Home Exercise Program Compliance of Service Members in the Deployed Environment: An Observational Cohort Study

Transcription

1 MILITARY MEDICINE, 180, 2:186, 2015 Home Exercise Program Compliance of Service Members in the Deployed Environment: An Observational Cohort Study CPT Timothy Eckard, SP USA*; CPT Joseph Lopez, SP USA ; CPT Anna Kaus, SP USA ; James Aden, PhD ABSTRACT Background: Home exercise programs (HEP) are an integral part of any physical therapy treatment plan, but are especially important in theater. The primary aim of this study was to determine if the number of exercises prescribed in a HEP was associated with compliance rate of Service Members (SM) in theater with a secondary aim of determining variables associated with compliance and noncompliance. Materials/Methods: Subjects were 155 deployed SM undergoing physical therapy in Iraq and Afghanistan. Clinical evaluation and prescription of a HEP were performed. Pathologic, demographic, and treatment data were obtained. Subjects returned to the clinic 1 week later to demonstrate their HEP. Subjects performance of each prescribed exercise was rated on a 12-point scale to quantify compliance. Results: 2 variables were found to be significantly associated with rate of compliance. These were the number of exercises prescribed ( p = 0.02) and if a subject left the base at least once per week ( p = 0.01). Conclusions: SM prescribed 4 or more exercises had a lower rate of compliance than those prescribed 2 or fewer. SM who left the base at least once per week also had a lower rate of compliance. INTRODUCTION Musculoskeletal injuries are a significant contributor to lost manpower in deployed military units and are one of the leading reasons military Service Members (SM) seek medical care. 1,2 As a means of mitigating the impact this has on unit readiness and combat effectiveness, physical therapists are deployed in combat support hospitals, brigade combat teams, and special operations units. 3 Deployed SM requiring physical therapy often have duties that do not allow for regular attendance of in-clinic rehabilitation sessions. This unique challenge makes military therapists treatment plans heavily reliant on home exercise programs (HEP), as they are often the only source of regular treatment a SM receives. It is well established in the literature that regular performance of therapeutic exercise has a significant impact on pain and function levels in patients with musculoskeletal conditions. 4,5 Several studies have found HEP compliance rates to be near or below 50% of prescribed levels. 5 8 When *Physical Therapy Department, U.S. Army Health Center, Vicenza, Italy, Unit Box 13 APO AE Physical Therapy Department, Blanchfield Army Community Hospital, 650 Joel Drive, Fort Campbell, KY Physical Therapy Department, Brooke Army Medical Center, 3551 Roger Brooke Drive, Fort Sam Houston, TX Department of Statistics and Epidemiology, Institute of Surgical Research, 3551 Roger Brooke Drive, Fort Sam Houston, TX This study protocol was approved by the U.S. Army Medical Research and Materiel Command Office of Research Protections Institutional Review Board at Fort Detrick, Maryland. The views expressed in this manuscript are those of the authors and do not necessarily reflect official policies or positions of the Department of Defense, the U.S. Army, the U.S. Army Medical Command, or the respective institutions to which the authors belong under the Army Medical Command. Upon completion of this study, the name was changed from Home Exercise Program Compliance of Service Members in the Deployed Environment to Home Exercise Program Compliance of Service Members in the Deployed Environment: an Observational Cohort Study. doi: /MILMED-D surveyed, physical therapists have estimated patient compliance to be 64% with short-term and 23% with long-term exercise prescriptions. 8 Factors relating to the patient, pathology, types of exercises prescribed, and patient provider interactions have all been found to be associated with HEP compliance. 5,6,9,10 In this body of literature, the authors found only 1 study examining a military population. Coppack et al 11 found that goal setting led to greater exercise adherence in UK military personnel rehabilitating from nonspecific chronic low back pain. No studies were found examining compliance in deployed military SM. Conducting research on compliance has proven difficult. One barrier to conducting high-quality studies has been a lack of agreement on what it means to be compliant. Meichenbaum and Turk, 12 in their seminal work on the topic, defined compliance as abiding obediently by the practitioner s prescribed treatment protocol. However, many authors have used differing definitions of compliance and despite representing disparate constructs, the terms compliance and adherence are often used interchangeably in the literature. 