun" into the team during team development, and this can work well enough if he does not dominate. The truth is that local talent is too easy to undersell. The team can develop well through the mutual education that goes on spontaneously and formally in any com-munity large enough to warrant a biomechanics clinical team. Academics appreciate the assoc-iation with clinical research activities, especially when there is an education component and so here we use Master programmes to get the information we need for problem-solving and to encourage academic commitment. At the same time, we would like to see a portion of the other team mates' time earmarked for research on relevant aspects of clinical problems as a natural part of their service commitment. Academics, committed to science can also benefit from having clinical responsibilities when their professional area is relevant.
Another temptation is to believe that a sophisticated laboratory and equipment are basic requirements, this is sometimes fostered by people who are experienced in the biomechanics field because of their acquisition of skills in unison with the development of facilities in previous positions. The truth is that, for the engineers who are backed by academic personnel and their facilities, little is needed that cannot develop out of the sort of research that must be part of problem-solving in the clinic. What really is required is improvement of facilities, such as occupational and physical therapy, that deliver the services to the patient. Such improvements include instrumentation for measuring various functions (motion, shape, etc.) and equipment with which to make things, especially simple things that can be used to supplement existing aids and equipment. An example might be a vacuum former in an occupational therapy department. Approaching the facilities problem in this way has forced us as engineers to fraternize productively with both the academics and the treatment staff for mutual gain. As engineers, we have only office facilities and storage space. For everything else we depend on existing facilities tailored to other ends. You can imagine the benefits gained in terms of overheads, keeping machines busy or maintained, etc., but on the other hand there are the demands made on us in order to maintain productive good will. Research is a good cement for all.
When seeding of the team through exploitation of what exists and what can be conveniently improved has reached a stage where demands on the team indicate it, then is the time to expand backup facilities to serve the treatment and research functions of the team. Thus, having reached close to that stage here, we would expect facilities which would allow us to make more complicated things of a one-off nature or to make pre-production runs which cannot be put into a manufacturer's hands and which are inappropriate in a university or other institutional facility such as a limb shop.
An inhibiting organizational fact is that biomechanics clinic teams are too strongly associated with research when in fact they should be identified with service at the interface between knowledge and ignorance. Besides the research and service function, their relationship to education is obvious. For the engineers, even more strongly identified as researchers, the big advantage is that they cross institutional and professional boundaries at will. Such mobility needs to be the right of all team members. Eventually, when the benefits that can derive from biomechanics clinic teams and the association with engineers are assured through the establishment of rehabilitation engineering, the aim should be to keep the engineers community and patient oriented and to keep access to technology open through a process of continuing education, a prolongation of post graduate studies.
Functions and responsibilities
In the day to day operation of the biomechanics team, among the influences on productivity are the willingness of those on the team to share functions, their flexibility in terms of time use, and their relationships with patients serving as test wearers or subjects. We set a morning aside for clinical activities, that is seeing patients, and followed up with a very flexible time allotment using "free" staff (usually the researchers) or treatment people with open time assignments.
On patients who are good representatives of the problem, the team will try out existing and new ideas in a mutually beneficial way. The purpose of this, besides improving a solution which is developing, is to uncover the prescription criteria which govern use of a particular solution at any stage. This is achieved by fitting patients who vary from ideal in ever increasing degrees of complexity to determine the boundaries beyond which the solution does not work. The patient on whom it does work defines the criteria for its use! This is an important function of the team. It may be evaluating the solutions of others as well as their own. The purpose is to have available for such a clinic a variety of solutions to a range of problems so that decisions can be made quickly and effectively to improve patient function and management.
Biomechanics clinic teams are responsible for developing prescription criteria for the devices they evaluate. When they have so established criteria for the clinical application of any device or procedure they are in a position to educate, field-test, and move on to other researches ... their role is that of problem-solving. Others pick up where their functions are complete.
We estimate that a biomechanics team including a rehabilitation engineer can be supported by a population base of 200,000 or over, and that one engineer per 200,000 is a reasonable ratio.
The best setting is within a rehabilitation unit with occupational and physical therapy staff and associated social workers. The skills of these people will expand as they relate to engineers and to each other, each taking on some of the functions of the other. The skills of the engineers must be extended by further technical training as well as adopting the skills of others. Such administrative details as establishing engineering positions for rehabilitation will come in time.
