| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Research Reports |
KD Henry, PT, is Physical Therapist II, Physical Medicine and Rehabilitation Department, The Johns Hopkins Hospital, Osler 159, 600 N Wolfe St, Baltimore, MD 21287.
C Rosemond, PT, GCS, is Clinical Assistant Professor, Division of Physical Therapy, The University of North Carolina at Chapel Hill, Chapel Hill, NC
LB Eckert, PT, is a graduate of the University of North Carolina at Chapel Hill
Address all correspondence to Mrs Henry
Submitted November 25, 1996;
Accepted October 13, 1998
 | Abstract |
|---|
=72.8), who were living independently in their communities. Methods. Subjects were randomly prescribed 2, 5, or 8 general strengthening home exercises. They were instructed on their exercises at an initial session and asked to record the number of repetitions performed each day in a self-report exercise log. At a return session 7 to 10 days later, subjects were scored on their performance of the prescribed exercises using a newly designed assessment tool. Results. The group that was prescribed 2 exercises performed better, as defined by their performance tool score, than the group that was prescribed 8 exercises. The group that was prescribed 5 exercises was not different from the groups that performed 2 or 8 exercises. No differences were found among groups regarding the self-report measurement of compliance. There was a moderate correlation between performance scores and the self-report percentage rates. Conclusion and Discussion. Subjects who were prescribed 2 exercises performed better than subjects who were prescribed 8 exercises. The question of an optimal number of exercises to prescribe to elderly people warrants further study.
Key Words: Compliance • Exercises • Older adult • Optimal performance
| Introduction |
|---|
 TopAbstract Introduction MethodResultsDiscussionConclusionsAppendixReferences |
|---|
The exercises prescribed for self-care often depend on the number of areas or muscles involved, the interval between clinic visits, and the types of exercises needed. Many of the older adults seen by physical therapists have cognitive losses and complex medical problems preventing them from completing all of the exercises prescribed. For many individuals, a sedentary lifestyle has been the norm. Factors such as these may influence an older person's willingness and ability to carry out a home exercise program with a large number of exercises. By the year 2000, 20% of the population will be over 65 years of age.1 More physical therapists will be working with this population and need to understand their patients' unique attributes in order to provide meaningful treatments.
The terms "compliance" and "adherence" are often incorrectly used interchangeably in reference to exercise and medication prescription. Many of the existing studies involving exercise use the term "compliance" to mean performing the exercises as directed by the physical therapist in their mode, type, duration, and frequency.2 According to Guccione, the term "compliance" "...implies that the patient must do as the therapist instructs in order for the patient education session to be successful."3(p336) Many authors, however, prefer the term "adherence" because of its implication of freedom of choice and action on the part of the patient.3 The term "adherence" often is used to describe patients adhering to their medication prescription regimen.4 We chose to use the term "compliance" in our study because the traditional sense of that term more closely matched the artificial setting of the study.
Motivation is needed for adherence to a regimen and is defined as a "complex attitude of wants, beliefs, and rewards versus cost of the behavior."3(p336) The participants in our study did not have a specific deficit at the impairment level and, therefore, did not independently search out physical therapy exercise prescriptions. They were all volunteers who were offered a chance (by their internal medicine physician) to participate in a study that involved general strengthening exercises. The participants may have been motivated to volunteer for a variety of reasons, including to improve their general strength, to participate in a local university study for altruistic reasons, or to help their physician in any manner they could. Even though volunteering implies a willingness to participate, we decided to use the term "compliance" in this report for 2 main reasons. First, the participants were asked to obey unquestionably the exercise instructions and perform them in a specific manner. Second, the motivational and independent problem-solving aspects were not assessed, and, therefore, adherence was not fully studied.
The participants were evaluated on whether they completed the exercises in the requested manner and at the requested frequency and, therefore, on whether they complied with the exercise prescription. In order to perform the exercises accurately for the evaluation, the patients needed to learn the exercises. Guccione defined learning as an "acquisition of knowledge or skills achieved by study, instruction, practice, and experience."3(p340) As described in this report, the participants used all of these methods to learn the exercises and perform them accurately. Learning is a required part of compliance.
