The Hypothesis-Oriented Algorithm for Clinicians II (HOAC II): a guide for patient management. (Perspective).

In 1986, Rothstein and Echternach (1) published a clinical decision and documentation guide called the Hypothesis-Oriented Algorithm for Clinicians (HOAC), which they contended offered clinicians a pragmatic, scientifically credible approach to patient management. Since that algorithm was first published, radical changes have occurred in the health care system. For example, there is now widespread discussion of the importance of physical therapists making diagnoses, (2) and there is also general acceptance of the need to view patients and clients within the context of one of the disability models. (3,4) In addition, therapists often have to relate to practice guides and guidelines. (5) We argue that what is needed is a patient management system that involves the patient in decision making and can be used to provide payers with better justifications for interventions, including occasions when therapists may disagree with practice guidelines. Compatibility with the Guide to Physical Therapist Practice's (Guide's) patient management model, including the formulation of diagnoses, is also desirable. (6)

The purpose of this article is to present HOAC II, a revised algorithm designed to meet the needs of contemporary practice. The algorithm, we believe, is compatible with the American Physical Therapy Association's (APTA's) Guide to Physical Therapist Practice, (6) including the therapists' need to diagnose and to offer interventions designed to prevent problems. In the context of the HOAC II, a problem is almost always a functional deficit. Although we attempted to be consistent with Guide terms, there are instances where we used alternate terms for the sake of clarity.

Although the original HOAC was a first effort at bringing scientific decision making into a user-friendly practical context for clinical decision making, it has some cumbersome elements as well some logical and procedural flaws. The algorithm offered no guidance on how to determine when an intervention designed primarily for prevention was appropriate and how risk factors could be eliminated. The algorithm also did not adequately provide a means for identifying problems and addressing goals noted by someone other than the patient.

The focus on patient-centered outcomes was, however, an innovation in HOAC and laid a foundation for the implementation of the HOAC in clinical decision making in the context of currently used disability models. The disablement model that we believe currently offers the greatest utility for clinical practice is the Nagi model. (7(pp223-241)) A common element in both the old and new versions of the HOAC is that therapists using the terms of the Nagi model are called upon to identify impairments, when appropriate; to examine how these impairments relate to functional deficits; and to examine whether interventions designed to ameliorate or reduce impairments result in changes in function and changes in levels of disability. In some cases, therapists also can hypothesize that factors other than impairments may lead to functional loss. For example, a societal limitation such as high curbs may contribute to a patient's inability to walk to school. We also believe therapists have a role in prevention (7(pp84-89) and that in a responsibility-focused health care system clinicians should identify the hypotheses that underlie interventions used for prevention.

We believed that the original HOAC could serve both as a template for documentation and as a conceptual model for decision making and, therefore, could link documentation and practice. This does not mean, however, that we believe either the original HOAC or the HOAC II must be implemented in the exact form we have written it, for all patients, in all settings. Rather, We contend that elements can be selected based on practicality and the expected benefit of using a system in which all elements of patient management are explicitly detailed. The HOAC II, we contend, provides a means for not only using evidence in decision making, but also for documenting the nature and extent of evidence used. Within the new version, elements related to justification of interventions (eg, where evidence can be cited) should be part of any credible system of documentation.

Overview of Elements in HOAC II

In developing the revised algorithm, we recognized that there are usually 2 major types of patient problems: (1) those that exist when the patient is being seen and that require remediation and (2) those that may occur in the future and that require prevention. We also realized that even though clinicians do not necessarily routinely discuss these differences, clinical management of the 2 types of problems is different, and assessment of the outcomes for each must differ.

While there are 2 types of problems (existing and anticipated), there are also 2 ways that problems are identified. There are patient-identified problems (PIPs) and non-patient-identified problems (NPIPs). Patient-identified problems, which usually consist of functional limitations and disabilities, often exist when the therapist sees the patient. The patient identifies the problem. The therapist, however, needs to generate hypotheses as to the cause of problems and to establish testing criteria, which can be used to evaluate the outcomes of interventions, and the correctness of the hypothesis and patient care strategies. The patient may identify existing problems as well as express concerns relating to problems that do not yet exist and could, therefore, be the source of an anticipated problem. For example, a patient may complain of shoulder pain and express a concern about the development of limitations in movements that could be disabling. The limitations in function caused by the pain would be an existing problem (eg, an inability to cook a meal because repetitive use of the shoulder caused intolerable pain). Any loss of function that could occur if motion became even more limited would be an anticipated problem.

Non-patient-identified problems are problems that are not identified by the patient. They are problems that may occur as well as existing problems. For example, children may not be able to identify problems secondary to central nervous system deficits. A child might, for example, routinely sit in a position that compromises his or her ability to breath because of decreased thoracic excursion. The child is unlikely to see this as a problem, but a family member or a member of the health care team could believe that a problem (NPIP) will develop. In this case, either the therapist or the caregiver will be the most likely person to identify the problems. Similarly, patients who have had a stroke may have difficulty communicating about their problems, and others will need to identify these problems. Justification for anticipated problems, regardless of whether they are PIPs or NPIPs, can, in the HOAC II, only be based on theory or arguments that are data (evidence) based. Hypotheses that guide intervention to eliminate existing problems (PIPs or NPIPs) can be tested because a change in something can be measured (eg, changes in impairment levels and disability). Changes in what is measured will be identified in the part of the algorithm where reassessment occurs.

A problem is kept from occurring when anticipated problems are correctly managed. Therefore, no observable change usually relates directly to the problem. More importantly, in the absence of' anything observable or measurable, a justification based on an outcome is not possible for interventions aimed at prevention, because even without intervention a problem may not have arisen.

