Development of a Theory-Driven Rehabilitation Treatment Taxonomy: Conceptual Issues

Archives of Physical Medicine and Rehabilitation journal homepage: www.archives-pmr.org Archives of Physical Medicine and Rehabilitation 2014;95(1 Suppl 1):S24-32

ORIGINAL ARTICLE

Development of a Theory-Driven Rehabilitation Treatment Taxonomy: Conceptual Issues John Whyte, MD, PhD,a Marcel P. Dijkers, PhD, FACRM,b Tessa Hart, PhD,a Jeanne M. Zanca, PhD, MPT,b Andrew Packel, MSPT,a Mary Ferraro, PhD, OTR/L,a Theodore Tsaousides, PhDb From the aMoss Rehabilitation Research Institute, Elkins Park, PA; and bDepartment of Rehabilitation Medicine, Icahn School of Medicine at Mount Sinai, New York, NY. Current affiliation for Zanca, Kessler Foundation, West Orange, NJ.

Abstract Many rehabilitation treatment interventions, unlike pharmacologic treatments, are not operationally defined, and the labels given to such treatments do not specify the active ingredients that produce the intended treatment effects. This, in turn, limits the ability to study and disseminate treatments, to communicate about them clearly, or to train new clinicians to administer them appropriately. We sought to begin the development of a system of classification of rehabilitation treatments and services that is based on their active ingredients. To do this, we reviewed a range of published descriptions of rehabilitation treatments and treatments that were familiar to the authors from their clinical and research experience. These treatment examples were used to develop preliminary rules for defining discrete treatments, identifying the area of function they directly treat, and identifying their active ingredients. These preliminary rules were then tested against additional treatment examples, and problems in their application were used to revise the rules in an iterative fashion. The following concepts, which emerged from this process, are defined and discussed in relation with the development of a rehabilitation treatment taxonomy: rehabilitation treatment taxonomy; treatment and enablement theory; recipient (of treatment); essential, active, and inactive ingredients; mechanism of action; targets and aims of treatment; session; progression; dosing parameters; and social and physical environment. It is hoped that articulation of the conceptual issues encountered during this project will be useful to others attempting to promote theory-based discussion of rehabilitation effects and that multidisciplinary discussion and research will further refine these rules and definitions to advance rehabilitation treatment classification. Archives of Physical Medicine and Rehabilitation 2014;95(1 Suppl 1):S24-32 ª 2014 by the American Congress of Rehabilitation Medicine

It has long been recognized that rehabilitation is in need of more precise and coherent means of defining, specifying, and measuring the contents of interventions for both clinical and research applications.1-4 Attention to this need has lagged behind the development of measures of both the inputs to the process (eg, patient characteristics) and its outcomes.5-9 As described elsewhere,10 we believe that a *rehabilitation treatment taxonomy (RTT) that is based on *treatment theory will lead to a better understanding of rehabilitation interventions and their effects. (Words and phrases that are specifically defined in supplemental appendix S1 [available on page A9 of this supplement and online at http://www. archives-pmr.org/] are marked with an asterisk and italicized when initially used.) No commercial party having a direct financial interest in the results of the research supporting this article has conferred or will confer a benefit on the authors or on any organization with which the authors are associated.

An important first step in creating such an RTT is to develop a set of rules or principles about what the taxonomy should cover and how *treatments are to be defined and grouped. As we grappled with these issues in the context of the wide variety of treatments relevant to rehabilitation, we recognized that they could not be entirely resolved with empirical data on the efficacy/effectiveness of existing treatments. Although empirical evidence will certainly affect its development, an RTT needs to be conceptually coherent, whether treatments are effective or ineffective; indeed, one of its main purposes is to guide systematic exploration of the efficacy of coherent classes or groups of treatments that share similar mechanisms of action. Thus, we assembled a range of treatments from published research articles, rehabilitation textbooks, and the clinical experiences of the authors. These sources were not reviewed as evidence along the lines of a systematic review (indeed, many of the treatments we reviewed were relatively poorly described and/or their hypothesized mechanisms

0003-9993/14/$36 - see front matter ª 2014 by the American Congress of Rehabilitation Medicine http://dx.doi.org/10.1016/j.apmr.2013.05.034

Rehabilitation taxonomy conceptual issues

S25

minimally discussed). Rather, we used them as examples of treatments to which our definitions and conceptual rules could be applied and tested. We used the empirical rehabilitation literature in an iterative process in resolving these conceptual issues, in which initial decisions were arrived at through discussion and documented in writing. Descriptions of treatments were then reviewed in light of those decisions. In this way we uncovered problematic implications of previous decisions and worked to refine or revise them until they appeared robust to the examples examined. In this article, we delineate the conceptual issues encountered and their implications and discuss how we have resolved them in our work on the RTT to date. There is no doubt that new conceptual problems will be discovered in the additional steps of taxonomy development11 and that new eyes will identify problems with the formulations developed to date. Although we are far from a practically useful RTT, we summarize the conceptual issues encountered to date in the belief that an understanding of these issues is critical for clinicians and researchers seeking to enhance theory-based treatment definition.

theory, which is why we believe that an RTT should be built around treatment rather than enablement/disablement theories. Treatment targets in rehabilitation are varied, including aspects of body structure, body function, activity, and participation, to use International Classification of Functioning, Disability and Health (ICF) terminology.13 Consequently, treatment theories relevant to rehabilitation come from many different scientific domains; the mechanisms by which muscle function might be changed are distinct from the mechanisms by which a new skill might be acquired. We also expect that specific rehabilitation disciplines or specific functional domains14 will not be isolated from one another in the taxonomy. For example, both physical therapists and speech and language pathologists may recommend devices that compensate for body function impairments (eg, prosthetic limb, cochlear implant). Therapists in both disciplines provide education and skill training of various kinds, many of which involve similar principles of change.