The biomechanics clinic team needs to establish a position on the development of devices and procedures of potential economic value. The main purpose of the team is to get help to their patients and other considerations are secondary. The broader responsibility is to extend the use of good solutions as widely as possible. If this conflicts with economic interests, economic interests must yield.
Part II—Information Needs For Biomechanical Problem-Solving
For rehabilitation engineers to be relevant beyond the ordinary in the clinic, they must aim to obtain hard data for use in problem-solving whenever possible. Some data can be found in scientific literature. Some must be found directly by measurements made on the patient.
When we measure for problem-solving in the clinic, we gain the advantage of information which is in a form easily grasped and which can be used to describe, compare and classify in universally useful terms. We mean by useful that the information be as exact as the end use requjres. The basic aim is to use measurements to define the patient's existing level of function and to estimate his needs so that the gap can be narrowed through clinical action. Visualize a typical clinical situation in which a bio-mechanical approach to problem-solving is called for. We can easily recognize that the ability to move is a crucial factor to the patient; forces between body parts and between the body and the environment are also relevant. Neural function, the basis for control, is important. In order to do anything useful by means of external support of body parts, additional factors which are significant include the shape of body parts against which forces must be directed and the quality of the tissues which must bear them.
So, we recommend that rehabilitation engineers concentrate on measuring; motion, force, neural function, shape and tissue quality.
The objectives of making such measurements should be clinical and be intended to aid in assessment, defining status, identifying changes, classifying disability, providing design data and comparing results.
Such information will be used to aid in making predictions, giving protection and developing means for improving function.
We recommend that rehabilitation engineers approach clinical problems of a biomechanical nature with the view that the patient is central and that the measurements are incidental in the sense that their exactness is defined more by the results achieved for the patient than on their reprpducability, exactness in detail, or how nearly they match ideals established by more exact scientific means. It is remarkable, for example, how much insight can be derived by nothing more complicated than comparing function on one side of the body to function on the other. We see asymmetry of gait as important for defining quality of function in prosthetic knee implants and have used this asymmetry to make predictions. We make this recommendation for a sort of opportunism in the use of measurements as a counter balance to the scientific obsession with exactness.
Measuring such quantities as motion, force, neural function, shape and tissue quality will then rest on the pragmatic premise that the quality need only serve the end. Often no more than go-no-go information is sufficient to aid in making an intelligent and useful clinical decision. For example, we note that people with prosthetic knees are maintaining their knees in a position of fixed flexion throughout stance phase. An obvious possibility is that they are splinting the knee with their muscles for some protective purpose. It would be very easy to demonstrate this with simple off-on muscle myography and then use biofeedback to aid them in judging the position of the knee as they stand on it.
If immediate usefulness is the objective for data collection in the clinic, then some boundaries are placed on how data is processed. We recommend that data should be delivered while the patient is available, preferably within seconds or minutes, and that it be in a form which is easy to read. It can then be used for deciding right on the spot while attention on the person is keen, and all the other factors are in view. In our studies of motion we get a direct readout on a strip chart recorder. The records are considered at the time in some instances, and in others pondered over for the preparation of an analytical report. Records of motion, for example, are easy enough to make some sense of when the discrepancies are gross; the eye alone is also a good instrument at such times. The difference lies in the capacity to remember for purposes of later comparisons. We find that there are obvious things only the very best eye will detect but some which the eye may not detect yet which the instrument will easily find. As the patient is more important than the measurement or the system of measurement, so the measurement is more important than the reference system when we speak clinically! We realize that this may not suit the purist, but in fact there are many measurements that can be made which are strictly related to the patient at an instant in time and which are not considered as fodder for future analysis or statistical collections. Just recognizing a pattern can be clinically useful. Just knowing that the force through a cane is under or over a given factor can be useful. What we look for in any information collection process is enough insight to make the intuitive mind click for a better clinical result. If such inspiration can be derived from relative information it would be folly to be pure.
Of course, the question of equipment immediately arises. We see a number of high quality, technology-intense laboratories studying gait, for example, and we ponder on how the products of such laboratories can be brought directly into the clinic. If we allowed the demand for measurement to be generated out of patient needs we would see something quite different developing for clinical use. First of all, measuring equipment would be portable and have different modular elements (especially transducers) to permit varied, direct and convenient use in the clinic. Equipment would evolve which ensured that at every stage in development something of clinical relevance was gained. There might be motion measuring equipment that quantified only a small amount of information at a time, to be assembled as needed. Biofeedback equipment could be refined into force measuring systems in response to obvious clinical needs.