Testing criteria are used to examine the correctness of hypotheses related to problems that currently exist. For NPIPs or PIPs that are anticipated, however, the therapist establishes predictive criteria, which, if met, indicate that problems will most likely be avoided because risk factors were reduced or eliminated. A predictive criterion for a patient with low back pain, for example, may be that a patient is considered no longer at risk when the patient can perform stretching exercises at a suitable level of performance on a regular basis (the predictive criteria would detail the specific exercise and how often it should be performed).

To justify any predictive criterion, the therapist should base the criterion on best available evidence. Patients with spinal cord injuries, for example, might no longer be considered at risk (ie, they have achieved the predictive criteria) for developing skin ulcers when they have shown that: (1) they will spontaneously do wheelchair pushups a given number of times per hour, and (2) they will monitor the status of their skin by having someone check for red marks or abrasions at specified intervals. In each case, the predictive criteria relate to an observable behavior, not just increased awareness or knowledge. The behavior ideally is justified based on identified evidence or sound theory and not just on assumptions. Circumstances may make it impossible to achieve goals with observed behaviors, and in these special circumstances knowledge may be a reasonable predictive criterion (eg, when the therapist cannot visit the patient's workplace but teaches the patient strategies for avoiding injury).

The dual problem lists, one for PIPs and one for NPIPs, are merged into a single problem list as one proceeds through the algorithm. Throughout the rest of the algorithm, the source that identified the problem is not a concern. What is critical, however, is that therapists manage anticipated and existing problems differently, and the algorithm provides parallel paths for the management of the 2 types of problems. This is particularly important in the reassessment phase (Part 2 of the algorithm). The existence of a list of anticipated problems and the predictive criteria allows for the identification of which interventions are designed for prevention, how these interventions are justified, and when intervention can be stopped. A novel element of the HOAC II, one that we believe has not previously been seen in physical therapy literature, are mechanisms to make interventions designed for prevention goal oriented and of determinate duration (ie, there is a stated goal that must be achieved and there is an expectation as to how long this will take).

The algorithm provides clinicians with a mechanism for planning and evaluating activities designed for prevention. This approach encourages therapists to work to minimize risks through prevention, but, more importantly, it allows them to evaluate their efforts and to describe and justify their efforts to one another, payers, managers, and others. Because in HOAC II prevention activities are goal driven and are planned for specified periods, of time, therapists can, through use of the algorithm, identify to payers the resources they will need to achieve prevention. This should, in our opinion, assure payers that interventions will not continue indefinitely, unless that can be justified before the initiation of the intervention. The algorithm also allows the therapists to document when, in their professional opinions, prevention activities are needed and the consequences of what will occur if these are not carried out (eg, due to a lack of patient adherence or because they are not authorized by payers).

In a continued effort to keep focus on what are truly the patient's goals, one problem list (the PIP list) is generated before the examination. In the HOAC II, there is a record, at least initially, of who identified the problem. A complete problem list, however, including problems identified by the therapist and others, and a complete set of goals are not generated until later in the process. Figures 1 and 2 illustrate Part 1 of the HOAC II.

[FIGURE 1-2 OMITTED]

In Part 2 of the HOAC II, there are 2 reassessment paths, one for existing problems and one for anticipated problems (Figs. 3 and 4). In each case, there are questions on a flow diagram that direct therapists through relatively simple steps that are taken in response to questions. Two flow diagrams are used to describe the reassessment, one for existing problems and one for anticipated problems. A list of commonly used terms operationally defined for the HOAC II is provided in the Appendix.

[FIGURE 3-4 OMITTED]

Using the Algorithm

Part 1 of the algorithm deals with all 5 elements of the patient/client management model described in the APTA's Guide to Physical Therapist Practice (6) (examination, evaluation, diagnosis, prognosis, and intervention). The Guide is not specific about issues related to the use of an evaluative strategy to modify interventions and to test hypotheses. In the Guide, however, under "Intervention," it is stated: "Decisions about intervention are contingent on the timely monitoring of patient/client response and the progress made toward achieving the anticipated goals and expected outcomes." (6(p46)) We believe, therefore, that Part 2 of our algorithm is an elaboration on one vital element of what the Guide refers to as "intervention." In the HOAC II, issues related to monitoring intervention effects and altering the plan of care are covered in Part 2.

Part 1

Collect Initial Data (includes the History)

Early in an episode of care, clinicians start to obtain information that they will use to guide all elements of patient management. Practitioners appear to approach each patient with a set of hypotheses and collect data to confirm or refute those hypotheses (8,9); therefore, even initial data collection is hypothesis driven. During the interview, for example, questions about activities that may have caused an injury are one sign that the clinician is seeking to confirm or deny hypotheses. More experienced clinicians can be expected to generate hypotheses earlier than less experienced practitioners (9) and, in our experience, more effective clinicians often feel a greater freedom to discard hypotheses and consider alternatives as early as the interview phase of the patient examination.

The algorithm does not specify the type and scope of information gathered during the initial data collection phase. This remains the choice of clinicians, depending on their approach to practice. The algorithm simply requires clinicians to note what they do in this process. Information that will be used to create a PIPs list needs to be obtained during the initial data collection.

Patients seeking physical therapy have expectations of what therapy should offer them, and these may differ from what their therapists feels are reasonable. A patient may believe that walking without an assistive device should be the goal, for example, whereas the therapist may contend that this would be impossible and walking with a device would be a reasonable goal. Among the essential data that clinicians must collect are clear nonmedical descriptions of expectations, particularly descriptions of those disabilities and functional limitations that need to be eliminated. Incongruence among expectations of the patient, a payer, a referral source, and a therapist will be considered when problem lists are generated, but information about the differing expectations needs to be obtained early in the process.