Numerous attempts have been made to isolate factors that influence compliance. A literature review by Sluijs et al2 revealed over 200 factors that may be related to patient compliance. These factors include sociodemographic variables (eg, age, sex, education), accessibility of setting, patient attitude, and type of regimen.5–12 Lorenc and Branthwaite13 used a person's living condition (ie, whether living alone or with someone) as the main demographic variable related to compliance. People over 65 years of age who live alone are less compliant.13 According to Mayo10 and Shephard et al,12 problems with compliance increase in individuals over 65 years of age. Mayo's explanation is that exercise regimens are more difficult for medical professionals to prescribe to people in this age group because of their impaired physical status.10 Sluijs et al2 also reported that older patients exhibit decreased compliance with exercise regimens because of multiple pathologies.
Accessibility to a treatment setting and patient attitudes can become critical barriers to exercise compliance. According to Martin,14 compliance decreases when individuals exercise at inconvenient locations (eg, fitness gym located across town). In contrast, Thompson et al15 reported that compliance is enhanced with home-based exercise programs. Some authors2,10 reported that noncompliance increases when clinic hours require the patient's lifestyle to be disrupted. The time and psychological investment and internalization of the exercises can affect the patient's willingness to comply. Decreased compliance is associated with preventative exercises, as compared with therapeutic exercises.2 This relationship may be attributable to direct results not being seen by the patient with preventative exercises.
Some aspects of prescribed regimens, such as the complexity, intensity, and meaning of the exercise program, have an effect on patient compliance. Haynes et al6 reported that treatments requiring more than one step or item task produced a higher rate of noncompliance. Drug studies dominate as a major type of research on complexity of prescriptions. Haynes et al6 described a study in which compliance decreased when more than 3 prescribed medications were taken daily. Sackett and Haynes16 stated that compliance with behavioral treatments (eg, an exercise program) is similar to drug compliance; as complexity increases, compliance decreases. Additionally, patients are less compliant if the exercise programs are not seen by the patients as meaningful or if the programs not individually tailored to their situation or routine.2,17
A major problem for health care professionals is measuring compliance with the medical regimen. Sluijs et al2 stated that physical therapists lack a measurement tool for reliably assessing patient compliance with exercises. Some methods of measuring compliance for drug studies are pill counts and blood traces.10 Exercise studies, by convention, rely on patient self-report. Self-report may include the use of a questionnaire, a patient diary (eg, log sheet), or an interview. Massie and Shephard9 and Gerber and Nehemkis18 showed that the degree of compliance estimated by self-report exceeded the degree of compliance estimated by other methods (ie, learning contracts and checklists). Although selfreport methods may tend to overestimate compliance, they are still the most cost-effective, feasible, and frequently used methods available in assessing compliance with an exercise regimen.6,8,11,19
Observation of exercise performance in the clinic is another way of measuring compliance. Observation in the clinical setting is meaningful only if it is assumed that the structured situation of the clinic is similar to the home environment of the older person.6 Few studies on exercise regimens have used direct observation as a method of measuring compliance.10 Friedrich et al20 used a 3-grade scale when observing patients performing their exercises after instruction from a brochure or a therapist. Their grading scale and definition of compliance were related to whether the goals were met and whether there were detrimental effects from the exercises being performed incorrectly.20 There are no scales for measuring compliance based on motor control, alignment, and amount of cueing during demonstration of the exercises. Scales of this type would provide more kinesiological information and potentially more reliable measurements. Mayo10 stated that evaluating the degree of compliance with direct observation would be difficult, but some measure of quality could be made. By using composite indices of self report and direct observation, more accurate portrayals of patient compliance with exercise prescription may be possible.10 In our study, we attempted to use self-report and demonstration to measure compliance.