Treatment Theory as a Guide to Treatment Definition

It is generally agreed that rehabilitation goals should be selected, to the extent possible, through a collaborative process between patient and clinician. However, it is often the case that the patient will express goals in activity and participation domains, but they will not have the knowledge or expertise to assess body structure or functional limitations that contribute to the activity or participation restrictions, nor will they be able to identify the specific treatment ingredients that they should receive to achieve those goals. Thus, in this discussion we refer to the clinician’s selection of treatments and treatment ingredients. This is not to deny the importance of maximizing patient autonomy, but to reflect the complex reasoning process that a clinician must undertake to organize specific therapeutic actions in line with those larger negotiated goals.

We propose to define rehabilitation treatment in terms of treatment theory. (Here and throughout the article, we use the singular theory to refer to a class of theories that address a kind of questiondin this case, how treatments exert their immediate effects. We use the plural theories to refer to multiple specific theories in that class.) A treatment theory defines and links (1) the *target of treatmentethe aspect of the treatment *recipient’s functioning or personal factors that is directly altered by treatment; (2) the *ingredients that produce change in the target; and (3) the hypothesized or known *mechanism of action by which those ingredients cause change in the target of treatment. This tripartite structure of treatment theory is illustrated in figure 1. Many rehabilitation treatments are undertaken in the hope of enhancing aspects of functioning that are distal to the target of treatment. We refer to these distal functional changes as *treatment aims. For example, one might treat hypertonia (the target, an impairment of body function) with the goal of improving gait speed (an aim). This expectation, however, rests on a different theoretical foundation, *enablement/disablement theory.12 Enablement/disablement theory concerns itself with the relations of abilities and limitations located at various levels of functioning, within individuals. It defines the interrelations among domains of functioning but does not identify the methods by which changes may be made in any one of these domains. In the previously mentioned example, it is quite possible that gait speed is primarily limited by muscle strength, with only a minor contribution of hypertonia. Enablement/disablement theories would address how muscle strength and hypertonia relate to gait speed, and in so doing might help one determine whether muscle strength or hypertonia should be the target of treatment. However, enablement/disablement theories do not identify the specific treatment ingredients or mechanisms of action that could be applied to effect change in muscle strength or hypertonia. These are addressed by treatment

List of abbreviations: ICF International Classification of Functioning, Disability and Health RTT rehabilitation treatment taxonomy

www.archives-pmr.org

Setting Treatment Goals and Selecting Treatments

Target of Treatment A treatment target, as previously defined, is not simply an area of functioning that needs to improve (eg, reaching, preparing a hot meal), but a measurable aspect of functioning in that area. For example, for reaching, one could measure progress on speed, distance, or precision; for meal preparation, one could measure time required, avoidance of hazards, or the success of the final product, as rated by others. This degree of specificity is critical because the mechanism of action of a treatment is described with reference to the change in a specific target. But then how do we distinguish among closely related treatment targets within the same organ, or person? For example, is the target of progressive resistance exercises improved muscle metabolism, increased central drive, muscle hypertrophy, or increased muscle strength, given that all of these may be occurring simultaneously? For practical reasons, we propose to define the target as the most directly affected functionally relevant aspect that is hypothesized to change. If we found a change in muscle metabolism without a change in strength, the treatment would be considered a failure; if we found an increase in muscle strength without an improvement in muscle metabolism, we would still consider the treatment a success (but would need to reconsider our assumptions about mechanisms). Therefore, the target should be expressed as increased muscle strength. In identifying the target of treatment in this way, we consider physiological steps that are precursors to this target to be part of the mechanism of action of treatment. In this example, we recognize that physiological contributions (ie, changes in central drive, or changes in muscle metabolism)

S26

J. Whyte et al

Fig 1 Illustration of the basic structure of all treatment theories. Ingredients (on the left) exert forces of change on a treatment target (on the right) through a known or hypothesized mechanism of action. A treatment must contain the defining essential ingredients pertinent to the specific treatment theory and may also contain other active ingredients, each of which contributes to the effects on the treatment target through the specific mechanisms of action. The inactive ingredients, although delivered with the treatment, exert no effects on the treatment target.

might be part of the mechanism of action. If the treatment was being pursued by a bodybuilder for cosmetic reasons, the target would likely be size (muscle hypertrophy) rather than strength, and the specific ingredients and mechanisms of action, although related, would likely differ as a result. Treatments whose targets are changes in the level of skill with which activities are performed, or changes in knowledge or attitudes, are particularly challenging to conceptualize. There is an infinite number of activities or categories of knowledge that might be addressed in rehabilitation. However, we wish to classify a finite set of treatments according to their active ingredients, and avoid the need for defining, distinguishing, and investigating mechanisms of action for shirt donning treatment, underwear donning treatment, and sock donning treatment (or teaching side effects of aspirin, teaching side effects of warfarin, etc). On the other hand, there very well may be meaningful differences in the training technique for donning socks versus donning underwear, such that one cannot completely ignore the activity when considering the nature of the treatment. In order to cope with this issue, we propose to define the treatment primarily with respect to the training approach, but to retain information about the content of the activity or knowledge pursued. In a future system for treatment notation, one might consider a format in which the content that is trained or practiced is represented in bracketed form. Thus, all of the dressing activities might receive the same treatment, roughly corresponding to guided practice of a sequential task with coaching (donning/doffing socks, underwear, shirt). This type of notation would maintain the relative importance of mechanism versus context. Such notation requires more development before being put to actual use to avoid listing numerous contextual items without a guiding principle. The ICF makes an important distinction between body functions and activities. In our initial efforts at organizing rehabilitation treatments, we sought to differentiate treatments that have as their

target aspects of functions (eg, coordination of voluntary movements) versus aspects of activities (eg, buttoning a shirt), both of which improve with practice, feedback, and other ingredients associated with learning. However, operationally specifying how treatment ingredients differ when addressing a function target versus an activity target proved very challenging. We reasoned that this is because one cannot practice a function in isolation. Coordination of voluntary movements can only be practiced in the context of the performance of some task or activity that also requires the integration of many additional functions (eg, visual perception, involuntary movement reactions). Thus, whenever practice is an important treatment ingredient, it will typically be practice of an activity rather than a function. In some cases that activity is functionally significant in its own right, but sometimes it is merely the delivery vehicle for practicing or strengthening the function. How are these to be distinguished? In the example of exercises performed to increase strength, the activity of lifting an object (eg, a weight) is the means by which a mechanical stimulus is provided to the muscles that results in increased production of force. Ultimately, performance-based treatments that have improvement in a function as their treatment target will likely pursue *progression of treatment along an identifiable function dimension (eg, when the target is coordination of [small] voluntary movements, the therapist will likely prompt repeated handling of smaller and smaller items), whereas a target, such as quality of dressing, will likely be addressed using treatment progression in relation with the internal structure of the activity (eg, adding articles of clothing to the sequence as each is mastered).