Generate a PIPs List

Generating the PIPs list is one of the easiest things to do in the HOAC II. It requires clinicians simply to record patients' reports of the problems that led them to seek physical therapy (or the medical care that led to a referral for physical therapy). Therapists ask patients about what they can and cannot do (ie, what limitations they have in function). A functional limitation or disability is a problem, and, in some cases, patients also may express concerns that they have a condition that could lead to the development of loss of function in the future. In this way, a patient could be the one who identifies an anticipated problem. The therapist, however, with consultation with the patient, needs to determine whether the patient's concern is realistic and, if the anticipated problem is justified, add it to the problem list.

Because the HOAC II emphasizes accountability, we believe therapists should never assume that any patient concern about the future means that an anticipated problem will occur. Only when the therapist can supply evidence or a sound theoretical argument to support the possibility of the anticipated problem occurring should it be placed on the list, which is true regardless of the source. Evidence is preferred over theory when the therapist believes that the patient's concern about future events is not warranted. The therapist needs to discuss the reasons with the patient and, to enhance accountability, document that the discussion took place (ie, if the patient's concern was not added to the problem list, explain why).

Formulate Examination Strategy

Based on the initial data collected and the nature of the PIPs, the therapist needs to determine what other information is needed. This is an examination strategy, and it cannot exist independently of hypotheses. When generating the examination strategy, the therapist is not yet able to identify a best hypothesis as to the cause of the patients' problems (both PIPS and NPIPS). The therapist may have several competing hypotheses and needs to develop an examination strategy that will obtain information to confirm correct hypotheses and negate nonviable hypotheses. Unless a therapist has some tentative ideas (hypotheses) as to what may be causing the problems (eg, the potential impairments or pathologies causing functional limitations or disabilities), there can be no examination strategy.

Experienced clinicians appear to generate hypotheses more readily than less expert clinicians, and they also appear to be better able to identify sources of data needed for hypothesis testing. (10) We believe many new therapists, like many new physicians, often conduct a plethora of tests because: (1) they have been taught methods of patient management that require suspension of hypothesis generation until all the data are collected, or (2) they do not have enough experience to generate a tentative idea (hypothesis) on which to base a focused examination strategy. We recognize that, for some patients, therapists may be unable to generate examination strategies, and the algorithm calls for consultation when this occurs and provides a mechanism for documenting and justifying the use of a consultant.

When using the HOAC II as a guide to documentation, therapists must describe their examination strategies, including how they arrived at these strategies (based on available data) and why they believe the chosen examination techniques will lead to information that can be used to confirm or deny hypotheses. This may appear to require a lot of information. Notes in the patient's medical record, however, may be as simple as "the patient's inability to walk down stairs may be due to balance problems. Testing of balance appears to be most important, and tests of muscle force and range of motion will be conducted to rule out less likely causes of the functional limitation." In this example, the balance testing directly addresses the hypothesis, whereas muscle force measurements and range-of-motion measurements could lead to rejection of the hypothesis. The important element is that a link exists between the logic that guides the examination strategy, the information available, and the therapists' hypothesis. This does not require elaborate documentation on the part of the therapist.

Conduct the Examination and Analyze the Data

Examination procedures for a given type of patient may be governed by departmental policies, critical paths, or a variety of other influences. Ideally, approaches should be data driven (evidence based) and based on research suggesting best methods of examination and data analysis. (11) The HOAC II does not specify how or what to examine, but, for the process to be useful, the examination must follow logically from the examination strategy and not include extraneous procedures if they are not part of the examination strategy. That is, examination procedures should be related to the tentative hypotheses, either to confirm or to reject those hypotheses. The measurements obtained during this phase should be of the type and quality specified by the APTA's Standards for Tests and Measurements in Physical Therapy Practice. (12)

For documentation, all descriptions and analysis of the data obtained during the examination should be clear. Reasons why hypotheses were supported or rejected need to be specified, and, when findings call for additional examination procedures, this should also be described.

Add NPIPs to the Problem List

Just as existing and anticipated problems are on the PIPs list, existing and anticipated problems are on the NPIPs list. The anticipated problems require special consideration because they involve prevention, whereas the current problems are those, including functional limitations and disabilities, that were not initially described by the patient. NPIPS may be identified as early as the initial data collection phase, but they do not formally appear in the HOAC II until after the examination, when the NPIPs list is completed.

Sometimes, particularly with children or those with communication disorders, caregivers or family members may describe current problems. In this case, the problems may be described in the initial data collection phase. These problems are placed on the NPIPs list, but, in the context of the HOAC II, will be managed in a similar way to the other existing problems on the PIPs list. The problems are not different in nature, but only in the source of identification. The therapist, however, is responsible for the management of the problems on the NPIPs list regardless of the source that identified the problem.

Anticipated problems are different than existing problems, and, in the HOAC II, management of anticipated problems is a central feature. Following a transtibial amputation, for example, a person is likely to develop a knee-flexion contracture. (13) The therapist is likely to know this, and the patient is not likely to know this. The therapist also will know that if a contracture develops, the patient may be unable to use a prosthesis and may lose function.

