TY - JOUR
AU - Leonard,, Hakeem
AB - Abstract Total knee arthroplasty (TKA) is a common orthopedic surgery known to be very painful. Emphasis has been placed on TKA pain management for postoperative care and during rehabilitation. Music therapy is used as a nonpharmacologic intervention for pain management and to promote rehabilitation exercise adherence. The objective of this study was to explore the effects of music therapy/physical therapy co-treatment using live music-supported exercise on pain and exercise adherence during a lower extremity pedaling exercise to facilitate range of motion (ROM). The researcher randomized 32 TKA inpatient rehabilitation participants to an intervention or control group. Following baseline measures, two study intervals occurred with the intervention group receiving live music for the first interval followed by no music during the second interval; the control group received no music during both intervals. Self-reported pain measures, observed pain measures, and observed measures of pedaling adherence were collected for each participant. A mixed analysis of variance (ANOVA) with repeated measures showed no significant effects for self-reported pain perception. For observed pain, ANOVA results did show a significant interaction (p < .05) between group and study interval. There were no statistically significant effects for pedaling adherence. Conclusions show an important role for live music therapy intervention on observed pain while engaged in co-treatment during this lower extremity ROM exercise. Additional implications and limitations are discussed. Introduction As of 2010, researchers estimated that 4.7 million people (1.52% of the total population and 10.38% of those in their eighties) had received a knee replacement in the United States (Kremers et al., 2015). Regarding total knee arthroplasty (TKA), the prevalence of both primary (initial joint replacement) and revision (replacement of initial joint, usually after failure due to wear) surgeries are increasing. By 2030, it is projected that TKA surgeries could increase by 673% to 3.48 million procedures per year, compared with 402,100 primary and 32,700 revisions performed in 2003 (Kurtz, 2007). During rehabilitation from TKA, range of motion (ROM) is an important indicator of progress and long-term functional ability. ROM goals for inpatient rehabilitation are near 0 degrees extension and 110 degrees flexion as measured by a goniometer. Extension is especially important for proper gait during heel strike, prior to midstance, and toe-off to have proper balance and support. Flexion is important for many functional activities like navigating stairs and getting in and out of cars. After more passive exercises during the first days of postoperative care, physical therapy (PT) patients work on ROM using more active exercises, including devices with pedaling action, such as a stationary bike or a restorator (McLeod & Blackburn, 1980; Meier et al., 2008). Cycle or pedaling exercise can be an important early and continuing intervention for increasing ROM after knee replacement (Artz et al., 2015; Bakkum, 2015; Kendelhardt, 2003). However, active PT interventions can increase patient pain levels (Dahlen, Zimmerman, & Barron, 2006). Postoperative pain following orthopedic surgery can be very uncomfortable and sometimes unbearable. The natural response to this pain may be reduced activity and movement, which will eventually lead to increased long-term pain and decreased ROM and functional ability (Dahlen et al., 2006; Lingard, Katz, Wright, Sledge, & Kinemax Outcome Group, 2004). Pain can interfere with patients’ functioning, mood, walking ability, interpersonal relationships, and rest (Rastogi, 2007; Strassels, Chen, & Carr, 2002). In orthopedic patients, the intensity of pain can reach severe levels of aches, pains, and irritation. This pain may increase to even greater levels when patients participate in PT activities such as flexion exercises (Dahlen et al., 2006). When subjectively quantified, orthopedic surgeries can register at pain levels of nearly “worst pain imaginable.” Strassels and colleagues (2002) found that the mean worst pain severity of TKA’s was 7.6 on a 0–10 scale. The researcher found that though analgesics relieved 60–78% of the pain, that pain still hindered general functioning, walking, and sleep. There has been concern by medical professionals in recent years that orthopedic patients have not received adequate management of pain (Lamplot, Wagner, & Manning, 2014; Moucha, Weiser, & Levin, 2016). The drugs used most to treat moderate to severe pain in orthopedic rehabilitation are opioid analgesics (Elvir-Lazo & White, 2010). The analgesic (pain killer) effects of opioids may be reduced pain perception, reduced response to pain, and/or increase of pain threshold (Babos, Grady, Wisnoff, & McGhee, 2013). Despite this amelioration of pain, many have to strongly weigh the negative side effects that accompany taking opioid analgesics for pain. Normal side effects of opioids may be nausea, urinary retention, vomiting, respiratory depression, and hypotension (Moucha et al., 2016). Side effects, along with worries about development of drug dependence and the effectiveness threshold of unimodal drug treatment approaches, have led to under-treatment of pain (Sloman, Rosen, Rom, & Shir, 2005; Tighe et al., 2015). Medical professionals have responded to these multifaceted contributors to overall pain with emphasis on a multimodal approach to pain management. Multimodal therapy is facilitation of at least two forms of analgesia with separate action modes (Elmallah et al., 2016). This has been especially needed in orthopedic surgery (Lamplot et al., 2014). If inflammation is not treated properly in damaged tissues (including bone), it can cause a greater sensitization to pain and lowering of the pain threshold in peripheral nociceptors (Ekman & Koman, 2004). Modulation of pain sensitization in the central nervous system may contribute to the development of chronic pain. Ekman and Koman (2004) recommend nonpharmacologic approaches as part of the multimodal treatment method. Nonpharmacologic approaches include both physical and behavioral interventions. Melzack’s theories of pain support the use of music for active focus or distraction from pain through effect on psychological variables (Melzack, 1999; Melzack & Wall, 1965). The neuromatrix theory expands the understanding of nonpharmacologic pain relief (Trout, 2004). The gate control theory has been updated by the neuromatrix theory, which encompasses more readily the integration of body-self elements of the stress system and brain cognitive functions. This theory accounts for how subcortical brain structures connect and facilitate inputs that include the “sensory-discriminative, affective-motivational, and evaluative-cognitive dimension of pain experience” (p. S121). Patients can benefit greatly