Treatment Ingredients Treatment ingredients are observable (and, therefore, in principle, measurable) actions or pieces of information or chemicals or www.archives-pmr.org

Rehabilitation taxonomy conceptual issues devices or forms of energy that are selected or delivered by the clinician. Ingredients are systematically applied by the clinician toward specific targets, even if their delivery is contingent on recipient behavior (eg, a specific consequence to some behavior must, by definition, follow that behavior, but can do so according to a plan). Treatments contain an infinite number of ingredients, if we include the specific words used by a clinician during treatment, the time of day of the treatment, and so forth. However, as is the case with drug treatments, we presume that some of those ingredients are active, in the sense that they carry the therapeutic power of the treatment (specific drug compound), and some are inactive (fillers and capsule color). Although we lack complete understanding of which ingredients are active in rehabilitation, for most of them we have at least implicit hypotheses. After all, we speak of serial casting, not of serial casting in a large room or serial casting with purple casting material. Thus, active ingredients are those that are hypothesized or proven to exert an effect on the target of treatment, and inactive ingredients are known or presumed not to have such effects. We find it useful to further subdivide active ingredients into *essential ingredients and other active ingredients, and to define *dosing parameters for each. By essential, we refer to ingredients that define a particular treatment category: if the essential ingredients are present, we have treatment X; if they are absent, we do not have treatment X. In contrast, (nonessential) active ingredients are hypothesized to moderate (enhance) the effects of treatment X, but they are not necessary for the effects to take place. This is not meant to imply that essential ingredients are necessarily more important or more potent than other active ingredients, simply that they are essential to the classification of the treatment in question. This distinction between essential and other active ingredients prefigures a categorical nature of the RTT’s ultimate structure (ie, essential for this category of treatment). However, as seen in Hart,14 Dijkers,11 and colleagues, we anticipate that treatments whose targets are in the skilled performances and cognitive/ affective representations domains, whose mechanisms of action involve volitional learning, are unlikely to be subdivided in such a categorical manner. Indeed, we would argue that the only essential ingredients of these large categories of treatment may be akin to getting the recipient to perform and getting the recipient to acquire novel information. Although the performances and representations that lead to learning are not within the clinician’s control, the clinician’s selection of activities and information, and the methods of facilitating practice and information receipt, respectively, are the relevant ingredients. We anticipate that more detailed treatment specification in these categories may be achieved by identifying differences in the amount and type of active ingredients, such as instructional and motivational sets, repetition, feedback, and reinforcement. In contrast, we anticipate that the tissue- and organ-based treatments may subdivide in a more categorical manner. For these, something considered a broad class of essential ingredients at one level of the RTT may be subdivided into more specific essential ingredients at a finer level. For example, delivery of energy in the form of heat might be considered an essential ingredient for a set of interventions directed toward increasing tissue elasticity, but different methods of delivering energy in the form of heat (eg, hot packs, ultrasound, microwaves) might be considered essential ingredients to more specific treatment categories, all subcategories of the heat ingredients category.14 www.archives-pmr.org

S27 We use dosing parameters to refer to quantitative values of essential ingredients or other active ingredients that are hypothesized to moderate their impact. For example, if the essential ingredients of a particular form of progressive resistance exercise are repetitive muscle contraction against a challenging load, a clinician must decide on the number of repetitions and how to select the specific load and advance it. It is very likely that these decisions affect the efficacy of the treatment. Nevertheless, defining exercises that involve 8, 9, or 10 repetitions as completely different treatments seems unwarranted. Defining the number of repetitions as other active ingredients does not make sense either, because one cannot deliver 8 repetitions separately from muscle contractions. Further exploration of dosing parameters, however, is warranted. For example, there may be instances where a certain dose is insufficient to produce a treatment effect; therefore, a dosing threshold becomes an essential ingredient of treatment. In other instances, a certain form of dosing may activate new or different mechanisms of action, such that a new form of treatment must be defined. The interaction between active ingredients of treatment and their dosing is complex, and will likely require extensive empirical examination to be well understood.

Mechanism of Action The mechanism of action specifies how the active ingredients of a treatment produce the predicted change in the treatment target. A clear specification of the mechanism of action often lags behind a definition of the treatment’s ingredients because empirical research may confirm that certain ingredients reliably produce change in a target before the mechanism is understood. In fact, much research is dedicated to identifying and characterizing the mechanisms of action, even when the link between ingredient(s) and target is firmly established because knowledge of the mechanism of action may help discover new treatments, or assist in understanding which specific changes in quantity or timing of ingredients will bring about what changes in the target, over what time frame.