Identification of anticipated problems often requires therapists to consider anticipated impairments to prevent functional limitations and disability, but anticipated problems may also be pathologies. A therapist, for example, should be aware that returning a patient with a compromised cardiovascular system to full activity without the patient being able to monitor his or her own vital signs could cause a cerebrovascular accident (CVA) or myocardial infarction. Here the anticipated problems are pathologies that could be prevented by teaching the patient how to monitor his or her cardiovascular status. The patient also may be the source for anticipated problems, and, although these are in the PIPs list, they are managed in a way that is similar, in the context of the HOAC II, to the way all other anticipated problems are managed.

Anticipated problems are usually risk factors, for future pathologies, impairments, functional limitations, and disabilities. The problem is the risk factor, and the intervention will be aimed at eliminating the risk factors. Sometimes an exacerbated risk factor may be contributing to functional limitation or disability. In a person with low back pain, for example, inappropriate lifting techniques may be the reason for an existing problem (ie, activities cause pain, which limits function), but continued use of poor techniques following the current episode could lead to recurrence. Poor lifting techniques may be a cause of an existing problem, and this needs to be addressed when the therapist generates hypotheses regarding the causes of existing problems. The poor lifting techniques also could be the cause of an anticipated problem because they put the patient at risk for future disability.

Justification for Hypotheses

The therapist makes 2 types of justification based on the nature of the problem (existing or anticipated) and chooses one of 2 paths in the algorithm. Existing problems require one type of argument, that is, hypotheses about the diagnosis that detail what needs to be changed to eliminate existing problems. Anticipated problems require a different kind of justification for the elimination of risk factors and a case as to what may happen without intervention. Both types of justification should be evidence based to the extent possible.

Generate a Hypothesis (or Hypotheses) as to Why the Problems Exist

Each existing patient problem has an underlying cause or causes. In the HOAC II, the cause is usually due to an impairment that is present, but in some cases the cause could also relate to pathology, functional limitations, societal limitations, or disabilities. Interventions, we believe, need to be focused on eliminating causes of problems. However, unless clinicians state, during clinical problem solving and documentation, why they believe problems exist, it is often difficult to justify interventions or to see how they relate to problems. Students, for example, often find it difficult to see how their clinical instructors determined what intervention to use. Similarly, payers may not be able to discern why a therapist focuses on an isolated motor skill instead of a functional task during intervention unless the therapist hypothesizes that a relationship exists between the isolated skill and functional activities. The hypotheses generated during this step provide the link between the therapist's diagnosis and the intervention. No intervention for an existing problem should be conducted unless it relates to the hypothesized cause of a problem.

Often the causes of disabilities will be the presence of pathologies, impairments, and functional limitations. Many physical therapy interventions focus on impairments and functional limitations, and therefore most diagnostic hypotheses will be directed at the impairment and functional limitation dimensions. Sometimes, however, therapists will attempt to eliminate a pathology. When this occurs, the pathology is the hypothesized cause. When a therapist believes that a wound fails to close because of infection, for example, the hypothesis could be at the level of a pathology; that is, unless the sepsis is eliminated, the wound will not close. This would be a testable hypothesis because wound cultures could be requested. Hypotheses that identify suspected pathologies often cannot be tested by physical therapists because most therapists are unable to request invasive tests or radiological diagnostic tests. Therapists usually need to consult with and possibly refer patients to a physician to determine when a pathology is identified as the diagnosis in a hypothesis. (10)

The HOAC II, like the original algorithm, places an emphasis on hypothesis generation and requires the therapist not only to determine what may be causing the problem (eg, loss of muscle force, loss of motion), but to also postulate as to the magnitude of the deficits (eg, how much weakness a patient has and how much force would be needed for the problem to be eliminated). The amount of force needed will serve as the testing criteria for the hypothesis. Therefore, when generating hypotheses, therapists must understand that in a subsequent step they must quantify what must be achieved to eliminate the problem. One way of determining whether a hypothesis is appropriate is to consider whether such testing criteria could be generated. In the wound example, the criteria would be a report of a negative culture. This example demonstrates that even when the hypothesis is at the level of pathology, generation of testing criteria must be possible.

Hypotheses that identify impairments as the cause of disabilities and functional losses are even easier to generate. If a person cannot walk following a CVA, for example, it would be incorrect in the HOAC II to hypothesize that the cause is damage to the motor cortex. Although this may be true, the quantification of the type and extent of pathology is not observable and measurable by physical therapists. The diagnostic hypothesis may be that the person cannot walk because he or she lacks the ability to generate sufficient quadriceps femoris muscle force during stance. In this example, the problem is a functional deficit, and the hypothesis relates the functional deficit to an impairment. The testing criteria will be the amount of force the therapist believes the patient needs to be able to generate to eliminate the problem (ie, to walk). Had the hypothesis identified a pathology (damage to the motor cortex), the pathology could not be measured by physical therapists, and, more importantly, the intervention is not designed to change the pathology, but rather the impairment and disability that the pathology caused. In addition, the pathology (as measured with magnetic resonance imaging, for example) would likely be unchanged, even though the intervention successfully dealt with the impairment or functional limitation.

The critical elements of hypotheses are that they deal with elements that would be affected by the intervention and that they must be sufficiently clear to allow for the generation of testing criteria. The testing criteria that therapists generate must represent pathology, impairments, or functional loss that can be measured in clinical practice. As discussed earlier, when a previously undiagnosed pathology is hypothesized to be present, consultation with or referral to a physician may be required to confirm the hypothesis. A problem may have more than one underlying cause, and, in these cases, the therapist may generate multiple hypotheses. The therapist also would generate testing criteria for each hypothesis. This might occur, for example, when weakness and a lack of coordination are hypothesized to be the reasons why a person can no longer ambulate independently.