The purpose of our study was to determine whether adults over 65 years of age perform and comply better with 2, 5, or 8 home exercises. Three hypotheses were formulated prior to this study: (1) Subjects who are prescribed 2 exercises will perform better than subjects who are prescribed 8 exercises, (2) subjects who are prescribed 2 exercises will comply on their self-report exercise log more than subjects who are prescribed 8 exercises, and (3) self-report percentage rates will highly correlate with performance assessment tool scores.
| Method |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionsAppendixReferences |
|---|
72.8). People were excluded from the study if they were under 65 years of age; had difficulty following directions; were currently receiving physical therapy; or had heart, lung, neurological, or orthopedic problems that prevented safe performance of general exercises. The Fearrington Village residents were the subjects who were participating in an exercise program 3 times per week. Their generalized exercise program consisted of a 1.6-km (1-mile) walk around the pool and light calisthenics. Each subject provided informed consent to be a volunteer.
Materials
Exercise packets were prepared for each subject. Each packet contained a list of 2, 5, or 8 exercises (Tab. 1), a written and pictorial description of each exercise, and a self-report exercise log. The exercises were selected by an experienced practicing physical therapist (CR, with 10 years of experience practicing physical therapy) for their functional relevance, simplicity, and commonality as general strengthening exercises. An alternate exercise (ie, shoulder flexion, corner push-ups, straight leg raises, shoulder shrugs, or hip extension) was used if the any of the original exercises were not safe for particular volunteers to perform. A chair with arm rails and a supportive back, an exercise mat, and a towel roll were made available to each subject.
|
Reliability
Interrater reliability for measurements obtained with the assessment tool was established using the following approach. Five physical therapist students ranging in age from 21 to 26 years from the University of North Carolina at Chapel Hill were instructed on knee extension and scapular pinch in a sitting position. The range in age was acceptable, in our opinion, because there is no defined and expected performance level for people over 65 years of age. The students were scored on their performance by 3 raters using the assessment tool. Between the raters' scores of the participants, there was one 1-point difference and one 2-point difference. The performance scores of the participants ranged from 3 to 12, indicating that the raters reliably scored all levels of the participants' performance. Two of the raters in the interrater reliability study were the same raters as those in the actual study. The one additional rater in the reliability study was not part of the actual study. One exercise instructor (nonrater) described the study to each participant in both the reliability study and the actual study. The instructor read from a script for consistency. Once interrater reliability was established as high, as defined by Munro et al,21 data collection began.
An intraclass correlation coefficient (ICC[2,1]) model was used to analyze the interrater reliability data. The ICC analyzed 3 raters' results on the performance scores of 5 students, each demonstrating 2 exercises. The ICCs for the interrater reliability test represent a comparison of the 3 raters using the performance assessment tool. The ICCs (2,1) for exercise 1 (knee extension in a sitting position) and exercise 2 (scapular pinch in a sitting position) were .87 and .93, respectively. Thus, the data obtained for these 2 exercises were used for subsequent data analysis.
Procedure
Subjects participated in 2 sessions that took place at exercise facilities that were equally accessible and convenient to the subjects. The first session was for exercise instruction. The return session was for subject demonstration of the assigned exercises, with a rater scoring each exercise using the performance assessment tool.
During the initial session, each subject was randomly assigned to 1 of 3 groups: subjects who were prescribed 2 exercises, subjects who were prescribed 5 exercises, and subjects who were prescribed 8 exercises. Exercise packets were prepared prior to the instructional session. The instructor randomly selected an exercise packet for each subject and in this way determined the subject's group assignment. Subjects within a group were instructed on their exercises in a fixed order (Tab. 1). Range of motion was assessed using Kendall and colleagues' description of "within normal limits."22 If a subject, in the view of the instructor, was limited by pain or an orthopedic condition, the instructor selected an alternate exercise that avoided the area of concern. Two individuals received instruction in an alternate exercise. Each subject was allotted 20 minutes maximum to learn the exercises and understand how to use the self-report exercise log. The subjects were asked to perform 10 repetitions of each exercise, once each day, until the return session 7 to 10 days later. The subjects were instructed to record information about their exercise sessions using the self-report exercise log, especially noting any pain with the exercises or any reason for noncompliance. Following the exercise instruction, a rater, who was not the same person as the instructor and was one of the reliability study raters, used the performance assessment tool to score each exercise. The subjects were scored on correctness of alignment and exercise quality of movement (Appendix). The cueing component was eliminated from the scoring because instruction for correct technique was considered as "cueing" and, therefore, required by all participants. The rater ensured that the subjects learned the exercises by demonstrating the correct alignment and quality of movement with a score of 8 out of 8 on the performance assessment tool (Appendix).
