Purpose of Examination of Function
Analysis of function focuses on the identification of pertinent activities and measurement of an individual's ability to successfully engage in them. In essence, functional testing measures how a person does certain tasks or fulfills certain roles in the various dimensions of living described in the framework of the ICF. Application of selected functional tests and measures yield data that can be used as (1) baseline information for setting function-oriented goals and outcomes of intervention; (2) indicators of a patient's initial abilities and progression toward more complex functional levels; (3) criteria for placement decisions, for example, the need for inpatient rehabilitation, extended care, or community services; (4) manifestations of an individual's level of safety in performing a particular task and the risk of injury with continued performance; and (5) evidence of the effectiveness of a specific intervention (medical, surgical, or rehabilitative) on function.
Physical therapists possess a unique body of knowledge related to the identification, remediation, and prevention of movement dysfunction. Thus, they have traditionally been involved in the examination of physical function. Other members of the rehabilitation team, including the occupational therapist, nurse, rehabilitation counselor, and recreational therapist, are also typically involved in administering and interpreting functional tests. Some formal instruments were designed to be completed collectively by the team. Other tests are compiled in separate sections by specific health professionals and housed together in the patient's chart. Where teams exist, physical therapists are typically responsible for the testing of aspects of function related to mobility, such as bed mobility, transfers, and locomotion (wheelchair mobility, ambulation, negotiation of stairs and graded elevations, walking for longer distances in the community, and so forth). Instruments to measure activities of daily living (ADL) may be administered by a physical therapist alone or cooperatively with other health professionals. When overlap among team members exists, for example, the performance of toilet transfers, the data may be collected by the physical therapist, an occupational therapist, or a nurse. In these instances, testing should be coordinated to reduce duplication and unnecessary patient stress. In noninstitutional settings or where there is no team, the physical therapist is often responsible for determining all aspects of these instruments.
Function tests can utilize two highly divergent perspectives on what is to be tested or measured by the physical therapist. It is extremely important that the therapist determine in advance whether data are needed to describe the habitual level of a patient's ability to do certain tasks and activities, or to identify the patient's capacity to perform certain tasks and activities, whether the patient habitually performs up to that level or not, or even performs them at all. These perspectives are incorporated within the ICF by the constructs of performance and capacity and the ICF allows for the separate coding of both constructs.
These divergent viewpoints directly affect what types of tests and measures should be chosen and what parameters of measurement are appropriate to yield data useful to making clinical judgments. Most important, physical therapists must consider the differences between capacity for function and habitual function in determining the prognosis for rehabilitation and estimating the likelihood of the success of an intervention. Patients accept a therapist's recommendations regarding the anticipated goals of treatment only if there is the perceived need and motivation to function habitually at the highest level of ability. Understanding the difference between what a person actually does or would be willing to do and what that person potentially could do is an essential component of designing realistic, and achievable, functional goals. For example, even though a person might have the capacity to climb stairs, there may not be any willingness to do so. Ultimately, physical therapists must abide by each patient's own decision regarding which tasks and activities will be incorporated into a daily routine and what is a meaningful level of function, regardless of the therapist's professional opinion.
Irrespective of the particular instrument used, there are several basic considerations to be kept in mind. The setting chosen must be conducive to the type of testing and free of distractions. Instructions should be precise and unambiguous. Testing may be biased by fatigue. If a patient performs best in the morning but tires by afternoon, an accurate determination of functional ability must consider the variation in the patient's performance. Therapists should be aware of patients whose energy fluctuates during the day, and interpret the data accordingly. In general, information related to body functions and body structure, activities, and participation, as well as personal and environmental factors, should be generated during the initial examination (or as soon as feasible) so the information may be considered together to develop a picture of a patient's function. Retesting should occur at regular intervals during treatment to document progress and before discharge from the episode of care.
A performance-based test involves observing the patient during the performance of an activity. Generally speaking, the therapist who chooses a performance-based test is searching for an indication of what a patient can do under a specific set of circumstances, which may or may not be similar to the natural environment in which the patient functions. If a performance-based test is chosen with the intention of making inferences about how the patient will perform at home, the conditions and setting should be as similar as possible to the actual environment in which the patient usually performs the tasks and activities. A performance-based approach may be used either to describe the patient's current level of function or to identify the maximum level of function possible.
During the administration of the test, each task is presented and the patient is asked to perform it. For example, to examine current level of function in wheelchair mobility, a patient would receive this instruction: "Push your wheelchair over to that red chair and stop." To determine the patient's maximum level of function in this activity, the instruction might specify a particular manner of performance: "Push your wheelchair over to that red chair as quickly as you can and stop." Understanding the difference between these two commands, even though both are observation-based performance of wheelchair mobility, is essential to sound clinical decision making. Data from the first example identify only what the patient can do under specific circumstances, but does not support the inference that the patient will be able to wheel across a busy intersection in the short time span allotted to a typical pedestrian walkway. The form of the instruction determines whether an inference can be made about the patient's maximal level of function in formulating the goals for intervention and the plan of care.
