Comparison of Prosthetic Mobility Outcome Measures
Trina Boyce, Prosthetic Resident
June 2003
ABSTRACT
According to the Random House Webster's College Dictionary, prosthesis is defined as "a device, either external or implanted, that substitutes for or supplements a missing or defective part of the body". In the field of Orthotics and Prosthetics, prosthesis is used synonymously with artificial limb. Prostheses can be very beneficial to amputees when worn and actively used, but amputees commonly do not wear or use their prostheses. It would be beneficial to practitioners if a method or tool were available to predict an amputee's mobility with a prosthesis. The benefit of such a tool would be its use in assisting practitioners in choosing an appropriate prosthetic prescription. There have been a variety of measurement tools used to assess the prosthetic outcome of amputees; however, some of these outcome measures are not necessarily specific to the amputee population and measure rehabilitation outcome in general. They range from performance tests to self-reported capabilities. Despite the various outcome measurement tools used, there isn't one that is universally accepted which measures prosthetic mobility outcome. It is difficult to create a prosthetic mobility outcome measurement tool because it has not been agreed upon which factors should be considered in the process. To be acceptable an outcome measure needs to be valid, reliable and efficient. Validity refers to the ability of the outcome tool to measure what it is designed to measure. Reliability refers to the consistency of the results from the outcome measurement. Efficiency refers to the relative ease with which the outcome measure can be applied. If the outcome measure is difficult to apply or takes too long to administer, it would not be appropriate for clinical settings. The validity, reliability and efficiency of various outcome measures such as the Prosthetic Evaluation Questionnaire and the Amputee Mobility Predictor will be compared and contrasted. By comparing a few of the outcome measures, it is hopeful that one will prove to be superior over the others.
INTRODUCTION
Two categories of outcome measures will be analyzed in this literature review. The first category consists of outcome measures that are specifically designed for the amputee population such as the Amputee Mobility Predictor (AMP), the Prosthetic Profile of the Amputee questionnaire (PPA) and the Prosthetic Evaluation Questionnaire (PEQ). The second category of outcome measures consists of more general rehabilitation measures in which the amputee population is tested. These outcome measures are the Functional Independence Measure (FIM), the Barthel Index, and the Goal Attainment Scaling (GAS) measure. The various measurement tools test different aspects of the functional capabilities of amputees and thus the outcome measures have both advantages and disadvantages.
Amputee Mobility Predictor
The purpose of the AMP is to predict the ability of a patient to ambulate with a prosthesis before prosthetic fitting.2 There are two forms of the AMP available. The first, AMPPRO, is meant to predict ambulatory potential of patients who already have a prosthesis. The AMPnoPRO is meant to predict ambulatory potential of patients who do not have a prosthesis. In order to participate in the AMP testing, patients could not participate in any rehabilitation program at the time of testing and they had to have pain free, well fitting prostheses.2 There were also contraindications for taking part in the AMP test which were as follows: patients who had 1) obvious mental deterioration which was based on an interview 2) advanced neurological disorder 3) compromised circulation in the sound limb which resulted in ulcers and/or infections 4) severe knee or hip flexion contractures which could not be reduced and 5) pulmonary disease, congestive heart failure or angina pectoris.2 Those patients who were thought to be medically stable enough to perform the test were included. With the presence of inclusion and exclusion criteria, human error in misinterpretation is always a possibility. For example in the AMP it is unclear why the five specific exclusion criteria were selected as opposed to other medical conditions. Also the term "medically stable" is vague and can be interpreted differently. Misinterpretation may also result when determining if the patient's prosthesis is pain free. Patients may interpret the term "pain free" differently. For example, a patient may suffer from bearable pain and take this to mean pain free, or a patient may incorrectly identify pressure as pain.
The AMP consists of performing tasks already used in two preexisting outcome measurement tools. AMP also includes tasks that are included in the US Health Care Financing Administration (HCFA) K-level classification system.2 The various tasks tested include transfers, sitting and standing balances and traversing obstacles. The scoring system consists of three scaling options: 0 indicates inability to perform the specified task, 1 signifies that minimal level of achievement or assistance was needed to perform the task and 2 denotes the task was completed independently.2 The advantages of the design of the AMP test is that depending on the score received, it parallels with the K-level classification system. Also, the AMP can be administered in a clinical setting and requires no advanced equipment. In my opinion the scoring system enables the test to be more accurate. By having fewer scoring options, there is less room left for misinterpretation.