For Each Anticipated Problem, Identify the Rationale for Believing Anticipated Problems Are Likely to Occur Unless Intervention Is Provided

Physical therapists, like many other health care professionals, share beliefs about what is happening and what may happen to their patients. Some of these beliefs are based on data that identify risk factors, factors that once eliminated should reduce the possibility of future negative health outcomes. The Framingham study, for example, identified many risk factors for cardiovascular disease. (14) Epidemiological studies of this type are usually the means for justifying interventions designed to eliminate risk factors. Unfortunately, data often are lacking for beliefs that health care professionals have about risk factors.

On what data do physical therapists act? The question is a legitimate patient management and resource allocation query. Without evidence to support the value of elimination of risk factors, the possibility of excessive intervention exists. Too little intervention for risk factors also is a possibility. The HOAC II provides a mechanism for therapists to use either epidemiological data or theoretical constructs to justify interventions aimed at reducing risk factors. The former is data based or evidence based, whereas the latter uses argumentation and logic that should have some scientific basis. Evidence-based arguments are preferred. (15)

By using the algorithm, justification is explicit rather than implicit and can be discussed by all relevant parties. We contend that unless physical therapists can provide either good presumptive arguments or data to support interventions aimed at reducing or eliminating risk factors, the role of therapists in prevention will be increasingly challenged and possibly eliminated. When therapists provide evidence to support their decisions, however, little reason exists to deny interventions if the risk-benefit ratio is reasonable. When therapists can only provide arguments, the case is less clear.

When using the HOAC II, we believe therapists must discuss all anticipated problems in the documentation and provide arguments and evidence as to why they believe that a problem would occur without intervention. This applies to justification for all interventions related to anticipated problems and diminishes the likelihood of unnecessary interventions or of interventions continuing after they are no longer necessary. We believe this documentation will not only enhance patient care but also will make our interventions more credible, particularly those aimed at prevention.

Refine Problem List

The problem list at this point in the algorithm contains 2 types of problems (existing and anticipated) derived from 2 sources (the patient and all other sources). The therapist needs to determine whether the problems can be addressed by physical therapy interventions. If the patient needs the intervention of another health care practitioner, the therapist needs to make a referral and document why the referral is necessary. If the therapist believes that the problem cannot be addressed, such as when no intervention would help, the therapist needs to discuss this with the patient and: (1) remove the problem from the list of problems to be addressed by physical therapy, (2) document why the problem could not be eliminated, and (3) document the discussion that took place and describe what was agreed on with the patient. The therapist may believe that some problems can only be modified and not be fully eliminated. Again, the therapist should make this modification in the problem list, discuss it with the patient, and document the nature of the discussion.

For Each Problem, Establish One or More Goals

Existing problems. In the HOAC II, like the original algorithm, there is one type of goal--something that the patient needs to achieve. Goals are almost exclusively expressed in terms of functional activities that the patient wants or needs to perform. Often therapists and others have used the term "short-term goal" not only to indicate something that can be achieved in less time than long-term goals, but also to indicate changes in levels of impairments they believe are related to long-term goals. A therapist might say, for example, that a short-term goal is teaching the patient an exercise to strengthen a paretic limb. Strengthening would relate to a long-term goal in which that limb might be used for ambulation. We believe this approach is confusing. Changes in the force-generating capacities of muscles may indeed help some patients to achieve functional activities, but the goal is the function--strengthening may or may not be a means to that end. We contend that if all a therapist achieves is increased force capacity, the patient has gained little or nothing from therapy. To consider a change solely in an impairment as meeting a goal, in our opinion, is almost always inappropriate.

In the HOAC II, impairment changes are monitored through the testing criteria and usually are not goals. All of the goals used in the HOAC II must represent meaningful accomplishments. (16) That is, meeting a goal as written in the algorithm means the patient's function has changed meaningfully. Some functions may be recovered sooner than others, and these can be identified as short-term goals. The overriding issue is that long-and short-term goals represent the same kind of phenomenon (meaningful change for the patient) and the only difference is the time needed to achieve them. The simplest way of checking whether a goal is really meaningful in the HOAC II context is to consider: (1) whether anyone would feel therapy was worthwhile if this is all that is achieved and (2) whether the payer would find therapy to be worthwhile if this is all that is achieved.

Many patients, such as people with CVAs, may, in theory, have many problems, and they might have a rather long list of goals they want to achieve to perform activities of daily living (ADL), instrumental activities of daily living (IADL), and other activities. Goal lists for such patients might seem almost infinite in scope and impractical in length. For patients such as these, the therapist needs to work with the patient to identify those goals that are most important and those that are indicative of various levels of difficulty. A therapist may list as a goal "independence in brushing teeth," for example, and use this to represent a variety of similar tasks requiring eye-hand coordination, such as using utensils for eating. In this way, not all goals have to be listed, but rather there should be those that are especially important to the patient and those that represent a hierarchy and diversity of motor skills that could serve as goals.

Anticipated problems. Therapists and patients need to work together to eliminate existing problems to achieve goals that they have delineated together. The goal for an anticipated problem is to prevent the problem from occurring.

For Each Existing Problem, Establish Testing Criteria

In the HOAC, the word "hypothesis" is used because it has a mechanism for therapists to test whether their ideas about causes of problems (ie, their diagnoses) may be correct. Only controlled studies can provide data as to whether interventions lead to desired effects. In clinical practice, however, the issue is whether we can provide interventions that we believe are effective. We believe one mechanism by which that can be done is the use of a systematic approach to patient management. The HOAC II, we believe, allows such an approach for the integration and use of the best evidence available. The HOAC II requires hypothesis testing in clinical practice. In the context of the HOAC II, the therapist has a hypothesis as to what is causing a problem, and usually that is an impairment leading to diminished function. An impairment is a loss of function in an organ or system, such as a loss of motion, strength, or coordination. These losses are all measurable. Therefore, if the intervention is focusing on the cause of the problem, as the impairments lessen, function should improve. The problems should be diminishing, and goals should be closer to attainment. At times, a problem may be hypothesized to be caused by multiple impairments. But how do we know whether we have identified the correct diagnostic hypothesis?