In either case, a patient is given no additional instructions or assistance unless he or she is unable or unsure of how to perform. Then only as much direction or assistance as is needed is given. Appropriate safety precautions should be taken during the session so that the patient does not attempt tasks that are potentially dangerous.
A number of tests of impairments are sometimes also referred to as functional performance measures, including the 6-Minute Walk Test,13 the Physical Performance and Mobility Examination,14 the Functional Reach Test,15,16 the Get Up and Go Test,17 the Timed Up and Go Test,18 and the Physical Performance Battery.19 A performance instrument of this sort typically measures either a complex integration of impairments, the performance of actions, or a combination of both by direct observation. Overall, the tests do provide some insight into the individual's capabilities to maintain a posture, transition to other postures, or sustain safe and efficient movement. The data from such a test, gathered under controlled conditions, characterize a person's performance limitations as a result of impairments, and may purport to predict the success or failure of an individual in performing goal-directed tasks or activities under natural conditions, using a score that summates the combined effects of impairments throughout and across systems on movement dysfunction. Each of these tests can contribute to an understanding of an aspect of a person's function, but they should not be used to represent all aspects of function. Although these tests employ the method of direct observation of performance, they most often do not measure the task or activity as it might be accomplished in the "real" world of the patient, which is also influenced by motivation and habit.
In contrast to the method of direct observation, useful data on how a person functions may also be gathered by self-report, in which the patient is asked directly either by the therapist or a trained interviewer (interviewer report) or through the use of a self-administered report instrument. The critical issue in the ability of a self-report to capture function correctly and completely lies in providing clearly worded questions without language bias, concise directions on completing the questions, and a format that encourages accurate reporting of answers to all questions. Self-report is a valid method of determining function and may be preferable to performance-based methods in some circumstances.20 Self-reports should be designed so that questions are asked in a standard format and answers are recorded as specified by the predetermined choices. Long paper-and-pencil tests may be difficult for those with upper extremity disability.
Clinical personnel who will act as interviewers must be trained to administer a questionnaire. Interviewers should practice until they have reached a high degree of agreement with expert examiners of the same cases. Periodic retraining may be necessary if interviewers do not have frequent practice administering the instrument. The interview should be scheduled with the patient in advance and conducted in an environment conducive to complete concentration. Interviews may be conducted by phone or in person, but the mode of administration should be kept consistent if comparisons of the data are to be made. Ad lib prompting by the interviewer or caregivers for answers is discouraged because these intrusions into the patient's self-report tend to bias results. If the patient has had help in filling out a form or responding to questions, this should be noted. Similarly, if the data have been provided by a spouse, family member, or caregiver, this should be documented as well.
The distinction in perspectives on function that was discussed regarding performance-based measures of function also holds for self-reports. It is extremely important to distinguish between questions that indicate a person's habitual performance (e.g., "Do you cook your own meals?") and those that identify a person's perceived capacity to perform a task (e.g., "If you had to, could you cook your own meals?"). It may also be important to distinguish between an individual's performance of an activity and his or her confidence in performance of an activity. For example, confidence and performance for 21 items are measured in separate scales in APTA's Outpatient Physical Therapy Improvement in Movement Assessment Log (OPTIMAL).21
The time frame reference of self-reporting is also a relevant consideration. A therapist should decide in advance if the relevant "window" on a person's functional level is the past 24 hours, last week, last month, or the previous year. One can easily imagine how the same person might respond differently regarding the same functional activity depending on frame of reference. Instruments that examine only short-term objectives may not relate well to the long-term objectives of a rehabilitation program.
Instrument Parameters and Formats
Performance-based and self-report instruments grade performance on a number of different criteria in a variety of formats. There is no one parameter or format that is perfect for every type of clinical encounter or research need. It is particularly important that documentation of a patient's progress not be blunted by floor or ceiling effects. For example, if a therapist wishes to measure changes in function among generally well elderly patients and the most advanced functional activity on an instrument measures "independent ambulation on level surfaces," there would be no room to demonstrate either progression or decline except around ambulation on level surfaces. Similarly, a patient who was severely debilitated might improve in transfers from needing the maximum assistance of two persons to maximum assistance of one. If the instrument only measures change from "maximum assistance" to "moderate assistance," this patient's real improvement will not be recorded.