Prosthetic Evaluation Questionnaire
The PEQ is a self-report instrument that measures lower extremity prosthetic mobility. Unlike the AMP, the PEQ measures the patient's potential rather than what the patient can perform. In addition to assessing mobility, the PEQ assesses overall well-being, emotional and social health and prosthetic function. For the purpose of testing reliability and validity, only items concerning mobility were used in the study. A total of 13 items, which addressed ambulation and transfer situations with the use of the prosthesis, were used.3 Specific activities addressed included walking up and down stairs and hills, walking on various surfaces and sitting and rising from a chair or toilet. A numeric scale ranging from 0 to 10 was the response scale used. The number 0 signifies inability to perform the specified task and 10 signifies ability to perform the task.3 I believe this response scale introduces a problem in that too many response options are available and thus misinterpretation may result. Another problem with the PEQ, as with all self -report measures, is that a 100% response rate is not guaranteed, and those that do respond most likely don't represent the general amputee population.
Prosthetic Profile of the Amputee
The PPA is a self-report measure whose purpose is to obtain information regarding the frequency of prosthesis use and wear. The PPA includes questions concerning predisposing, enabling and reinforcing factors that relate to frequency of prosthesis use and wear. Since we are interested in predicting prosthetic mobility, the Locomotor Capabilities Index (LCI), which is a subscale of the PPA, was used in the various studies.3 The PPA-LCI is similar to the PEQ in the sense that it measures the mobility potential of an amputee. The PPA-LCI contains 11 items that pertain to mobility tasks such as rising from a chair, climbing stairs or curbs and walking on uneven terrain.3 Similarly to the PEQ response scale, the LCI uses a scale ranging from 0 to 3. The corresponding responses range from ability to perform the task independently to inability to perform the task.3 The highest score possible is 33. The higher the score received, the higher the level of potential predicted. In general a possible fault of self-report measures is the fact that the respondent may be either over or under confident when assessing his/her capability of performing the specified tasks.
Barthel Index
The Barthel Index is a 10 item self-report that measures functional independence. With the Barthel Index, functional independence is measured through a person's perception of his/her ability to perform specific activities of daily living (ADLs). Such activities include bathing, feeding, mobility, toilet use and bladder control. The response scale ranges from 2 (independent) to 0 (with help). The highest possible score is 20; the higher the score, the higher the level of functional independence. Unlike the previous three outcome measures, the Barthel Index measures functional independence rather than mobility independence. Also the Barthel Index, similarly to the two following outcome measures to be discussed, is not designed specifically for the amputee population.
Goal Attainment Scaling
Goal Attainment Scaling (GAS) is another self-report measure whose purpose is to measure rehabilitative outcome on an individualized basis. For reliability and validity testing, exclusion criteria included individuals with cognitive impairment. On the other hand, persons who had a Symes or higher level of amputation and could speak, read and understand English were able to participate in the study.4 With GAS, the subject is able to set his/her own goals with the assistance of an investigator, and then rate the importance and difficulty of the specified goals.4 Different from scoring in the previous outcome measures, the GAS scoring included the subject rating the importance and difficulty of the goal, and the investigator rating the subject's ability to perform tasks having to do with the subject's specified goals. For rating the goals, responses varied from 1, "not important/difficult at all" to 5, "extremely important/difficult". The rating scale for ability to perform specified tasks varies from -2, "much less than expected" to +2, "much more than expected". Since the goals set are based on the subject, the scoring has to be standardized through the use of a mathematical formula.4 The major advantage of the GAS seems to be the individualized nature of the self-report.