Identification of Theory Based on Observable Aspects of Treatment We aim to design an RTT in which it is possible to classify treatments by objective means, just as we can examine a plant and determine where it fits in the Linnaean taxonomy of living things. Classification based on observable or measurable characteristics of treatment would aid research and clinical practice. In research, we need to be able to verify adherence to distinct treatment categories when conducting clinical trials and to measure the amount of a particular treatment that has been delivered when conducting observational studies. In clinical practice and education, we need to be able to provide supervision and quality assurance that individual clinicians are delivering treatments appropriately. Although treatment theories are not themselves observable, they tell us what we should expect to observe about treatments of a specific kind. In principle, it should be possible to identify and classify a treatment through observation or other measurement of the specified active ingredients because it is only those treatment ingredients that can directly affect the treatment recipient. In some cases, observation of multiple treatment *sessions will allow relevant ingredients to be identified. Many treatments differ from

S28 one another in how they evolve over time, such as when weights are increased over time when the target is increased strength, but exercise duration is lengthened over time when the target is increased endurance. Such across-session forms of treatment progression are a common way of delivering the essential ingredients (to accomplish some form of continued challenge that is a presumed component of the mechanism of action) of many treatments, as subsequently discussed. A corollary of the objective observability of treatment ingredients is that treatments should be defined in relation with the behaviors of the rehabilitation practitioner, rather than the practitioner’s general intent or clinical goal, such as a goal to work on a certain problem area. As noted elsewhere in this supplement,10 other systems for classifying rehabilitation treatments have defined and named treatments based on the area of functioning that the practitioner believes to be addressed by the treatment (eg, gait training and memory remediation). Such treatment definition and naming tells us about the desired outcome of treatment (its target or aim) but typically gives little information about the specific contents of treatment. Furthermore, there may be cases in which the therapist’s clinical goal refers to the target of treatment (that which is being directly affected), and other cases in which it refers to distal aims of treatment that are hoped to be affected. Nevertheless, it can be useful to understand the clinician’s intent or treatment goal as a short cut to understanding what treatment theory is likely to be driving the treatment, what treatment attributes one should look for, and so forth.

Scope of Treatments to be Classified and Treatment Boundaries At the outset of this effort, we faced the need to determine the scope of treatments that we intended to address and to develop logical rules for determining the boundaries between individual treatments within that scope. The following sections describe conceptual decisions and challenges related to those issues.

What is a rehabilitation treatment? What is not a rehabilitation treatment? Rehabilitation treatments and services are those designed to change (increase or enhance) functioning of individuals or prevent decline in functioning in those at risk. This definition is extremely broad because functioning, as defined in the ICF,11 includes a range of outcomes from cellular structure to social participation. We believe that the principles we are defining can be applied to this full range. Nevertheless, in order to focus our work, we made a number of pragmatic decisions. We chose to focus only on treatments and services provided by specific clinicians to specific patients or clients. Thus, we have not included public health, policy, or environmental interventions designed to affect groups of individuals in society. For example, an elevator in a public building is not a treatment; an elevator or stair glide selected by a rehabilitation practitioner for the home of a specific patient is, however, a treatment designed to improve the patient’s functioning at home. Some rehabilitation treatments aimed at the body structure or function levels are essentially symptomatic treatments that differ very little from traditional medical treatments. For example, one can consider treatments for low back pain as medical treatments

J. Whyte et al aimed at a symptom (pain) of a disorder (mechanical low back pain), or as rehabilitation treatments aimed at an impairment (pain) that may limit activity (bending or lifting). Thus, it is impossible to draw a clear boundary between rehabilitation treatments and medical treatments. For practical reasons, we have focused our energies on typical medical rehabilitation interventions. We believe that the framework we have adopted is relevant to the full range of medical and psychiatric treatments, but we have not yet evaluated our conceptual decisions with regard to how well they accommodate these other treatment domains. There are many instances in rehabilitation where a treatment is initiated to enhance function, but some aspect of the intervention must be continued long term to maintain the functional gains. For example, providing a prosthesis and training in its use are followed by long-term use of the prosthesis; training in the use of a memory notebook or internal mnemonic strategy is followed by its long-term application in daily life. At some point, such activities become integrated with the individual’s everyday life and can no longer be isolated as treatments. In an effort to distinguish health-promoting behaviors occurring in everyday life from treatment, we will consider rehabilitation treatments to be interventions selected or designed by and overseen by a rehabilitation practitioner to increase function or prevent decrease in function of an identifiable patient. This is not to deny the importance of preventive or health maintenance activities performed independently by persons with disabilities, but merely to place our focus on situations that are clinical in nature, as indicated by the involvement of a rehabilitation practitioner.

Treatment versus assessment Assessment is an important part of the rehabilitation process, including assessment of current function, of the underlying factors and environmental influences that support and limit function, and of the potential to respond to treatments that might be selected by the clinician. Nevertheless, we do not consider it treatment. In principle, assessment can be distinguished from treatment by its purposedto clarify or understand versus to changedbut distinguishing assessment from treatment in actual rehabilitation practice is more challenging. Assessments performed by rehabilitation clinicians, in contrast with laboratory-based assessments, are often intertwined in space and time with treatment. Treatments may be continued, discontinued, or modified in response to the results of ongoing assessments, and specific forms of cuing or feedback may be added or subtracted during treatment to assess the source of functional difficulties. Moreover, the process of assessment itself undoubtedly has some behavioral effect in some learning-based treatments. For example, if behavioral performance by a patient is used to assess acquisition or retention of an information-based treatment (eg, how to take one’s medications), this episode of performance may provide some opportunity for additional learning. These points are more troublesome in practice than in principle. We can still define treatments with respect to their theoretically relevant ingredients, without regard to how they were selected for application. At the same time, we can acknowledge that in clinical practice, treatments may be started, stopped, and modified in response to interwoven assessment, in ways that might make them difficult to identify and isolate through observation. www.archives-pmr.org

Rehabilitation taxonomy conceptual issues

Who or what is being treated? In most cases, the recipient of treatment is the person with, or at risk of, a disability. However, in some instances, the treatment recipient is a caregiver, employer, or friend of this person. When a therapist trains a caregiver in how to safely transfer a mobilityimpaired patient, a learning-based treatment is delivered to the caregiver. To the extent that the caregiver acquires this skill and is available to apply it to the person with a disability, it has an indirect effect (via enablement/disablement theory) on the transfer activities of the person with the disability. Similarly, an employer might be taught useful ways of interacting with a new employee with a cognitive limitation. The treatment recipient, then, is the employer, but if the employer makes use of these new skills, there will be an indirect effect on the employment success of the individual with a disability. The target of treatment in both of these examples is acquisition of a skill (broadly defined) by a person without a disability, which is expected to affect the physical or social environment of the person with a disability, enhancing their day-to-day functioning. When the clinician manipulates the physical environment, however, rather than interacting with the person with a disability or a caregiver, we do not regard the environment as the recipient of treatment. Rather, we regard the recipient to be the person with a disability, who receives an intervention to lessen the effects of an impairment or activity limitation. This distinction between how we consider the patient’s physical and social environments reflects the reality that the clinician can directly alter the patient’s physical environment at will, but can only alter the social environment with the cooperation and behavior change of others in that environment.