9 Another difficulty in defining compliance stems from the multifaceted nature of the concept. Overall compliance can be considered a product of frequency, duration, accuracy, or quality of performance of each exercise, or some combination thereof. In a study examining patients with neck and low back pain, Medina et al 5 demonstrated that there were different predictive factors for each facet of compliance. Accurately quantifying exercise compliance has also proven difficult. The Sport Injury Rehabilitation Adherence Scale has been demonstrated to be a valid way of quantifying therapeutic exercise adherence. 13 However, all questions on this scale refer to in-clinic exercise sessions and thus are of no use when examining HEP compliance. In a study examining the effect of number of exercises prescribed on the quality of performance, Henry et al 7 developed the Henry-Eckert 186 MILITARY MEDICINE, Vol. 180, February 2015

2 Performance Assessment Tool (PAT), a 12-point scale in which cueing, alignment, and exercise quality are quantified for each exercise on a scale of 1 to 4. Although this scale has not been validated, it has been found to have good inter-rater reliability in the form of an intraclass correlation coefficient ranging from 0.87 to 0.93 for 3 graders of 2 exercises. Additionally, in their study, Henry et al 7 found that PAT score was moderately correlated (r = 0.54) with self-report and therefore had the potential to be used to assess HEP compliance. 7 The authors elected to use this scale along with a simple 1-question self-report questionnaire as 2 ways to quantify HEP compliance. They also elected to score any exercise that was prescribed but not performed at follow-up as a 0 on the PAT to properly weigh the lack of performance of that exercise into the average score. Determining the association between the number of exercises prescribed and the rate of HEP compliance may help deployed therapists choose the optimum number of therapeutic exercises to prescribe to maximize both patient compliance and treatment efficacy. Additionally, identifying key demographic, pathologic, and treatment variables associated with compliance may help therapists to better tailor HEP to individual patients for improved outcomes. The primary objective of this investigation was to determine if rate of HEP compliance differed significantly between SM prescribed 1 or 2 therapeutic exercises from those prescribed 3 or more. Our hypothesis was SM prescribed 1 or 2 exercises for a HEP would demonstrate a significantly higher rate of compliance than those prescribed 3 or more. A secondary objective was to determine demographic, pathologic, and treatment variables associated with compliance and noncompliance. METHODS Research Design Prospective observational cohort study. Subject Population Consecutive SM presenting to the physical therapy clinics in Iraq and Afghanistan in the locations to which the first, second, and third authors were deployed in Inclusion Criteria (1) At least 18 years of age. (2) Able to read, write, and speak sufficient English to be able complete the outcome measures. (3) U.S. military SM presenting to the physical therapy clinic for evaluation and treatment of a neuromusculoskeletal complaint of the shoulder, spine, hip, knee, ankle, or foot region. (4) Chief complaint of pain and/or dysfunction of a neuromusculoskeletal nature of the shoulder, spine, hip, knee, ankle, or foot, diagnosed by a licensed physical therapist. Exclusion Criteria (1) Chief complaint of elbow, forearm, wrist, or hand symptoms determined to be local and not primarily originating from one of the regions listed above. (2) Pathologies requiring interventions from other services (i.e., upper motor neuron disorders, systemic disorders). (3) Symptoms in the region of chief compliant determined to be of a nonneuromusculoskeletal nature. (4) Inability of a patient to follow-up between 4 and 14 days in the physical therapy clinic. (5) Diagnosis of concussion or traumatic brain injury within 8 weeks before the date of recruitment. Recruitment Process Standard physical therapy evaluation of the region of chief complaint to include past medical history, history of chief complaint, red flag screening, observation, range of motion, manual muscle testing, and special tests were performed to determine diagnosis. Potential subjects were instructed in, and demonstrated correctly, their prescribed exercises and provided an opportunity to ask questions before recruitment. All components of the evaluation and treatment met the standard of care and were performed by a licensed physical therapist. Consecutive patients were asked if they would like to participate in the study at the conclusion of their initial evaluation. Patients were assured that no