Functional Independence Measure
The purpose of the Functional Independence Measure is to measure rehabilitative progress in persons who have many disabilities. The FIM was derived from the Barthel's Index, but it is more detailed in nature. The FIM consists of 18 tasks that have to do with self-care, mobility, sphincter control, locomotion, social cognition and communication.5 These skills are rated on a 4-point scale ranging from complete dependence to complete independence. The FIM score is determined by summing the scores received for each skill. The lowest possible score is 9, which signifies complete dependency. The highest possible score is 72, signifying complete independence. In the study analyzed the FIM was administered on both admission and final discharge. The outcome measure used in this study was the difference between FIM scores received upon admission and discharge. The FIM was also divided into an "amputation" and "all other skills" FIM sub score. The amputation FIM consisted of completing 8 of the 18 skills that dealt more with issues of ambulation such as transfers, climbing stairs, and bathing. The remaining skills that dealt with communication and social cognition were part of the "all other skills" FIM.
RESULTS
Amputee Mobility Predictor
In order for the AMP to be valid, scores from AMP must vary for subjects in differing K levels of the MFCL classification system.2 The AMP was able to differentiate between subjects in different K levels and thus it is valid tool to predict a patient's ability to ambulate with a prosthesis. Despite its ability to differentiate between MFCL levels, there was some overlap in the scores. This problem could be solved by using a larger sample size and a more representative sample of the amputee population.2 Unlike the general trend, the sample used in the AMP study contained more trauma related amputations than vascular related amputations.
In the AMP study, intra- and interrater reliability were tested. The results concluded that both intra- and interrater reliability were excellent for both the AMPPRO and AMPnoPRO.2 I believe the AMP is so reliable because the tasks to be performed are very straightforward along with the scoring system used. Since a person other than the patient determines the scoring, it is more likely to be impartial.
The AMP is designed to be performed in fifteen minutes or less, but I don't see this being true for patients in level K2 or lower. The major advantages of the AMP are:
it can be administered in a clinical setting and the results will help determine prosthetic prescription.
Prosthetic Evaluation Questionnaire and Prosthetic Profile of the Amputee
To test convergent validity, both the PEQ and PPA were compared with results from the 2 Minute Walk Test, the Timed Up and Go (TUG) test and the Activities-Specific Balance Confidence Scale (ABC).3 The 2 Minute Walk Test measures the distance the subject can walk at his/her own pace in the time frame of two minutes. This test is usually used to determine prosthetic progress. The TUG measures the performance of mobility skills such as balance, transfers and walking. For example, the time it takes for a subject to stand from a seated position may be recorded. The ABC scale is a self-report measure of how individuals perceive their balance confidence when performing tasks such as climbing stairs, getting in/out of a car or picking up an object from the floor. Based on previously set hypotheses, the PEQ and PPA are expected to distinguish amputation level (transfemoral vs. transtibial), amputation cause (vascular vs. nonvascular), reliance on an assistive device and other variables.3 Both the PEQ and PPA correlated with the 2 Minute Walk Test, TUG and ABC tests. The PEQ and PPA were able to distinguish between all groups with the exception of amputation level. A possible explanation of this result is that both the PEQ and PPA measures one's potential rather than actual ability to perform a specified task. For instance, a transfemoral subject may be able to traverse uneven terrain, but doesn't because of fear of falling. The PEQ was able to better distinguish specifically walking distance (unlimited vs. less than one block) and automatic walking. This may imply that the PEQ would be better suited for use as predicting the frequency that the prosthesis is used.3 This information would be very helpful when determining prosthetic prescription. The PPA was able to better distinguish specifically amputation cause and use of an assistive device. This information would also aid in determining prosthetic prescription.
Both the PEQ and PPA exceeded the minimum test-retest reliability standard.3 The minimum standard that was used is acceptable for comparisons between groups. The PPA had the higher reliability estimate, but it wasn't overly superior to the PEQ. The reliability of each item of the PEQ and PPQ were also tested. Out of the total items only one item from the PPA exceeded the accepted minimal standard for item test-retest.
The efficiency of self-report measures would depend on when it is administered. If the self-report measure is being used to assess prosthetic prescription, and is administered during initial evaluation, it would be very efficient. The advantages of giving the questionnaire while in the clinic are that you receive the results in no more than fifteen minutes, patient questions can be answered by the prosthetist and a higher response rate would be obtained. On the other hand, if the questionnaire is being used for the purpose of measuring prosthesis progression, most likely the questionnaire will have to be mailed. In this case a great deal of time is spent in preparation of the mailings and a lower response rate is likely.