One treatment or many: boundaries between treatments In clinical practice, it is common to direct multiple treatment ingredients toward the same overall clinical goal. For example, a clinician might discuss with a patient the importance of aerobic fitness in avoiding future health problems, provide instruction on specific aerobic exercise routines, and supervise a patient actually exercising. How, then, does one determine whether these are multiple essential ingredients of a single treatment versus multiple treatments used in combination? First, one must consider whether the multiple ingredients are truly directed toward the same treatment target. In the aforementioned example, providing the rationale for a set of exercises, and guiding the performance of the exercises, have different targets (a change in the knowledge of or attitude toward exercise vs a change in aerobic fitness). Thus, although it may often be useful to package these ingredients together for delivery to the same patient, they should be thought of as different treatments. In other cases, we may think of one treatment as paving the way for another. For example, a patient with protein deficiency may have a poor response to exercise; as a result, one might provide a protein supplement first. In another example, a patient with shoulder weakness may be unable to perform exercises designed to improve accuracy of reaching; as a result, one might place the arm in a balanced forearm orthosis in order to facilitate reaching exercises. In both of these cases, the 2 treatments have different targets (muscle metabolism vs strength; arm positioning vs reaching), and are, therefore, separate treatments that need only be given in combination to certain patients. www.archives-pmr.org

S29 There are cases, however, when multiple ingredients work synergistically and must be considered part of the same treatment. For example, vibration, when combined with muscle contraction against a load, may increase the load that can be lifted, and hence the strength that is gained.15 In this case, vibration has no treatment target when given alone, but it acts synergistically with exercise on the target of muscle strength. If this synergistic relation is relatively general across patients, then these ingredients would likely be given together to most patients. In these examples, attending to the target of treatment helps to clarify the boundary between treatments. Sometimes identifying the target or targets of treatment is itself difficult. Many *courses of treatment involve breaking an activity or skill down into component processes and gradually building toward a more complex routine. For example, one might guide a patient in practicing putting on underwear and then putting on a shirt, and so on. In another example, one might help a patient practice rolling in bed to the easy side and then to the difficult side. Assuming that the ingredients (specific instructional techniques) are the same and the mechanisms of action (learning processes) are the same, are they directed at one target (dressing or bed mobility) or different targets (donning underwear and donning socks; or turning to the left and turning to the right)? Because we have already established that the ingredients and mechanisms are the same, it is inconsequential whether we consider treatment X to have been applied to multiple subtasks one by one or to have been conceived as a larger-scale treatment from the outset. However, in the case where the clinician plans from the outset to build a larger behavioral routine, they may include additional ingredients, such as instruction in how to move from one subtask to the next, or may train in a strict order to emphasize their continuity. In such cases, the treatment target should be considered to be the larger behavioral routine, as revealed in the delivery of these additional ingredients. Mechanism of action is also important to consider when determining if a single treatment or multiple treatments are taking place. Ingredients that are directed to the same target are considered to belong to different treatments if they act on that target via different mechanisms of action. For example, one might provide education about the rationale for a certain behavior change. One might also prompt the recipient to engage in goal setting to encourage behavior change. If the theories of how factual knowledge versus personal goal setting resulting in behavior change specify different mechanisms of action for each of these ingredients, then these would be considered separate treatments. As noted previously, in the current state of knowledge it is not always possible to know when treatment ingredients operate by the same versus different mechanisms. Therefore, any RTT will inevitably change along with the state of the science.

Importance of the environment In the ICF and other disability conceptualizations, the environment plays a key role in interaction with the individual’s characteristics and in supporting or impeding activities and participation.13 Accordingly, clinicians often modify the environment by providing assistive devices, recommending changes to the physical environment, and seeking to influence the supportiveness of the social environment in the home and community. Because multiple aspects of the environment affect participation, are these services all treatments? When a clinician devises such treatments, is the target of treatment the environment?

S30 We noted previously the exclusion of global changes in the environment that are designed to serve the public in general rather than an individual with specific needs. We propose to identify the target of an individually tailored environmental modification as the broadest aspect of functioning that is restored by that modification. In the case of a stair glide installed in a person’s home, the target of treatment is moving upstairs and downstairs in the patient’s own home, not walking up and down stairs. The stair glide does not restore walking on stairs globally, it makes it possible for a person to move upstairs and downstairs in his/her home via an alternative method that does not require walking. Thus, the target of treatment for assistive devices is rarely the impairment that shaped their selection because most such devices act by reducing the effect of a given impairment on other domains of functioning rather than reducing the impairment itself. However, we believe it will still be useful to retain information about the impairment or activity limitation for which the device is compensating in order to understand the treatment’s mechanism of action. This may be accomplished through notation (eg, a different form of bracket than that described previously) that indicates the underlying impairment(s) for which the device or strategy is compensating. The aforementioned discussion addressed environmental modification treatments and the target of such treatments. The environment is also relevant in another waydin determining the breadth of generalization achieved in treatment. Consider 2 patients with gait dysfunction: one is largely housebound with dementia and the other wishes to participate in a range of community activities. With the first patient, one may set a treatment goal of safe ambulation within the home, whereas for the second, one would strive for safe ambulation in a range of community environments. One might conduct gait training only in the home for the first patient but conduct walking training in progressively more uneven and challenging environments for the second. Choices related to environmental variation, therefore, alter the active ingredients being delivered and demonstrate that different treatment targets are being pursued. Finally, the environment may play a role in motivation to engage in treatment. Consider a patient with tightness of soft tissues around the shoulder joint, which limits active shoulder range of motion. One might design stretching exercises intended to lengthen these soft tissues (the target of treatment). A plausible treatment theory would define the essential ingredients along the lines of repeated and/or prolonged tension on soft tissues of the shoulder. Any treatment that creates such tension would be effective, whether the stretching is done by the patient, the therapist, or a machine. There would be no reason to hypothesize that reaching to numbered lines on the wall should have any different effect than placing dishes on a high shelf in the occupational therapy kitchen if both create similar tension. However, a clinician might choose the latter approach to help the patient visualize his or her impairment more concretely and have a more intuitive sense of progress. Choices about the task materials and environment in which the treatment is conducted may contribute active ingredients that affect the degree of patient engagement in the task, without being essential (definitional) for treatment. The environment is an important factor in the enablement process and in the design and selection of rehabilitation treatments. We believe that the environment is important in qualitatively distinct ways and, therefore, should be incorporated into the RTT in more specific ways than simply acknowledging its generic importance for functioning. Environmental modifications may

J. Whyte et al encompass a variety of specific treatments that operate by different mechanisms and belong to different taxonomic groups.