change would be made to their plan of care if they declined enrollment. No attempt to recruit a patient was made after a patient s first refusal. All subjects signed an informed consent form and Health Insurance Portability and Accountability Act authorization statement. Subject rights were protected at all times. No compensation was provided to participants. Study Procedures Immediately following enrollment, initial data collection was performed. Data collected included gender, age, rank, Army Military Occupational Specialty (MOS) series (combat arms or noncombat arms with combat arms referring to the infantry, artillery, special operations, and cavalry series) referral source, last Army Physical Fitness Test score, hours spent exercising per week, if current duty hours were predominantly day or night in the past month, if the majority of their exercise was during unit physical training, and whether or not the subject left the perimeter fence of the base at least once a week. The job series of any Navy and Air Force personnel were converted to their Army MOS equivalents. Clinical data collected included region of chief complaint, number and type (mobility, strength, or proprioception) of exercises prescribed, Numeric Pain Rating Scale at rest and at worst in the past 48 hours on a 0 to 10 point scale, the Epworth Sleepiness Scale, and the Global Rating of Change (GROC) questionnaire. All questionnaires were completed by the subjects. All other data were requested verbally by the treating therapist and written on a data collection form to ensure completion. MILITARY MEDICINE, Vol. 180, February

3 Subjects were then scheduled a follow-up appointment as close to 7 days as possible based on subject availability. A range of 4 to 14 days was considered acceptable for follow-up. Other physical therapy appointments such as modality or exercise performance sessions were scheduled if warranted by the condition and therapist s treatment program. At follow-up, subjects completed a survey asking them to answer the question How accurately, consistently, and thoroughly do you feel you have performed this treatment program since your last appointment? Subjects were asked to circle one of the following responses: Not at all, Somewhat, Mostly, or Completely. Subjects were informed that their physical therapist would be blinded to their answer to encourage accuracy in reporting compliance. Blinding was achieved through the utilization of a file folder containing all patient data from the follow-up appointment that was never accessed by a subject s treating provider and was input into the spreadsheet at the conclusion of subject participation in the study by a different investigator, thus blinding a subject s provider from knowledge of that subject s self-reported compliance. Subjects also completed a GROC questionnaire at follow-up. Subjects then demonstrated what they knew of their prescribed HEP for an investigator who was not their treating therapist. The PAT was used to quantify performance of each exercise demonstrated by the subject. The investigator asked the subject to perform each exercise they were prescribed for 30 seconds, scoring them on each subscale of the PAT. During exercise performance, patients had access to a high-low treatment table, chair, foam roll, towel, and resistance bands all displayed on a table in front of them for use in demonstrating their HEP. After exercise demonstration, subject participation in the study was terminated and the patient had a follow-up appointment with their treating physical therapist to continue their care. All subjects were prescribed a HEP and none were altered during the course of the study. Subjects were contacted if they failed to attend their scheduled follow-up appointment to encourage them to come to the clinic as soon as possible. Attempts to contact a subject ceased 14 days after the date of enrollment. Only data from complete data sets were used in analysis. Sample Size This study included 100 subjects enrolled in Iraq and 90 in Afghanistan. Complete data sets were collected on 155 subjects, with 35 lost to follow-up. Recruitment was stopped upon redeployment of the authors. Statistical Analysis Statistical analysis was performed using SAS version 9.2 (SAS Institute Inc., Cary, North Carolina). Average PAT scores were calculated as the mean PAT score of a subject. Scores ranged from 0 to 12. Any exercise prescribed that was not performed by a patient was assigned a score of 0. A subject was deemed compliant based on the PAT if they had a mean score of 10 or higher. PAT average scores were also compared to the self-evaluation questionnaire in which subjects were asked if they were not at all, somewhat, mostly, or completely compliant with exercises. This was treated as a 0 to 3 ordinal scale for statistical purposes. Kappa