Goal Attainment Scaling and Barthel Index
GAS was tested for reliability and construct validity. In terms of reliability, interrater reliability was determined based on GAS scores. The number and type of goals set were also analyzed. Out of a total of ten subjects, three set an additional number of goals. As expected, since the GAS measure is more individualized than other tests, the goals set by the subjects did vary. The interrater reliability of the GAS measure is described as "adequate to good".4 What may have contributed to a lower interrater reliability than what is ideally expected is the fact that there wasn't any training period for the investigators participating in the study.4 In addition, if help was needed when helping to set goals, the investigators may have been biased when choosing which goals were important in the rehabilitation process.
For testing construct validity, GAS was compared with the Barthel Index. Similarly to previously mentioned methods, construct validity was established by analyzing the strength and direction of a previously stated hypothesis. The hypothesis states that a subject's score would improve as a result of the subject's participation in the rehabilitation program.4 Construct validity positively correlated with the Barthel Index and was described as having "moderate" strength.4 The possible reasoning for only "moderate" strength is that the Barthel Index is a functional measurement tool whereas the GAS measures additional aspects of the rehabilitation process, such as caregiver stress and coping with pain. This makes it more difficult for the two tools to be compared. Unlike GAS, the Barthel Index was not as responsive to change. This was due to the high ceiling effect of the Barthel Index. On the other hand, the GAS was very responsive to change, probably resulting from the individualized nature of the measure.
Functional Independence Measure
Unlike the previous studies analyzed, the purpose of the FIM study is to determine if it is an effective prognostic indicator of prosthetic use in amputee subjects. The purpose of the studies involving other measurement tools is to determine their validity and reliability. In previous studies it has already been proven that the FIM is a reliable and valid tool, and that it has low interobserver variability. 6 The finding of the present study shows that the discharge FIM score was a better indicator of prosthetic use than the admission FIM score. In fact, the study showed that the admission FIM score did not predict prosthetic use at all.6 The study also revealed that level of amputation and the number of comorbidities play an important role in determining prosthetic outcome. In a second study using the FIM to predict rehabilitation outcome, it was found that the admission FIM predicted successful rehabilitation in higher functional individuals.5 The admission FIM score was also a predictor of improvement during hospitalization. From the above findings it can be concluded that the FIM has a ceiling effect. Higher functioning individuals are more likely going to achieve both high admission and discharge FIM scores, and thus their improvement is limited. In contrast, individuals who obtain a lower admission FIM score have more room for improvement than higher functioning individuals, thus contributing to the ceiling effect.
CONCLUSION
Until more research is done, it is difficult to say if and which measurement tool is superior. It can be concluded that different aspects of the various measurement tools could be combined to produce an even superior tool than each by itself. As displayed by the FIM, other factors besides mobility, such as level of amputation, will contribute to an individual's ability to function with a prosthesis.6 Identifying factors that affect prosthetic use by amputees is the purpose of a paper by Christiane Gauthier-Gagnon, MSc, Marie-Claude Grise, MSc and Diane Potvin, MSc. Their article focuses on people with unilateral transtibial and transfemoral amputations. The three dependent variables studied were frequency of prosthetic wear and active use of the prosthesis either indoors or outdoors. The three categories of factors that have been identified to affect prosthetic use are called predisposing, enabling and reinforcing factors.7 The enabling factors include locomotor skills and resources while the reinforcing factors have to do with social, physical and environmental factors. The predisposing factors analyzed in the article are physical condition, demographic profile, rehabilitation program and motivation.