In the HOAC II, changes in the impairment measure are almost always monitored. The level of improvement in impairment that the patient needs to achieve to eliminate the problem is called the "testing criteria." When multiple impairments occur, each will have to be measured, with testing criteria established for each. When testing criteria are met, the problem should have been eliminated and the related goals achieved. In this way, the therapist tests the original hypothesis for existing problems.

For Each Anticipated Problem, Establish Predictive Criteria

The conceptual basis for the testing criteria comes from the application of traditional scientific methods of inquiry into clinical practice. Unfortunately, this cannot easily be done for anticipated problems. In science, proving a negative is often seen as impossible because hypotheses are not testable in the usual sense. If we intervene to prevent a contracture, for example, we cannot prove that we achieved anything. The failure of a contracture to develop may be due to an intervention or because a contracture would not have developed anyhow. With an anticipated problem, however, we can argue that, based on what is known, something might have occurred had we not intervened. Therefore, the means of justifying interventions focused on prevention is not in this part of the algorithm, but rather it is described in the section where therapists supply the rationale for each anticipated problem (see section titled "For Each Anticipated Problem, Identify the Rationale for Believing Anticipated Problems Are Likely to Occur Unless Intervention Is Provided").

The testing criteria for existing problems are used to examine the viability of the hypothesis. The predictive criteria for the anticipated problems are different from the testing criteria. A goal for an existing problem can be achieved within a known time period. If we are trying to keep something from happening, when do we declare we have succeeded? In health care, often the best we can do is to eliminate risk factors; therefore, the predictive criteria relate to risk factors. If risk factors can be eliminated during some finite time period, the predictive criteria would reflect this possibility.

For example, if a patient is seen as being at risk for the postsurgical development of pneumonia (a pathology), pneumonia would be an anticipated problem. Physical therapy interventions may be gait training and breathing exercises. Both of these interventions are preventive because they relate to the potential reduction in risk of pneumonia. The predictive criteria for ambulation may be a certain distance walked per day, whereas the predictive criteria for the breathing exercises may be an inspiratory level with an inspirometer and an observed level of competence in generating a productive cough. When these predictive criteria are achieved, the patient should no longer need the preventive interventions.

This finite situation can be contrasted with people who have permanent disabilities and chronic injuries who may have to reduce risk factors for the rest of their lives. A patient with recurrent back pain, for example, may be taught prophylactic exercises, and the predictive criteria may be a level of competence and degree of adherence in doing those exercises. When the desired level of competence and adherence is achieved--that is, when the predictive criteria have been achieved--the patient would no longer need ongoing physical therapy intervention. The assumption is that the patient would continue to carry out the exercises as taught, or that the therapist might need to see the patient periodically to determine whether the predictive criteria are still being met (ie, the patient is still performing the exercises with the appropriate frequency and in the proper manner).

The predictive criteria are used to determine how long interventions designed for prevention should be carried out. In this way, predictive criteria are somewhat similar to goals, but they exist only for anticipated problems. They are not goals because they are worth achieving only if sufficient evidence indicates that a problem might occur. The value of achieving the predictive criteria is entirely dependent on the case made for anticipating that a problem might occur.

Establish a Plan to Reassess Testing and Predictive Criteria and Establish a Plan to Assess Problems and Goals

Hypotheses often require multiple testing criteria, and changes in the impairments measured for these criteria may not all change at the same rate. Similarly, achievement of various predictive criteria may not happen at the same time. Measuring impairments and disabilities and doing a re-evaluation at every session is time-consuming and impractical. Therapists should have reasonable expectations as to when meaningful and therefore measurable changes will occur and should plan a re-evaluation schedule accordingly. Similarly, not all goals can be achieved at the same time, and therefore they should be checked based on a logical plan (ie, short-term goals should be checked sooner than-long term goals). By committing to an evaluation schedule, a therapist using the HOAC II has identifiable points in time when the patient's status will be reconsidered. Without such a plan, re-evaluation is often chaotic, and measurements may be obtained at intervals that may make interpretation of data difficult.

Plan Intervention Strategy and Tactics

If the therapist thinks muscle weakness is the impairment contributing to a disability, the most obvious approach would be to use exercise to increase the force-generating capacity of the involved muscles. The strategy would be the use of exercise. Describing the strategy alone is insufficient, because many types of exercises exist. The HOAC II asks therapists to describe the tactics (specific exercises and frequency) they would use. If we were dealing with an anticipated problem (such as the development of postoperative pneumonia), there might be 2 strategies: (1) teach the patient how to clear his or her airway and (2) teach the patient preventive measures such as frequent ambulation and use of an inspirometer. The tactic for the first strategy (airway clearance) may be to have the patient cough a specified number of times per hour (and the patient could be shown how to determine if the cough is productive). The tactic for generalized prevention might be correct use of an inspirometer 5 times daily and ambulation 5 times daily. Strategies are broad statements of what types of things need to be done, whereas tactics are the elements of the intervention. Tactics specify the frequency, duration, and intensity of the interventions.