statistics with 95% confidence intervals were calculated for average PAT scores of 10 or higher against those subjects who answered mostly or completely on the questionnaire. Demographics and other factors were compared between those with an average PAT score of 10 or higher. Factors that were continuous were compared using a Kruskal Wallis nonparametric test and were summarized using the medians along with the interquartile range. Categorical variables were compared using the c 2 test and summarized using percentages. A test was deemed statistically significant if the p value was less than RESULTS Overall agreement between the more objective PAT and the self-reported compliance score was low. Comparing those with average PAT scores of 10 or higher to a compliance questionnaire response of mostly or completely fared poorly with an overall agreement of 45% and a kappa of 0.02 ( 0.10, 0.14). Because of this poor agreement, selfreported compliance was not used in any further analysis. Several key demographic, pathologic, and treatment factors were collected for the subjects enrolled in the study. Table I shows the results for these comparisons on all continuous variables. The only 2 factors that were significantly associated with a subject s compliance were the number of exercises he or she was asked to perform ( p = 0.02) and whether he or she left the confines of the base at least once per week ( p = 0.01). Evaluating this further, a clear trend emerged. Of those subjects prescribed 2 or fewer exercises, 86% were found to be compliant. In contrast, only 54% of those prescribed 4 or more exercises were found to be compliant. Figure 1 shows this result along with the corresponding c 2 p value of Even though total number of exercises was significantly associated with rate of compliance, it did not seem to matter which types of exercises a subject was prescribed. Strength and proprioception exercises were the least significant with p values >0.95 for both. Flexibility-improving exercises were also not significant ( p = 0.17). There were no differences between subjects in combat arms MOS and those not ( p = 0.95), those that worked primarily day duty hours versus night ( p = 0.70), and no differences between those with referrals from general practitioners versus self or orthopedicgenerated referrals ( p = 0.39). Table II shows categorical demographic, pathologic, and treatment variables of compliant and noncompliant subjects as measured by the PAT. DISCUSSION Our hypothesis for the primary aim of this study was that SM prescribed 2 or fewer exercises for a HEP would have a 188 MILITARY MEDICINE, Vol. 180, February 2015

4 TABLE I. Continuous Demographic, Pathologic, and Treatment Variables of Compliant and Noncompliant Subjects as Measured by the PAT Noncompliant Compliant Variable Median q1 q3 Median q1 q3 Kruskal Wallis p Value Age APFT Score a Hours Spent Exercising Per Week Number of Strength Exercises Number of Flexibility Exercises Number of Proprioception Exercises Total Number of Exercises in HEP Epworth Sleepiness Scale NPRS at Rest NPRS Worst in Last 48 Hours Days Between Intake and PAT GROC APFT, Army Physical Fitness Test; NPRS, Numeric Pain Rating Scale; GROC, Global Rating of Change. a Army Physical Fitness Test. A combined score of pushups, sit-ups and a 2-mile run given twice a year. Scored from 0 to 300. Values represent self-reported score of last test. higher rate of compliance at 1-week follow-up than those prescribed 3 or more. Our hypothesis was not supported by the results, as there was not a significant difference in compliance rate between subjects prescribed 2 or fewer exercises and those prescribed 3 or more ( p = 0.07). However, there was a statistically significant difference in compliance rate between subjects prescribed 2 or fewer exercises versus those prescribed 4 or more ( p = 0.02), with an overall strong trend demonstrating an inverse linear relationship between the number of exercises prescribed and the rate of compliance ( p = 0.01). Our hypothesis was not supported by the data likely because we failed to define groups different enough from each other in total number of exercises, our main variable of interest, to detect a difference between the groups. Once groups more dissimilar were compared, a difference was detected in agreement with our hypothesis. The strong negative trend demonstrated in Figure 1 supports this assertion. FIGURE 1. Rate of compliance in subjects prescribed 2 or fewer, 3, and 4 or more exercises as measured by the PAT. This finding is similar to those in the previous studies examining the relationship between HEP compliance and number of exercises prescribed. Henry et al 7 demonstrated that elderly patients who were prescribed 2 home exercises performed better than those prescribed 8, but no better than those prescribed 5 on the PAT. 7 Similarly, Medina et al 5 found that the number of exercises prescribed was a specific predictive factor for frequency adherence. In the same study, a significant difference was found in frequency adherence between those subjects with a HEP of 6 or more exercises versus those with 3 or fewer. 