Physical condition involves the cause and level of amputation, concurrent medical conditions, and the conditions of both the residual limb and nonamputated leg. It was found that the level of amputation and one's concurrent condition are significant factors of prosthetic wear while the condition of the nonamputated leg is a significant factor to both prosthetic wear and active use of the prosthesis.7 A larger percentage of transfemoral amputees discarded their prosthesis within five years of their discharge from a rehabilitation program than transtibial amputees. It was found that transfemoral amputees walked poorly compared with transtibial amputees due to the fact that energy expenditure when walking increases with the level of amputation.7 Increased energy expenditure also limits prosthesis wear for amputees with cardiac and respiratory problems.7 These problems result in less efficient cardiovascular and pulmonary systems, which are needed when compensating for increased energy expenditure that is due to prosthetic use. Limitations on prosthesis wear were related to vascular and arthritic problems in the nonamputated leg while muscle cramps and sores on the nonamputated leg affected the active use of the prosthesis.7 In order to reduce these problems, stresses on the nonamputated leg have to be reduced, and thus wearing and using the prosthesis have to be limited.
Demographic profile considers the age, income and educational background of the amputee. For those persons who actively wear their prostheses, it was found that as age increases the active use of the prosthesis for ambulation indoors decreased.7 It was also found that people who live alone tend to wear their prosthesis for longer periods of time than those who live in nursing homes.
The rehabilitation program actually refers to delays in prosthesis fitting after amputation and the duration of pre-prosthetic and prosthetic training. The general trend is that a longer delay between amputation and fitting, and longer pre-prosthetic and prosthetic training is associated with a decrease in prosthetic wear and active use.7 Surprisingly longer pre-prosthetic and prosthetic training doesn't necessarily result in more frequent prosthesis use. Delays in fitting and the duration of training will sometimes vary depending on the patient's condition. Having to deal with chronic illnesses and delays in wound healing are a couple of the factors that affect time delays.
Motivation deals with the amputee's adaptation to the amputation and the prosthesis. Results show that adaptation to the amputation and prosthesis correlate positively to prosthesis wear and active use of the prosthesis indoors and outdoors.7 About 57% of nonusers (of prosthesis) did not feel adapted to the prosthesis and 60.2% of users felt adapted to their prosthesis. Acceptance of a prosthesis was shown to be a factor in the time delay between amputation and fitting and the duration of pre-prosthetic training.7 The attitude of the patient is an important factor in predicting prosthetic wear and use.
In conclusion the factors that would be good predictors of prosthesis wear are level of amputation, absence of respiratory problems and adaptation to the prosthesis.7 The factors that would be good predictors of active use of the prosthesis indoors are short pre-prosthetic training periods and adaptation to the prosthesis.7 For active use outdoors good predictors are place of abode, presence of muscle cramps, delays in prosthetic fitting and duration of prosthetic training.7 With this added knowledge it is realized that many other factors beside mobility alone are needed to help predict how one will function with a prosthesis. Therefore different aspects of the various measurement tools can be combined to create a more complete measurement tool. For example, the mobility aspects of the AMP in addition to the social and emotional aspects of the Barthel Index could be used to aid in predicting prosthetic prescription for amputee patients.
References
Costello, Robert B. Random House Webster's College Dictionary. New York: Random House Inc, 1992. Gailey RS, Roach KE, Applegate B. The Amputee Mobility Predictor: An instrument to Assess Determinants of the Lower-Limb Amputee's Ability to Ambulate. Arch Phys Med Rehab 2002;83:613-26. Miller WC, Deathe AB, Speechley M. Lower Extremity Prosthetic Mobility: A Comparison of 3 Self-Report Scales. Arch Phys Med Rehab 2001;82:1432-39 Rushton PW, Miller WC. Goal Attainment Scaling in the Rehabilitation of Patients With Lower Extremity Amputations: A Pilot Study. Phys Med Rehab 2002;83:771-75. Muecke L, Shekar S, Dwyer D. Functional Screening of Lower-Limb Amputees: A Role in predicting Rehabilitation Outcome? Phys Med Rehab 1992;73:851-58. Leung EC, Rush PJ, Devlin M. Predicting Prosthetic Rehabilitation Outcome in Lower-Limb Amputee Patients With the Functional Independence Measure. Phys Med Rehab 2002;83:613-26. Gauthier-Gagnon C, Grise MC, Potvin D. Predisposing Factors Related to Prosthetic Use by People with a Transtibial and Transfemoral Amputation. JPO 1998;10:99-109
|
|