Implement Tactics

Once tactics have been identified, they need to be implemented. Most often the therapist will be doing the implementation. Sometimes, as when a person has a home exercise program, the patient may be doing the intervention. Family members, other health care personnel (eg, physical therapist assistants), and other caregivers all may have a role in implementing tactics. The physical therapist, however, should note who is implementing which tactics. We believe the therapist must recognize that, as long as these tactics are part of the physical therapy plan of care, the therapist must assume responsibility for overseeing, evaluating, and determining whether modifications should be made to tactics.

Part 2

In Part 1 of the HOAC II, the therapist working with the patient and others developed an intervention plan (a series of strategies and tactics that is conceptually similar to the plan of care as defined in the Guide).6 Justification for the interventions was based on the therapist's concepts of what was causing problems. Therefore, by definition, much of what occurs in Part I arises from conceptual models that can only be examined in the context of intervention (eg, did the intervention lead to a desired outcome?). Part 2 is far less conceptual in nature and consists of questions that are designed to provide insights into whether any aspect of patient management is deficient, including whether the original goals were viable.

The steps in Part 2 can be used for documentation, or they can be used to less formally guide decision making. The most important element, however, is that, by using Part 2, the therapist must account for all changes in goals, tactics, strategies, and hypotheses. In addition, the therapist needs to document whether the criterion measure chosen is still viable and whether it is still reasonable to expect to see the desired change in the criterion measure. Part 2 not only assists in the evaluation process, it provides the logical framework for examining the effects of all interventions. Use of Part 2 requires the therapist to document what happened to a patient, even if the result is an acknowledgment that the result was less than was expected. Documentation may be particularly useful on occasions when factors outside of the therapist's control led to a termination of the intervention. For example, by following the steps in Part 2, a therapist can make an argument to a payer that goals were not achieved (even though progression was being made on the criterion measure) because there was too little time allowed for the intervention.

Part 2 consists of 2 flow diagrams. The first diagram (Fig. 3) leads the therapist through a series of questions for all existing problems (regardless of who generated them). The second diagram (Fig. 4) also consists of a series of questions, but these questions relate to anticipated problems (regardless of who generated the problem list). The peculiar nature of prevention (ie, therapists may take credit for what does not occur by making sure a risk factor is reduced or eliminated) leads to somewhat different questions. The most notable difference is that the first question when dealing with anticipated problems is asking whether the problem has occurred. If it has, prevention did not work, and a new problem needs to be added to the existing problem list.

Examining the Hypothesis for Existing Problems

If a patient's goals are met (the problems are resolved), the question remains as to whether this occurred because of the intervention. Although causality cannot be claimed in the absence of controlled studies, the algorithm and use of the testing criteria allow therapists to gain some insights as to whether their approaches seemed appropriate and their interventions beneficial and therefore whether their hypotheses were appropriate.

When therapists set the testing criteria, they are stating that a level of performance (usually of an impairment measure) is needed for the goal to be achieved. If the goal is achieved and the testing criteria are not met, the therapist's hypothesis is incorrect (or the criteria's levels were incorrect). If the testing criteria are met and the goal is not achieved, the hypothesis is at best incomplete; that is, other causes may exist in addition to those identified, or those identified are irrelevant. These are absolute examples. What is less clear is what is happening when there is movement toward meeting goals and when there is also an indication that the impairments measured for the testing criteria are also becoming less pronounced. In these cases, no simple test of the viability of a hypothesis exists, so the therapist must extrapolate and consider the overall picture and determine whether the hypotheses and criteria should be maintained in the same form.

When a therapist thinks a problem has multiple causes and generates multiple hypotheses, it is impossible to say with certainty whether achieving appropriate levels of all the testing criteria led to attainment of a goal. The possibility exists, for example, that if there were 3 hypotheses, 2 of the hypotheses were correct and the third hypothesis was either redundant or unnecessary. When all testing criteria are achieved, the therapist has no way of knowing what would have happened with this patient if only 2 hypotheses had been met.

Following a CVA, a patient might be incapable of dressing. Among the many possible causes of this deficit could be: (1) weakness, (2) lack of coordination, and (3) poor position sense. All 3 might be hypothesized as causes of the problem. Testing criteria for weakness could be a force level obtainable on a hand dynamometer. For the lack of coordination, the testing criteria might be a level of performance on a coordination test, and, for position sense, the testing criteria might be the ability to place a limb on a target with less than a specified number of errors. If all 3 criteria were met and the patient achieved the goal of dressing himself or herself, the therapist could not be certain whether this goal still could have been obtained if only 2 of the 3 testing criteria were met. The therapist, however, may develop an opinion based on the time course of events; that is, how did the attainment of the goal over time relate to changes in the testing criteria? This case illustrates how, even in the absence of being able to definitively test hypotheses, therapists can better understand patient management by use of the algorithm. In this manner, the HOAC II serves as a means of ongoing feedback for professional development, independent of what occurs with each patient.

Summary

The HOAC II was designed to facilitate the use of science and evidence in practice, and to do so in a manner that is not intrusive on clinical practice. We believe that much of what we ask clinicians to do in the algorithm is already part of their practice but that it occurs in a less defined manner and without a context for documentation and discussions among colleagues. Among the differences between this version and the original HOAC are the mechanisms for justifying prevention and, more importantly, for developing measurable outcomes related to prevention as well as defining the time it will take to achieve reduction of risk factors.

Appendix.

Terms Used in the Hypothesis-Oriented Algorithm for Clinicians II (HOAC II)

Anticipated Problems:

These can be identified by the patient, the physical therapist, or any other person and are statements that describe deficits that the therapist believes will occur if an intervention is not used for prevention.