During the return session, we collected and reviewed the self-report exercise logs for completeness and reasons for omissions. The original instructor asked all subjects to "perform the exercises as you have been doing them at home." All subjects sat in the same chair in which they were instructed, and the same towel roll was in sight and available for the subjects to use. If subjects needed help remembering the next exercise or how to perform the exercises accurately, then appropriate cueing was provided by the instructor. A rater, who was not the instructor or the rater from the first session, used the performance assessment tool to score performance as the subject demonstrated each exercise. That rater was blind to the group assignment (2, 5, or 8 exercises) of the subjects. The rater, however, knew how to perform all of the exercises accurately. The rater continued to score the exercises until the participant stopped exercising. The participants' stopping point defined the completion of the prescribed exercises. None of the participants stopped their demonstration before completing all of their exercises. A total performance score for each exercise was tallied. After completing the exercises, all subjects were instructed about the purpose and rationale of the study and were permitted to ask any questions. The purpose of measuring compliance was not explained to the subjects before the study so as to not bias the participants' typical compliance with the exercise programs.
Data Analysis
The performance assessment scores were computed for the first 2 exercises for each subject across the exercise groups, and these 2 exercises were used for subsequent analysis. The first 2 exercises were analyzed because they were common to all groups. In addition, by analyzing the same 2 exercises, we attempted to ensure that the exercises were of the same level of difficulty and meaning to the participants. The self-report exercise logs were analyzed by a statistician (Dr William Ware) for completion and omissions from the first 7 days, as all participants performed exercises at least 7 days. A check mark on the exercise log for completion of 10 repetitions of one exercise on one day, for example, indicated compliance with that exercise for that particular day. Review of the 15 participants' exercise logs revealed that, if there was an omission of a check mark on one particular day for an exercise, then all exercises for that day also were omitted. The self-report percentage rates for all subjects were computed. The medians of the performance scores and self-report percentage rates were computed for each group (2, 5, or 8 exercises) and are listed in Table 2. The performance scores and self-report percentage rates across all exercises for each subject were not computed. The performance scores for exercises 1 and 2 and the self-report percentage rates were analyzed for exercise groups using the Kruskal-Wallis one-way analysis of variance (ANOVA).23 Differences were determined using Dunn's post hoc multiple comparison test.24 Finally, a correlation analysis was used to determine whether a correlation existed between the performance scores and self-report percentage rates.
|
| Results |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionsAppendixReferences |
|---|
Table 2 shows the group medians for the self-report percentage rates. The self-report percentage rate was based on the number of exercises performed per total number prescribed. The Kruskal-Wallis one-way ANOVA results indicated no differences among groups with 2, 5, or 8 exercises.23 A correlation analysis of the exercise compliance, as measured by an exercise log, and of performance, as measured by direct observation, yielded a correlation coefficient of .54.
| Discussion |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionsAppendixReferences |
|---|
The reliability of the performance assessment scores obtained for exercises 1 and 2 may not indicate that all performance scores of other exercises have similarly high reliability. Future studies on the reliability of scores for this performance assessment tool obtained for multiple exercises would allow therapists to more fully understand the differences in each exercise and whether they all are evaluated equally. Another limitation to the reliability study is that students ranging in age from 21 to 26 years were used from a convenience sample instead of people over 65 years of age. Often, people over 65 years of age demonstrate different performance abilities than people under 65 years of age. There is, however, no defined performance level according to age. We believe that the use of people with varying ages is acceptable for the reliability study, but it is a limitation.