The temptation is to embark on a programme of developing equipment to carry out such clinical measurements, thumbing through the myriad catalogues of electronic components and handbooks while keeping in mind the last scientific advance. Better to check the equipment that exists first and proceed through a course of minimal modification toward the end that you have rather than make a measurement system. When you must (as you well may) have equipment that is different and new then the kind of product wanted needs to be thoroughly established before design and development work is considered. Easily 80 per cent of the costs before usefulness are in such developments, only a minor part being costs of defining a solution. Any attempt to upgrade equipment should relate to ease of operation, convenience to the patient and delivery of information of more use in the clinic. It is worth remembering that while we may wish to replace the medical professionals with instruments there is no chance that we will, and such information as can be derived from measurements will only be used if it speeds up clinical processes, cuts costs, improves results for patients and provides new insights. The information provided through measurements is only part of the information being used to make judgements. Clinicians will establish the criteria for the design of special equipment which they will ultimately use or abandon. Their criteria will relate to the clinical function. As an example, we know that a single curve rather than a stream of similar curves is favoured. This could lead us to microprocessing or some different way of displaying data. But ultimately it is for the clinicians to choose.
That aspect of equipment development which has no clinical relevance—the hardware aspect —is best kept right out of the clinical environment. If fundamental design factors are involved academic people should take them on. If they involve adaptations of equipment and techniques already in use, industry should deal with them.
Use of resources
No one or half dozen rehabilitation engineers who are trying to solve clinical problems can cope with doing everything that arises as attempts are made to use objective means to supplement subjective findings in problem-solving. It is better that there is co-operation between various laboratories and clinic teams so that sharing evolves. It is quite feasible for one group to become masters of motion measurement for clinical service, and another, through an interest in biofeedback, to become expert in force equipment use, or on matters of tissue quality or shape. Highly developed expertise in a limited area by specific individuals or groups can open the way for other groups to adopt the application of techniques and equipment which the more experienced recommend. We recommend a modular approach to the assembly of skills as well as equipment and procedures. If such a mood were created or existed, specified laboratories would take responsibility for discrete parts of the whole process of using measurements in clinical processes until one day all would be able to use everything useful by this means, and therapists and doctors rather than engineers would be manning the pumps.
Well established laboratories designed to produce refined measurements would carry out work which provided statistical information and established standards. They would take data from clinics for analysis against standard information to give back results for use in the clinic and generally carry out what is accepted as scientific enquiry. They could also use their data bank information for modelling systems so that comparisons could be made between such models and patients for postulation of possible solutions.
Rehabilitation engineers should not be so equipment and process conscious as to forget the value of training and experience. Collectively we know very much more than any one of us. Part of this is training, part perception, and part experience of practice. All of this needs to be organized into a brain-bank which ensures that people with well defined problems can be matched as nearly as possible with people who possess solutions. This also means that there should be geographic fluidity so that people can move around and be shared without jealousy.
While we recommend that rehabilitation engineers become involved and informed in matters of measurement, and specifically in measuring motion, force, neural function, shape and the quality of tissues, we realize that they will also get involved in true research because they are involved in design, development and clinical measurements, but hopefully, they will resist the temptation to escape into purely scientific studies until they have ceased to identify themselves as rehabilitation engineers.
Part III—A Framework For Problem-Solving
Our interest in modern problem-solving techniques led us to examine our successful projects for a logical sequence. From this we derived the problem-solving framework shown below.
Stage 1—developing a hypothesis:
phase i —problem identification
phase ii —problem definition
phase iii —solution generation
phase iv —solution selection
phase v —modelling.
Check against criteria for this stage. Abandon, recycle or go on to Stage 2.
Stage 2—testing the hypothesis;
phase vi —clinical prototype fabrication
phase vii —prototype development
phase viii —prototype evaluation
phase ix —test batch production.
Check against clinical criteria. Abandon, recycle or go on to Stage 3.
Stage 3—solution transfer;
phase x —field test quantity production
phase xi —field testing, functional refinements
phase xii —dissemination of information
phase xiii —supply development.
Continuing improvements through user-maker dialogue.
The three main stages through which our successful projects seemed to pass were (a) the assembly of information which would permit us to form a hypothesis ; (b) collection of information which would establish the merit of the hypothesis; and (c) transferring the solution to users and producers.