Defining *volitional treatments and *nonvolitional treatments Treatments can be carried out on patients or by patients. We view this as an important distinction because the treatment theories that underlie treatments carried out on patients, nonvolitional treatments, rest on theories of the reactions of tissues or organs to external stimulation or manipulations of various kinds (eg, soft tissues that respond to prolonged stretch). Some nonvolitional treatments may require that the patient cooperate with the treatment or at least not actively resist it, whereas others may be carried out on patients even if they are unconscious. In contrast, treatments carried out by patients, volitional treatments, require some degree of effort to be effective, even if the target of treatment lies in the function of an organ system (eg, progressive resistance exercises). As a result, in the case of volitional treatments, a range of ingredients that pertains to learning and motivation becomes relevant as they influence whether and how the patient actually receives the treatment, and a range of cognitive and behavioral patient characteristics becomes relevant factors in treatment success. Defining volitional treatments poses particularly challenging issues. We earlier suggested that if ingredients affect different targets or work by different mechanisms of action, they should be defined as separate treatments, though we acknowledged that it may be common in clinical practice to deliver 2 treatments of this type together. By this strict definition, all volitional interventions would have to be subdivided into at least 2 treatments. To return to the example of progressive resistance exercise, one treatment target might be “increased knowledge of how to correctly perform X exercise protocol,” and the ingredients might focus on information about good exercise techniques, number of repetitions to complete, and so forth. Another might be “enhanced motivation to perform aerobic exercises as prescribed” and might include ingredients related to goal setting, encouragement to complete the target number of repetitions, and so forth. These treatment targets and ingredients essentially focus on progressive resistance exercise as an activity that the patient needs to learn and perform, not unlike treatments involving performance of activities of daily living, meal preparation, and so forth. However, the activity of progressive resistance exercise is, itself, a treatment (at least when directly supervised by a therapist), with the target of enhanced muscle strength. These 2 types of ingredients are potentially present in all volitional treatments. As a matter of convention, we emphasize the target-specific ingredients when the therapeutic activity is performed under supervision. In such instances, the clinician is typically most concerned with selecting the ingredients that will have the desired impact on the target and has the opportunity to convey more instructions and enhance motivation during the therapeutic activity if it is not being performed correctly or intensively enough. In contrast, when the therapeutic activity is instructed for performance at home or in the community, we emphasize the instructional and motivational ingredients because these are what the therapist delivers: there is no opportunity for modification based on the patient’s performance of the therapeutic activity if it is done without the therapist present. The specific therapeutic activity that is being taught is still relevant because the instructional and www.archives-pmr.org

Rehabilitation taxonomy conceptual issues motivational ingredients that are needed for success likely depend on the complexity and effortfulness of the therapeutic activity itself. Thus, we again propose to retain information about those contents of the treatment interaction that are informative but not pertinent to its classification; in this case, the therapeutic activity that has been taught or encouraged is “individual goal setting protocol X [home aerobic exercise protocol Y].”16 Further conceptual work is required, however, to address the question of whether volitional treatments should be automatically subdivided into 2 individual treatments, or whether we should view the instructional and motivational components as other active ingredients that are combined with the target-specific essential ingredients. It seems overly complex and impractical to split treatments in a way that is counterintuitive to most clinicians. Moreover, once we remove the instructional and motivational ingredients as separate treatments, what is left that meets the definition of treatment at all? In a certain sense, it is the patient who treats himself or herself by lifting the weights, repeating the learning trial, and so forth, once the patient knows how and is sufficiently motivated to do so.

Treatment progression Many rehabilitation treatments involve maintaining a level of challenge within the treatment task as clinical progress occurs: the position of a cast in serial casting is progressively extended over time as tissues stretch; the size of the item to be grasped may be reduced over time when attempting to remediate fine motor coordination deficits through practice. Such structured, systematic, and preplanned progression signals a defining ingredient of a treatment: the necessity to maintain a given level of challenge even as the target of treatment improves. For example, continuing to increase muscle strength requires continuing to lift a difficult (but not overwhelmingly difficult) weight, not just any weight; therefore, the weight must be increased over time to maintain that difficulty. The progression algorithm defines the creation and maintenance of the challenge. In order for progression to be meaningful as a description of a treatment, however, it must be systematic and rule-governed, and attributable to a priori choices made by the clinician rather than ad hoc responses to clinical improvement. For example, an algorithm that increases the weight to be lifted by 10% after the patient is able to complete 15 repetitions is an example of treatment progression. Providing less than expected physical support to an ambulating patient is not considered treatment progression because the clinician is prepared at all times to provide the maximum amount of support needed, but the actual provision of support is driven by the patient’s moment-to-moment needs. However, reducing the amount of support provided during practice of ambulation in a planned and systematic way, to elicit greater effort from the patient, is considered a form of treatment progression. Different forms of treatment progression can help us distinguish among similar-appearing treatments because they may reveal subtly different dimensions of challenge. Because treatment progression involves monitoring some aspect or aspects of the patient’s performance (rather than others), and advancing the treatment in specific directions (rather than others), treatments that are not distinguishable in a snapshot (eg, a single session) can be distinguished over time. For example, a particular physical exercise can be directed more toward strength or more toward www.archives-pmr.org

S31 endurance by progressing it in different ways, although the weight lifted in a single session might be the same.