Hypothesis 1 was supported because we found a difference in performance scores was among groups. An apparent tendency was that the median performance score of each exercise group decreased as the number of exercises increased (Tab. 2), and a difference occurred between the group that was prescribed 2 exercises and the group that was prescribed 8 exercises. The data suggest that performance may decrease when subjects are prescribed 5 exercises, but further study is needed. A few limitations regarding the analysis that led to the support for hypothesis 1 are discussed. First, there were 5 participants in each group, and a larger sample size may provide more information about the effect of various numbers of prescribed exercises. Second, 20 minutes was allotted for the instruction session for all of the exercises. In future research, perhaps each exercise should be allotted a set amount of time instead of a time for the whole exercise session. Third, the analysis of the exercises may be more informative if all of the exercises are evaluated and not just the first 2 exercises. Fourth, some participants were exercising previously and other participants were not. With these limitations, this study did not determine an optimal number of exercises to prescribe. It supports the idea, however, that subjects who are prescribed 2 exercises perform better (as indicated by higher scores on exercise quality of movement, alignment throughout the exercise, and cueing requirement) than subjects who are prescribed 8 exercises. If we can assume that higher performance scores are secondary to increased practice and thus increased compliance with an exercise program, then this assumption supports use of the performance assessment tool as a measure of compliance. In future studies, the performance assessment tool may be used in the clinic.
A moderate correlation was found between the performance assessment tool scores and the self-report percentage rates, making the correlation positive. Two possible reasons for the correlation are (1) the participants placed check marks on their self-report exercise log even when they did not perform the exercises and, therefore, did not practice the exercises and (2) the assessment tool was not sensitive enough to detect changes in performance when participants complied completely. In our view, the correlation provides evidence that the performance assessment tool is similar to the self-report in indicating compliance. Both scores, however, demonstrate high compliance rates.
There were, in our opinion, 4 factors present in our study that could positively influence compliance and performance. The first factor was that the subjects' overall level of fitness and attitude toward exercise may have been more favorable than for the average adult over 65 years of age. The majority of the subjects were recruited from an exercise class that met 3 times a week throughout the year. The median scores for each of the 3 exercise groups (2, 5, and 8 exercises) on the measurement tool were 11.5, 11, and 10.25, respectively. These scores suggest that the subjects' level of performance was good overall. Second, subjects volunteered for this study, which was advertised through their physician and exercise instructor. If the subjects had no intention of doing the exercises, they probably would not have volunteered. Often in physical therapy clinical settings, patients are receiving treatment not because they have volunteered, but because it was prescribed by their physician. Third, all the subjects in this study were from above-average socioeconomic settings. Some of the literature suggest that this factor may positively influence compliance.2 Fourth, the exercise regimen was short-term and had a definite ending date. Long-term exercise programs are more difficult to maintain and would require a longitudinal study to monitor compliance.2
One factor that we believe might have negatively influenced compliance was that the exercises prescribed were general strengthening exercises, which are considered preventative exercises rather than curative exercises. According to Sluijs et al,2 preventative exercises usually have less meaning to the patient than curative exercises, and people who consider their exercises as meaningful are more compliant. Second, the performance assessment tool may not be sensitive enough for all of the patients with various noncontributing past medical histories. A more sensitive scale may yield lower scores. Future studies with larger samples of subjects from differing socioeconomic groups, a variety of fitness levels, and more narrow age ranges would be beneficial. Different types of exercises could be examined, such as a cardiovascular program incorporating walking and bicycling. Finally, performing this type of study in a clinical environment may produce more generalizable results, because that is where physical therapists typically prescribe exercises.
| Conclusions |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionsAppendixReferences |
|---|
| Appendix |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionsAppendixReferences |
|---|
|
| Acknowledgments |
|---|
| Footnotes |
|---|
This study was approved by the Committee on the Protection of the Rights of Human Subjects at the University of North Carolina at Chapel Hill (approval #93-MAHD-287).
| References |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionsAppendixReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