5 Interestingly, we found no difference in the rate of compliance between subjects prescribed different types of exercises (strength, flexibility, proprioception), suggesting this is not something prescribing therapists need to be as sensitive to when attempting to enhance compliance. The other statistically significant difference in the rate of compliance was between those subjects who left the base perimeter at least once per week and those who did not ( p = 0.01). The group of subjects that left the base perimeter at least weekly had a significantly lower rate of compliance. Although we cannot say with certainty that those working outside the base perimeter at least once per week worked longer hours than those who did not, we can say that those who left the base perimeter experienced greater psychological stress, as SM outside the base perimeter faced increased risk of attack. The psychological toll taken by repeating such activities once or more per week may have negatively impacted a subject s motivation or ability to comply with their HEP, especially when compounded by a long and fluctuating work schedule. These results are simple and intuitive but informative for deployed physical therapists. First, to maximize the rate of compliance, therapeutic exercises selected for HEP in theater should be the minimum number capable of addressing patients identified impairments. Second, deployed therapists may be able to enhance outcomes by asking patients if they leave the base at least once per week as part of their initial MILITARY MEDICINE, Vol. 180, February

5 TABLE II. Categorical Demographic, Pathologic, and Treatment Variables of Compliant and Noncompliant Subjects as Measured by the PAT Variable Noncompliant Compliant p Value n Female Gender 22% 33% 0.14 Number of Exercises ³ 3 93% 82% 0.07 Number of Exercises ³ 4 60% 39% 0.01 Combat Arms MOS 16% 16% 0.95 Self-Referral 33% 35% 0.39 Majority of Exercise in 15% 8% 0.20 Unit Physical Training Duty Hours Primarily at Night 13% 15% 0.70 Leave the Base at Least Once Per Week 55% 34% 0.01 MOS, Military Occupational Specialty. evaluation and provide those patients who leave the base more frequently with means of potentially enhancing compliance such has more frequent in-clinic sessions or handouts. These results suggest that prescribing a higher number of therapeutic exercises may not always be to the benefit of the patient. Deployed therapists should strive to select the most parsimonious set of therapeutic exercises possible and factor patients duty requirements into their overall plan of care. These results may also generalize to SM compliance with other medical treatments in theater, such as taking prescription medication and attendance of behavioral health appointments. Limitations There are several limitations to this study. First, subjects may not have accurately or truthfully recorded their level of compliance on the questionnaire. Although blinding the treating therapist to a subjects answer was used to mitigate this risk, there is still a possibility that a desirability bias influenced the subjects response to this question. The tendency for people to under or overestimate exercise performance in self-reports also limits the usefulness of any self-report measure of exercise performance. 14,15 Also, the questionnaire did not attempt to separate compliance out into the facets of frequency, duration, and accuracy, which have been demonstrated to have differing levels of compliance in patients. 5 Finally, the questionnaire was developed only for the purposes of this study and was never validated. For these reasons it is not surprising that the results failed to demonstrate convergent validity between the 2 compliance metrics. Had the questionnaire been validated or another type of self-report measure, such as a log used by Henry et al 7 in their study, we may have had found the same correlation between the PAT and self-report as they did, which originally led them to believe that the PAT has merit as a measure of HEP compliance. However, the PAT itself also has several limitations. This scale has not been validated and quantifies quality of performance, not compliance. Although it is likely that quality of performance is correlated with compliance to some degree, they are not the same thing. Furthermore, the dichotomous separation of subjects into categories of compliant and noncompliant based on a PAT cutoff score may not perfectly capture the continuous nature of a construct such as