RTT and Actual Treatments The reader will have noted that the conceptual issues laid out here are complex, although we would argue that they are only as complex as the domain they seek to systematize. When we began this process, we had anticipated that currently used treatments could be sorted into specific taxonomic categories.14 Moreover, we envisioned that certain treatment categories would correspond closely to particular domains of clinical intervention. That, in turn, would allow specific subgroups of clinicians and researchers to undertake the further refinement and subdividing of treatments in their area of expertise. In the course of developing the conceptual framework and initial steps of a theory-driven RTT, however, neither of these predictions has held true. As previously noted, we believe that many activities that clinicians currently conceptualize as specific treatments are, in fact, combinations of treatments (sequentially or simultaneously administered) when considered from the perspective of theory-driven treatment definitions. For example, the provision of an assistive device typically involves selection of a device with the appropriate properties to help compensate for a specific impairment or activity limitation, providing some information about how and when the device is used (to create a positive attitude toward the device, and understanding of its proper use); having the patient practice the donning/doffing, fitting, or programming of the device (a set of very specific skills); and practicing the selected daily activities using the device (yet another set of skills). Although all of these ingredients make sense with respect to the functional goal that the device is intended to solve, they operate through different mechanisms of action. There is, therefore, a benefit to conceptualizing them separately because different patient-specific decisions would be relevant to different treatment components. Certain physical impairments or capacities might be relevant to selection of the device itself; a patient’s language abilities and declarative memory might be relevant to the didactic instruction; and different forms of feedback during practice might be appropriate for different patients. Although it may still be useful to create treatment packages that specify the use of several different treatments that commonly occur together, we would argue that acknowledging the separateness of treatment components supports more effective algorithms for matching treatments to patients, and invites clinician reflection on whether all of the treatment components are appropriate (or even necessary) for all patients.

Conclusions We undertook this effort with the goal of beginning development of a theory-driven RTT. The commitment to a taxonomy based on treatment theories is important for guiding future research and for categorizing treatments with respect to their mechanisms of action. However, attempting to develop a theory-driven RTT has also exposed a number of complex and challenging conceptual issues. Some of these are specific to a theory-driven RTT and some would be faced by any effort to coherently categorize rehabilitation treatments. We hope that identifying and articulating these challenging issues can support ongoing efforts to develop a useful and comprehensive system for defining and classifying rehabilitation treatments.

S32

J. Whyte et al

Keywords Behavior and behavior mechanisms; Classification; Environment; Patient education as topic; Rehabilitation; Social environment; Therapeutics

7.

8.

Corresponding author John Whyte, MD, PhD, Moss Rehabilitation Research Institute, 50 Township Line Rd, Elkins Park, PA 19027. E-mail address: [email protected].

9.

10.

References 1. van Heugten C, Grego´rio GW, Wade D. Evidence-based cognitive rehabilitation after acquired brain injury: a systematic review of content of treatment. Neuropsychol Rehabil 2012;22:653-73. 2. Dijkers M, Kropp GC, Esper RM, Yavuzer G, Cullen N, Bakdalieh Y. Quality of intervention research reporting in medical rehabilitation journals. Am J Phys Med Rehabil 2002;81:21-33. 3. Johnston MV, Graves DE. Towards guidelines for evaluation of measures: an introduction with application to spinal cord injury. J Spinal Cord Med 2008;31:13-26. 4. Whyte J, Hart T. It’s more than a black box, it’s a Russian doll: defining rehabilitation treatments. Am J Phys Med Rehabil 2003;82: 639-52. 5. Wilde EA, Whiteneck GG, Bogner J, et al. Recommendations for the use of common outcome measures in traumatic brain injury research. Arch Phys Med Rehabil 2010;91:1650-60.e17. 6. Adelson PD, Pineda J, Bell MJ, et al. Common data elements for pediatric traumatic brain injury: recommendations from the working

11.

12.

13.

14.

15.

16.

group on demographics and clinical assessment. J Neurotrauma 2012; 29:639-53. Johnston MV, Keith RA, Hinderer SR. Measurement standards for interdisciplinary medical rehabilitation. Arch Phys Med Rehabil 1992; 73(12-S):S3-23. Bryce TN, Budh CN, Cardenas DD, et al. Pain after spinal cord injury: an evidence-based review for clinical practice and research. Report of the National Institute on Disability and Rehabilitation Spinal Cord Injury Measures meeting. J Spinal Cord Med 2007;30: 421-40. Wilde EA, Whiteneck GG, Bogner J, et al. Recommendations for the use of common outcome measures in traumatic brain injury. Arch Phys Med Rehabil 2010;91:1650-60. Dijkers MP. Rehabilitation treatment taxonomy: establishing common ground. Arch Phys Med Rehabil 2014;95(1 Suppl 1):S1-5. Dijkers MP, Hart T, Tsaousides T, Whyte J, Zanca JM. Treatment taxonomy for rehabilitation: past, present, and prospects. Arch Phys Med Rehabil 2014;95(1 Suppl 1):S6-16. Whyte J. Contributions of treatment theory and enablement theory to rehabilitation research and practice. Arch Phys Med Rehabil 2014; 95(1 Suppl 1):S17-23. World Health Organization. International Classification of Functioning, Disability and Health (ICF). Available at: http://www.who.int/ classification/icf. Accessed November 30, 2012. Hart T, Tsaousides T, Zanca JM. Toward a theory-driven classification of rehabilitation treatments. Arch Phys Med Rehabil 2014;95(1 Suppl 1): S33-44. Rees SS, Murphy AJ, Watsford ML. Effects of whole-body vibration exercise on lower-extremity muscle strength and power in an older population: a randomized clinical trial. Phys Ther 2008;88: 462-70. Dijkers M, Hart T, Whyte J, Zanca JM, Packel A, Tsaousides T. Rehabilitation treatment taxonomy: implications and continuations. Arch Phys Med Rehabil 2014;95(1 Suppl 1):S45-54.