Examination Strategy:

This is the plan for examination that a physical therapist uses based on the therapist's experience, available data relating to the patient, and information on similar patients. Because not all possible tests and measures are used, the choice is considered a hypothesis-driven strategy in the HOAC II.

Existing Problems:

These can be identified by the patient, the physical therapist, or any other person and are statements that describe deficits in a person's function (disability).

Goals:

Functional deficits are problems, whereas goals are descriptions of function that will be recovered as a result of one or more interventions.

Hypothesis:

The reason that a patient's problems (which are usually at the disability level) exist is not necessarily known, but in order for a physical therapist to carry out an intervention, the therapist must have an idea as to the underlying causes. In the HOAC II, the therapist's conjecture as to the cause is a hypothesis. Often there will be more than one hypothesis, and usually the hypothesis will involve one or more impairments causing a deficit in function (ie, a disability).

Intervention Strategy:

These are the overall types of interventions that the physical therapist believes are needed to alleviate problems (eg, exercises designed to increase range of motion are a strategy, whereas the specific exercises are tactics).

Non-Patient Identified Problems (NPIPs):

These are problems identified (at least initially) by people other than the patient but that are added to the patient's problem list after consultation with the patient (these can be existing or anticipated problems).

Patient-Identified Problems (PIPs):

These are problems identified by the patient (these can be existing or anticipated problems), and because they are generated by the patient, they cannot be removed from the problem list without the patient's consent.

Predictive Criteria:

These are critical values (thresholds) for measurements, which, if met, would indicate that one or more problems will most likely be avoided because risk factors were reduced or eliminated. Sometimes the measurement may be how often someone does a task or whether a patient demonstrates competency in a prevention program (eg, does stretching or prophylactic back exercises).

Tactics:

These are the elements of an intervention. For instance, the exercises or techniques used to treat the patient or client are the specific elements of the intervention, whereas the overall purpose of the interventions is the strategy.

Testing Criteria:

These represent critical values (thresholds) for measurements, which, if achieved, would suggest the hypothesis (or hypotheses) is correct if the associated problem(s) is resolved (these are most often measurements of impairments).

References

(1) Rothstein JM, Echternach JL. Hypothesis-oriented algorithm for clinicians: a method for evaluation and treatment planning. Phys Ther. 1986;66:1388-1394.

(2) Delitto A, Snyder-Mackler L. The diagnostic process: examples in orthopedic physical therapy. Phys Ther. 1995;75:203-211.

(3) Jette AM. Introduction: physical disability. Phys Ther. 1994;74:379.

(4) Jette AM. Physical disablement concepts for physical therapy research and practice. Phys Ther. 1994;74:380-386.

(5) Feder G, Eccles M, Grol R, et al. Clinical guidelines: using clinical guidelines. BMJ. 1999;318:728-730.

(6) Guide to Physical Therapist Practice. 2nd ed. Phys Ther. 2001;81: 9-744.

(7) Pope A, Tarlov A, eds. Disability in America: Toward a National Agenda for Prevention. Washington, DC: National Academy Press; 1991:84-89, 223-241.

(8) Elstein AS, Shulman LS, Sprafka SA. Medical Problem Solving: An Analysis of Clinical Reasoning. Cambridge, Mass: Harvard University Press; 1978.

(9) Payton OD. Clinical reasoning process in physical therapy. Phys Ther. 1985;65:924-928.

(10) Goodman CC, Snyder TEK. Differential Diagnosis in Physical Therapy. 3rd ed. Philadelphia, Pa: WB Saunders Co; 2000.

(11) Sackett DL, Haynes RB, Guyatt GH, Tugwell P. Clinical Epidemiology: A Basic Science for Clinical Medicine. 2nd ed. Boston, Mass: Little, Brown and Co Inc; 1991.

(12) Task Force on Standards for Measurement in Physical Therapy. Standards for tests and measurements in physical therapy practice. Phys Ther. 1991;71:589-622.

(13) May BJ. Assessment and treatment of individuals following lower extremity amputation. In: O'Sullivan SB, Schmitz TJ, eds. Physical Rehabilitation: Assessment and Treatment. 4th ed. Philadelphia, Pa: FA Davis Co; 2000:632-633.

(14) Dawber TR. The Framingham Study: The Epidemiology of Atherosclerotic Disease. Cambridge, Mass: Harvard University Press; 1980.

(15) Straus SE, Sackett DL. Getting research findings into practice: using research findings in clinical practice. BMJ. 1998;317:339-342.

(16) Randall KE, McEwen IR. Writing patient-centered functional goals. Phys Ther. 2000;80:1197-1203.

JM Rothstein, PT, PhD, FAPTA, is Professor, Department of Physical Therapy, College of Applied Health Sciences, University of Illinois at Chicago, 1919 W Taylor St, 4th Fl, Room 456, Chicago, IL 60612 ([e-mail]). Address all correspondence to Dr Rothstein.

JL Echternach, PT, EdD, ECS, FAPTA, is Professor and Eminent Scholar, School of Physical Therapy, Old Dominion University, Norfolk, Va.

DL Riddle, PT, PhD, is Professor, Department of Physical Therapy, Medical College of Virginia Campus, Virginia Commonwealth University, Richmond, Va.

All authors provided concept/idea/project design, writing, and project management. The authors acknowledge the efforts of Andrew Guccione, PT, PhD, FAPTA, Julie Fritz, PT, PhD, ATC, and David Scalzitti, PT, MS, OCS, for reviewing an earlier draft of the manuscript.

This article was submitted March 12, 2002, and was accepted December 2, 2002.