compliance. However, we felt this division was necessary for statistical reasons and because we feel such terms more accurately reflect the way in which the concept of compliance is used in the clinical environment. A score of 10 or higher was chosen to represent compliance as this range represents the top 20% of the 10 possible PAT scores We believe this cutoff reflects the minimum level of compliance required for a HEP to be efficacious. This is, of course, an estimate, and therefore a limitation. Another limitation results from the 18% dropout rate. Since failing to follow-up can be considered a form of noncompliance in itself, there is a possibility that our results demonstrate a rate of compliance higher than exists in reality. It is likely that many of the subjects who failed to follow-up were noncompliant with their HEP. Reasons may have included fluctuating duty schedules or feelings that their therapist was unable to help them, among others. Because only complete data sets were used in analysis, these subjects likely noncompliance was not captured in the results. Future studies should examine the rate of HEP compliance over longer periods of time, such as weeks or months, to identify those factors associated with long-term compliance and noncompliance. Studies should also examine compliance with other in-theater treatments such as taking prescription medication or attendance of behavioral health appointments to see if our findings generalize to other patient populations. Finally, we recommend intervention studies be done to determine the impact of aids such as handouts, videos, Web sites, and games on HEP compliance using the results of this and other observational studies as baseline data for comparison. CONCLUSION The deployed environment is a challenging one to practice as a physical therapist. Evidence-guiding treatment parameters to improve compliance, and subsequently outcomes, in this environment are of great value. We demonstrated in a prospective cohort study conducted in the Iraq and Afghanistan theaters that a patient who is prescribed 4 or more exercises is significantly less likely to be compliant than a patient prescribed 2 or fewer exercises. We also demonstrated that a deployed SM who left the base at least once per week was significantly less likely to be compliant than those who did not. Thus, for best outcomes, therapists should strive for brevity and consider duty requirements when designing a plan of care. REFERENCES 1. Armed Forces Health Surveillance Center: Army injury surveillance: September 2012 report. MSMR 2012; 19: Belmont PJ: Disease and nonbattle injuries sustained by a U.S. Army Brigade Combat Team during Operation Iraqi Freedom. Mil Med 2010; 175: MILITARY MEDICINE, Vol. 180, February 2015

6 3. Rhon DI, et al: Clinician perception of the impact of deployed physical therapists as physician extenders in a combat environment. Mil Med 2010; 175: Kolt GS: Adherence to rehabilitation in patients with low back pain. Man Ther 2003; 8: Medina-Mirapeix F, et al: Predictive factors of adherence to frequency and duration components in home exercise programs for neck and low back pain:an observational study. BMC Musculoskelet Disord 2009; 10. Available at ; accessed January 16, Escolar-Reina P, et al: How do care-provider and home exercise program characteristics affect patient adherence in chronic neck and back pain: a qualitative study. BMC Health Serv Res 2010; 10. Available at accessed January 16, Henry KD, Rosemond C, Eckert LB: Effect of number of home exercises on compliance and performance in adults over 65 years of age. Phys Ther 1999; 79: Slujis EM, et al: Correlates of exercise compliance in physical therapy. Phys Ther 1993; 73: Bassett SF: The assessment of patient adherence to physiotherapy rehabilitation. N Z J Physiother 2003; 31: Mori DL, et al: Predictors of exercise compliance in individuals with Gulf War veterans illnesses: department of Veterans Affairs Cooperative Study 470. Mil Med 2006; 171: Coppack RJ, et al: Use of a goal setting intervention to increase adherence to low back pain rehabilitation: a randomized control trial. Clin Rehabil 2012; 26: Meichenbaum D, Turk D: Facilitating Treatment Adherence: A Practioner s Guidebook. New York, Plenum Press, Brewer BW, et al: Construct validity and interrater agreement of the sport injury rehabilitation adherence scale. J Sport Rehabil 2002; 11: Prince SA, et al: A comparison of direct versus self-report measures for assessing physical activity in adults: a systematic review. Int J Behav Nutr Phys Act 2008; 5: 56. Available at accessed January 17, Watkinson C, et al: Overestimation of physical activity level is associated with lower BMI: a cross-sectional analysis. Int J Behav Nutr Phys Act. 2010; 7. Available at accessed January 16, MILITARY MEDICINE, Vol. 180, February