www.archives-pmr.org

Rehabilitation taxonomy conceptual issues

Supplemental Appendix S1 Glossary of Terms Active ingredients e Attributes of a treatment, selected or delivered by the clinician, that play a role in the treatment’s effects on the target of treatment. These include the essential ingredients associated with the treatment’s known or hypothesized mechanism of action, and any other ingredients that moderate the treatment’s effect(s) but may be common to multiple treatments. Aim(s) (of treatment) e Aspect(s) of the patient’s or other recipient’s functioning or personal factors that is predicted to change indirectly (via mechanisms specified in enablement/disablement theory) as a result of the treatment-induced change in the treatment target. A single treatment may have multiple aims, and there may be a chain of treatment aims (eg, exercises leading to increased strength leading to improved ambulation leading to greater community participation) with at least 1 (in the case of treatments delivered to other recipients) involving the patient’s functioning. Although highly relevant to the ultimate clinical value of a treatment, these distal treatment aims are not relevant to the definition or classification of the treatment. (As previously noted, when the target of treatment is clinically and functionally significant in its own right, we avoid calling it a treatment aim to avoid confusion about direct vs indirect effects of treatment.) Course of treatment e Series of treatment sessions or therapistrecipient contacts that are pursuing change in a specific treatment target, sometimes with some form of treatment progression within or between contacts. Dosing parameters e Quantitative variations in the strength, intensity, frequency, and/or quantity of treatment ingredients; these are often expressed as an amount of time during which the recipient is continuously exposed to the ingredient, the number of times a discrete ingredient is administered, or the magnitude of the ingredient on a quantifiable scale. Enablement/disablement theory e Conceptual system that specifies how change in one aspect of a patient’s functioning (eg, at the level of an International Classification of Functioning, Disability and Health component: body structure, body functioning, activity/activity limitation, participation/participation restriction, personal factor, or environment) will translate into changes in another aspect, specifically a characteristic classified elsewhere in the framework being used. Essential ingredients (of a treatment) e Active ingredients, selected or delivered by the clinician, that define a particular treatment and distinguish it from other treatments. The essential ingredients are those that are specified by the corresponding treatment theory, and are hypothesized or known to be necessary for the treatment’s effects on the treatment target. Inactive ingredients (of a treatment) e Attributes of a treatment that do not define or moderate the impact of the treatment on the target. Ingredients may be presumed to be inactive because they are not addressed by a treatment theory (eg, the building in which the treatment is conducted) or have been empirically determined to be inactive. Ingredients e See Treatment ingredients Mechanism of action e Process by which the treatment’s essential ingredients induce change in the target of treatment. A treatment theory should specify how the essential ingredients

www.archives-pmr.org

S32.e1 engage mechanisms of action to bring about desired treatment effects, that is, specification of the mechanism of action explains how the essential ingredients alter the treatment target within the framework of the treatment theory. Similarly, additional mechanisms of action specify how other active ingredients moderate the effects of the treatment. Nonvolitional treatments e Treatments whose hypothesized mechanisms of action require no effort on the part of the recipient (other than cooperation/nonresistance). Unlike volitional treatments (subsequently defined), the recipient of nonvolitional treatment is always the patient/client undergoing rehabilitation, not a third party (eg, caregiver). Patient/client e Person with a disability or at risk of a disability who is the intended beneficiary of treatment. Progression e Clinician’s deliberate, systematic alteration of the treatment to maintain, over time, the degree of challenge to the body system/behavior(s) selected for change. Progression is often triggered by improvements in the target of treatment; therefore, the pace of progression (within a single treatment contact or a course of treatment) typically depends on the pace of change in the treatment target. The form that treatment progression takes (and hence the nature of the challenge that is being maintained) is often specified by the treatment theory; therefore, multiple treatment sessions may need to be observed to distinguish between one treatment and another. Recipient (of treatment) e Individual whose function/behavior is intended to be changed directly as a result of treatment. In most cases this is the person with a disability (patient/client recipient), but in some instances a caregiver or employer may be the other recipient who is changed by the intervention (eg, to provide care or intervention to the patient/client or to create a more supportive environment for the patient/client). Rehabilitation treatment taxonomy e System of classifying rehabilitation treatments based on a principle or set of principles that allows for distinctions between treatments, which have practical and/or theoretical utility. Session (of treatment) e Individual episode of treatment (typically minutes to a few hours), which may be repeated during a course of treatment. Target of treatment/treatment target e Aspect of the recipient’s functioning, or personal factor, that is predicted to be directly changed by the treatment’s mechanism of action. Specification of the target of treatment in a theory in terms of an International Classification of Functioning, Disability and Health variable(s) helps to define the scope of the treatment/treatment theory. (The aim of treatment refers to changes in functioning obtained in indirect waysdsee Aim(s) (of treatment), above. Although there are instances where the target of treatment is functionally important in its own right, without reference to distal enablement effects, we nevertheless reserve the term treatment aim only for the functionally relevant clinical effects that are distal to the treatment target.) Taxonomy e System of classification or categorization based on characteristics that have important pragmatic or theoretical implications. Treatment e Action taken by a health professional, in the context of contact with a treatment recipient, to alter the

S32.e2 functioning of an individual with a disability or at risk of a disability. Treatment is defined broadly to include provision of information, devices, and referrals, specific active experiences, and passive interventions.

J. Whyte et al ingredients, Dosing parameters, Essential ingredients, and Inactive ingredients.

Treatment grouping e Broad class of treatments that is similar in essential ingredients (eg, forms of energy) and is able to act on a class of similar treatment targets (eg, tissue properties).

Treatment theory e Conceptual system that predicts the effects of specific forms of treatment on their targets, specifying the law(s) that expresses the relations between essential ingredients and treatment target changes. (This is similar to Mechanism of action but may be broader and more inclusive.)

Treatment ingredients e Observable (and, therefore, in principle, measurable) actions, chemicals, devices, or forms of energy that are selected or delivered by the clinician. See also Active

Volitional treatments e Treatments where a hypothesized mechanism(s) of action requires some effort on the part of the treatment recipient.

www.archives-pmr.org