We consider this framework a good model for others in rehabilitation engineering to adopt. It will enhance communications within and between groups by referring to the stage or phase of a project in a way that avoids confusion as to what is happening. This would be particularly useful to people who are working on segments of a general problem which must eventually generate a total solution. Different workers could be paced to meet at a particular point in time and so be organized to optimize results. Granting agencies would find such a framework useful. They could fund to the level of each main stage, abandoning projects or continuing as results indicated, with the degree of risk well defined for each stage. Reporting would include reference to the stage and phase. The framework could also be usefully employed in support of grant applications. People within a group would have a clearer picture of project progress. This leads to other possibilities. When granting agencies or institutions have "skills" inventories combined with such a format as this framework, it would be possible to see that a project was getting beyond a group which had been handling it. Transfer to others more suitably organized for more advanced stages or phases could then be logically proposed and contracts let. This prospect of transferring projects to the best suited laboratory or workers suggests that grant applications themselves should be accorded a special status for which applicants could be rewarded, not necessarily by receiving it themselves, but by receiving funds for grant application development. When work on problems is considered in these terms it is quite apparent that different people suit different stages and phases. Rehabilitation engineering can be more fruitful by matching people and expertise to the most suitable phases.
It may be tempting to consider this as nothing new. We counter by noting confusion experienced in some of our own projects and in projects of others. Such an organizational form for problem-solving could start us on a path of enhanced communications without imposing a straightjacket. The outcome could be improved sharing of the work of rehabilitation engineering for more efficient delivery of solutions.
To clearly describe the problem-solving framework, we will work through the phases by example.
Stage 1—developing a hypothesis
Phase I—problem identification: We were told that lifting heavy flaccid children was a major problem and that a lifting device was needed. We looked at such children and watched them being handled by staff and parents; in and out of cars, on to work surfaces, into seats, beds, toilets, etc. We examined lifts and various other multipurpose devices. This involved about three months of strenuous effort and included development of many criteria against which we could evaluate some existing solutions. By the end of this exploratory period we felt confident enough to make a definition of the problem.
Phase II—problem definition: The result was to change the environment so that lifting was less necessary. We could see that by reorganizing our approach from "lifting" to "not lifting" we could solve the problem better.
More intense effort went into looking at what others had done to improve the environment as well as to lift such patients until we could, by checking these solutions against the criteria which were developing, see the advantages and shortcomings. It is important to do this because, not infrequently, something quite different emerges as a key factor. For example, it may be prohibitively expensive to "change the environment" to solve the lifting problem.
Phase III—solution generation: On another project, Fracture Bracing, we embarked on generation of solutions which involved going far outside the field of normal rehabilitation engineering. We looked at birds, snakes, geological processes, and a host of other things in nature to find a "bionic analogy" to our problem. The outcome was a variety of proposals for solving the problem that looked very different from what we had seen. The process of "measuring" with established criteria and defining the boundaries of possibilities or constraints ensured that any solution which rated high could be adopted. This solution could be something entirely new or something already in existence. When existing solutions are pertinent the rehabilitation engineer should be encouraged to adopt them. It is not necessary to reinvent the wheel.
Phase IV—solution selection: The solution that scores highest is the one to select. There is no need for procrastination at this point ("it will cost too much", "it will be too heavy", "etc.") because all the objections are accounted for in the criteria and constraints ! This includes a statement on costs and time and skills needed to raise it to succeeding phases. Thus, with a solution selected the evidence to support continuing to phase V will be included.
Phase V—modelling: Modelling is intended to be a super communication. There is more of "how it works" than "use it on a patient" in this phase, and there may well be more than one model because even though the solution has been stated in general terms, its eventual form is far from established. In the case of the CARS-UBC Brace for example, there were sketches, cardboard models and more sub-stantial models built over plaster models of legs as well as analogies formed to illustrate to other team members what was proposed. We even did "hands on" force demonstrations on each other and on patients. New criteria evolved out of this, some of which anticipated succeeding phases. Eventually, there was a demand to proceed to trials on patients (rather than to recycle through previous phases or abandon). The demand could as well have been to adopt an existing solution if it better met the criteria so far developed. Such an alternative solution would just as readily be cycled into the next stage.
Stage 2—testing the hypothesis
We had hypothesized that the knee brace for arthritics with medial or lateral instability should be an intermittent force application device which supports when the knee is extended and relaxes during flexion. With the experience of modelling, development of a clinical prototype is conceivable.