Therapists should use descriptive terms that are well defined and unambiguous. Meanings of descriptive terms should be clear to all others using the medical record. Box 8.1 provides a sample set of acceptable terms and definitions. Additional terms used to qualify function include dependence and difficulty. Most often, the term independent refers to the complete absence of a need for human or mechanical assistance to accomplish a task, but some scoring systems consider reliance on devices and aids as a modified form of independence when used without the help of another person. The use of equipment during the performance of a functional task should be explicitly noted; for example, "independent in ambulation with axillary crutches" or "independent in dressing with adapted clothing and a long-handled shoe horn."
Difficulty is a hybrid term that suggests an activity poses an extra burden for the patient, regardless of dependence level. It is unclear whether it is a measure of overall perceptual-motor skill, coordination, endurance, efficiency, or a combination of measures. Difficulty can be measured in two ways. One approach assumes that difficulty is likely to be present and quantifies the degree of difficulty that the individual experiences while performing the activity (e.g., "How much difficulty do you have while doing household chores? None, some, or a great deal?"). The other approach quantifies the frequency that the difficulty is encountered (e.g., "How often do you have difficulty putting on your shoes? Never, sometimes, very often, or always?").
Box 8.1 Functional Examination and Impairment Terminology
Independent: patient is able consistently to perform skill safety with no one present.
Supervision: patient requires someone within arm's reach as a precaution; low probability of patient having a problem requiring assistance.
Close guarding: person assisting is positioned as if to assist, with hands raised but not touching patient; full attention on patient; fair probability of patient requiring assistance.
Contact guarding: therapist is positioned as with close guarding, with hands on patient but not giving any assistance; high probability of patient requiring assistance.
Minimum assistance: patient is able to complete majority of the activity without assistance.
Moderate assistance: patient is able to complete part of the activity without assistance.
Maximum assistance: patient is unable to assist in any part of the activity.
A. Bed Mobility
Independent—no cuinga is given
| ||may require cues |
B. Transfers, Ambulation
Independent—no cuing is given
| ||may require cues |
C. Functional Balance Grades
|1. Normal || |
Patient able to maintain steady balance without support (static).
Accepts maximal challenge and can shift weight easily and within full range in all directions (dynamic).
|2. Good || |
Patient able to maintain balance without support, limited postural sway (static).
Accepts moderate challenge; able to maintain balance while picking object off floor (dynamic).
|3. Fair || |
Patient able to maintain balance with handhold support; may require occasional minimal assistance (static).
Accepts minimal challenge; able to maintain balance while turning head/trunk (dynamic).
|4. Poor ||Patient requires handhold and moderate to maximal assistance to maintain posture (static). Unable to accept challenge or move without loss of balance (dynamic). |
|5. No balance || |
Often it is helpful to qualify a person's performance by linking observations with nonspecific indicators of impairments such as the energy consumption required to complete the functional task and the degree to which patients must exert themselves to engage in the activity. Simple measurements of a patient's physiological response to activity generally include heart rate, respiratory rate, and blood pressure, obtained at rest (baseline measurements), during (as possible), or immediately after completion of the most stressful elements of the activity. For example, "heart rate increased to 100 beats per minute with independent ambulation on stairs; no increase in respiratory rate." In addition, the patient's perceived fatigue, perception of exertion, and overt signs of physiological stress, such as shortness of breath, also should be noted. These notations may assist the therapist in a quick identification of some obvious impairments that limit function, which should be followed by more specific tests and measures of impairment.
Additional descriptors frequently used to qualify functional performance further include (1) pain, (2) fluctuations according to the time of day, (3) medication level, and (4) environmental influences. Any factors that modify a patient's function should be carefully noted and considered by the physical therapist evaluating examination data.
The time it takes to complete a series of activities is often used to enhance a therapist's quantification of function when a given speed of performance is required or an improvement in performance speed is expected. A common example of timed functional skills is found in premedication and postmedication performance of individuals with Parkinson disease who are placed on l-dopa therapy. Examples of activities that may be timed include (1) walking a set distance; (2) writing one's signature; (3) donning an article of clothing; and (4) crossing a street during the time of a "Walk" light. Scores of timed tests should not be taken as absolute, but rather as one dimension of performance. Although the ability to complete a particular activity in a specified period of time does provide one kind of important data on a patient's overall ability, it may not always be correct to conclude that what is being measured as "quicker" can be interpreted as "better." For example, the patient may get dressed quickly (within seconds), but do so with poorly coordinated movements and a haphazard outcome. When the task is slowed down, the movements may become more coordinated, with a more satisfactory functional outcome, even though the time taken to do the task increases. Similarly, certain medical conditions that affect energy expenditure may require that the patient properly pace a functional activity to complete it successfully. Thus, time scores alone do not always yield the complete functional picture. When interpreted in light of other aspects of the patient's clinical presentation, they do provide an added dimension to the evaluation of data collected during a functional examination.