Phase VI—clinical prototype fabrication: A number of these can be made, each intended for a patient carefully selected to avoid extraneous factors. Thus for the knee brace, people with subluxation or flexion-extension instabilities were initially avoided. The braces were made on plaster cast models, with each prototype a variant on the theme in all but functional terms.
In another project, the Joint Motion Measuring System, reproducibility of curves was initially checked out on a joint motion simulator, but eventually came to be used on patients. This was a variation with the same aim, to develop confidence in those who apply it and in those who are subjects.
The emphasis in terms of team involvement is that non-engineering team members gradually shift from advisory to active roles as problem-solving reaches the stage of involving patients. Later, the emphasis shifts further as these members take over and the engineering members become advisory.
Phase VII—prototype development: This is very much the engineers' job but the other team members are getting involved via patient contact and through this are revising and adding criteria as they evaluate it. If the solution is fruitful, the engineers and supporting technologists are soon supplying devices in a fairly routine way to keep up with developing demand.
Phase VIII—prototype evaluation: In typical formal evaluation programmes everything is held constant as a device is tested. This is not the case in this phase. Rather, this is the time when engineers and other team members are on their knees around the patients. A great deal of pressure is felt at this phase (not only on the knees) because if the solution is good, it must be delivered in ever increasing quantity. This pressure is in itself a positive factor in design. It led us to mass production techniques, this being the only way we could cope. In making the shift from individual or custom fabrication to standardized fabrication, other improvements in design were adopted leading directly into the next phase.
Phase IX—test batch production: Just as first clinical trials represent the moment of truth in the clinic, test batch production is another moment of truth for potential production. The producer can inject new criteria into the design process to make the item producable and marketable. At the same time, if success is assured in the clinic, introduction of the device to a manufacturer relieves some of the pressure on designers and sets the scene for the final stage.
Stage 3—solution transfer
Barring obstacles that preclude further development, this stage is entered with the objectives of getting the process or device out of the hands of the engineers (as designers) and into the stream of clinical use. This involves essentially two things: (1) to ensure production; and (2) to ensure a market through education and the subsequent development of confidence among patients and treatment staff. There is no better way to prime the manufacturing or clinical pump than through field-test programmes. This time, the evaluation process is more formal because it is further from home and from the help of those who know most about it.
Phase X—field-test quantity production: This phase puts designers and the producer in close touch until they both deal confidently with the product. Emphasis will shift from the manufacturer advising to the engineers advising until the transfer is complete.
Phase XI—field-testing and functional refinements: At this phase, the designers advise on how the produced item can better serve its purpose as information arises out of field testing. There is now a three or four way stream of communication as the manufacturer faces the item as a product, the designers face users on issues of function and users face patients on issues of competence and function. As competence evolves at the clinical level through field testing, there can begin to develop a direct relationship between the users and producers to permit withdrawal of the engineers after successful transfer.
The dilemma designers face is when to let go. There is a reluctance to stay involved (tiring), especially if there is clearly success, but also some reluctance to make a clean break (anxiety —pride). This can be resolved by documenting and publishing results to aid in fostering general use of the system which has been developed.
Phase XII—dissemination of information: This takes many forms. Information is given in articles, through injection of new ideas and principles into the educational process, and so on. The manufacturer serves his interest by making his product known. Dialogue is generated between producer and user on an ever increasing basis. Enlarging programmes of field-testing may be necessary in some cases to improve the manufacturer's prospects for maintaining viability in the early stages and to develop familiarity among users and patients.
Phase XIII—supply development: The best compliment a design group can experience is demand for what has been designed. This is not necessarily a measure of merit however, nor is merit an assurance that this can happen. There will be of necessity a phase during which supply development may have to be promoted. This requires that the developers co-operate with others in furnishing information related to development of supply. Lecturing at conventions, writing articles, assisting on an advisory basis on demand are among ways supply development can be helped. Increasing geographic scope of field-testing in an outward wave from the design group using manufacturers and users as those who carry the message is warranted to keep a good idea from dying.
The problem-solving framework discussed is a viable way to organize solving any problem whether it is a process, a device or a programme. The terms may change slightly or be understood differently, but the process of evolution will endure. As an exercise, it was applied to the problem of alcoholism and absenteeism from work to see if it would hold up and it did. Within the phases of the framework are areas that require further consideration such as development of criteria for design in rehabilitation and more disciplined approaches to problem-solving but these will be discussed separately at a later date.