Function can be measured with tests that report data as nominal, ordinal, interval, and ratio measures. The clinician should consider the uses of the measure when deciding which format to use. In cases where the clinical decision is nominal, such as is the patient ready for discharged to home, a nominal measure such as whether the patient can independently ascend 10 stairs may be adequate. When ratio measures are obtained, such as the score on a Berg Balance scale, the clinician may interpret the score as a dichotomous measure related to the decision (e.g., Does the patient have or not have adequate balance for discharge to home?). In cases where the clinical decision is more complex, such as the amount of assistance a patient needs with activities, nominal measures cannot be used and the measure should reflect the type and amount of information needed for the decision.
One of the simplest formats in functional tests uses a nominal level of measurement by presenting a checklist of various functional tasks on which the patient is simply scored as able to do/not able to do, independent/dependent, completed/incomplete, or the like. The results are not particularly descriptive of the exact nature of an individual's limitations and usually require further examination before interpretation. Nominal measures, however, may be helpful in making dichotomous decisions. For example, knowledge of the ability to perform ADL skills by themselves is important in deciding if a patient can be discharged, or not, to living independently at home.
A few tests use descriptive scales that describe a range of performance or the degree to which a person can perform the task. Most commonly, the scales are ordinal or rank-order scales (e.g., "no difficulty," "some difficulty," or "unable to do"; or "always," "sometimes," "rarely," or "never"). Scales may be graded in ascending or descending order. The primary drawback in using such a system to score function is that these grades do not define categories that are separated by equal intervals. For example, it is not possible to tell whether the patient who went from maximal assistance to moderate assistance changed as much as a patient who also went one level between moderate assistance and minimal assistance.
Summary or additive measures grade a specific series of skills, award points for part or full performance, and sum the subscores as a proportion of the total possible points, such as 60/100 or 6/24 and so forth. Although the scales typically may include a score of zero, this value represents a floor effect of the scale and not necessarily the absence of the construct. One example of a summary measure, which is well known to physical therapists, is the Barthel Index.22
Some formal, standardized instruments for testing function summarize detailed information about a complex area of function into an overall index score. Use of these instruments facilitates the interpretation of complex data and enables the clinician to perform crossdisease, cross-program, and cross-population comparisons of function. Caution must be exercised in considering only summated scores, however, because potentially important individual differences in functional ability can be masked.23 A patient who is limited in only a few of the many tasks covered on a functional test will most likely score well, despite what could be substantial limitations in discrete functional activities that are pertinent to the physical therapist's anticipated goals of treatment. Similarly, two patients with the same numeric score might be quite different in their functional deficits, having gained (or lost) their points on different activities. Although these measures yield a "hard number," which is regarded statistically as an interval level of measurement, the degree to which "points" are truly equal intervals or only ordinal should be carefully scrutinized.
Visual or linear analog scales attempt to represent measurement quantities in terms of a straight line placed horizontally or vertically on paper (Fig. 8.4). The endpoints of the line are labeled with descriptive or numeric terms to anchor the extremes of the scale and provide a frame of reference for any point in the continuum between them. Some scales will also use descriptors or numeric intervals between the endpoints to assist the individual in grading responses. Commonly a visual analog line of 10 centimeters (100 millimeters) is used. The patient is asked to bisect the line at a point representing self-reported position on the scale. The patient's score is then obtained by measuring from the zero mark to the mark bisecting the scale.
A visual analog scale for measuring pain or other symptoms. The patient is instructed to mark the line at the point that corresponds to the degree of pain or severity of symptoms that are experienced.
Examples of the use of visual analog scales in rehabilitation settings may be to measure pain, dyspnea, function, or satisfaction with care. Since visual analog scales include a true zero and equal intervals (e.g., mm) they may be analyzed as ratio measures. In contrast, some clinicians may use a numeric rating scale (e.g., rate your function on a scale from 0 to 10) to measure similar impairments. Although providing a numeric rating may be quicker to obtain in clinical settings, scores obtained may not represent interval or ratio data as the reporting of a numeric value may not represent equal intervals. For example, a 4 reported by a patient may not be twice as much function as a 2 reported by another patient. This is due to the nature of interval and ratio scales because a ratio scale allows for the comparison of scores using addition, subtraction, multiplication, or division and an interval scale allows for the comparison of scores using addition or subtraction. These mathematical functions cannot be performed with ordinal or nominal scores.
Knowledge of the level of measurement is important in analyzing data from groups of patients, such as a rehabilitation unit wishing to summarize the functional status of patients admitted during a specified time period. For interval and ratio measures, means and standard deviations may be calculated (assuming the data follow a normal distribution). For ordinal measures, medians and interquartile ranges are appropriate, whereas nominal measures should be reported as modes or frequency counts.