from a multimodal pain management regimen including pharmacological and nonpharmacologic approaches (Nett, 2010). Music as a nonpharmacologic pain management method can be very effective. Cepeda, Carr, Lau, and Alvarez (2006) found that of 3,600 patients experiencing pain related to surgery or impairing conditions, those receiving music (primarily music listening interventions) had a 70% greater probability of at least a 50% reduction in pain and decreased opioid requirements compared to control subjects. Music has also been shown to significantly reduce pain in orthopedic postoperative recovery (McCaffrey & Locsin, 2006). Allred, Byers, and Sole (2010) found that a nurse-delivered music listening intervention used in the days immediately after TKA significantly reduced pain and anxiety over time. Allred and colleagues also mentioned that music “potentially could be opioid sparing in some individuals, limiting the negative effects of opioids” (2010, p. 24). However, Ko, Chang, Lee, and Lin (2016) found no significant difference in pain levels between two groups receiving TKA or total hip arthroplasty when the experimental group received pain education and a music listening intervention. These music medicine interventions were all facilitated with recorded music. Clinical music therapy (MT) interventions have also been used to target pain perception. Kim and Koh (2005) investigated MT’s effect on pain perception for those in stroke rehabilitation during upper extremity joint exercise using recorded music. Lee (2016) included studies incorporating both music medicine and MT as part of a meta-analysis on the effects of music on pain. Lee includes prior definitions to help distinguish music medicine, defined as “pre-recorded music listening experiences administered by medical personnel” (Dileo & Bradt, 2005, p. 5) and MT, defined as an intervention “involving a relationship between client and therapist, a therapeutic process” (Dileo & Bradt, 2005, p. 9). In comparing MT versus music medicine, Lee (2016) found that both showed a significant reduction in pain perception. The results suggested that MT showed a larger clinical effect on participants’ perception of pain. The effect of MT on postoperative pain has also been explored. Bradt (2010) found live music to be effective in decreasing pain in pediatric patients. Madson and Silverman (2010) used live music to examine effects on pain, anxiety, and nausea of individuals participating in organ transplantation. Fredenburg and Silverman (2014) found significant differences between groups for affect and pain outcomes in a study with those recovering from blood and marrow transplants. The study found patient-preferred live music to be effective, with the researcher singing and playing guitar. Kendelhardt (2003) found that MT significantly reduced postoperative pain in inpatient rehabilitation patients performing a pedaling exercise after orthopedic surgery compared to those in the control group. This study used live contingent singing with guitar, stopping the music for 3 s when the participant stopped pedaling. Observed measures of pain are also an important consideration. While Kim and Koh (2005) measured pain perception with upper extremity joint exercises, they provided recommendations for future MT studies to measure variables such as facial expression and verbal indicators of pain. Of 97 studies included in Lee’s (2016) meta-analysis, 76 assessed pain with some type of 0–10 scale. Six studies used an observational measure of pain assessment. A key consideration during pedaling and all other rehabilitation exercise is patient compliance or adherence (henceforth referred to as adherence) (Bennell, Dobson, & Hinman, 2014). In one review of PT studies, patient adherence in exercise following knee replacement hovered around 60% for studies collecting data on this outcome (Artz et al., 2015). Attitudes about exercise can play a part in adherence when exercise is accompanied by knee pain (Holden, Nicholls, Young, Hay, & Foster, 2012). Making exercise programs more patient-centered can be a positive strategy (Bennell, Dobson, & Hinman, 2014). Giving participants a choice in music genre, song, or artist could be considered a patient-centered practice (Daveson, 2001). Combining interventions with some type of psychological, cognitive, behavioral, or affective component is also a key approach (McLean, Burton, Bradley, & Littlewood, 2010). In a study of orthopedic rehabilitation patients, including those having had knee replacements, Kendelhardt (2003) found that there was a difference in perceived rehabilitation levels in those who received MT during a pedaling exercise. Using live and recorded music to investigate motivation, well-being, and comfort during exercise with those receiving treatment for a bone marrow transplant, Boldt (1996) found more cooperation and increased participation with long-term participants and increased comfort and relaxation with short-term participants. Boldt recommended future research to examine initial motivation and participation levels during the combination of music with PT exercises. Past MT research has determined that certain measures of exercise adherence, such as counting the frequency of repetitions, may be less effective observation of exercise across time intervals (Johnson, Otto, & Clair, 2001). Johnson and colleagues also found that live instrumental music was more effective than live vocal music at increasing adherence. Clark, Baker, and Taylor (2012) found that live patterned sensory enhancement did not have the hypothesized effect on exertion, adherence, and mood during multiple exercises, but was seen as enjoyable for participants. In addition to rating scales for adherence, the study sought an objective measure of adherence using rehabilitation attendance records. The current study sought to discover if a MT/PT co-treatment session using live music to support exercise can reduce pain during PT activity for people who have had TKA. While live MT has been used for PT co-treatment with the restorator/pedal exerciser (Kendelhardt, 2003), this was not isolated to a specific population (i.e., TKA). Additionally, the effect of live music during a single restorator treatment session was not explored. Therefore, the problem this researcher sought to explore was the question of the effect of a live MT intervention on the pain and adherence of individuals rehabilitating from knee replacement surgery during a single restorator pedaling treatment. The research questions were (1) Does live MT during restorator pedaling exercise cause a difference in (a) self-reported and (b) observed measures of pain between groups (music and control) or within the music group over time? and (2) Does live MT during restorator pedaling cause a difference in exercise adherence between groups (music and control) or within the music group over time? Methods Study Design This was a stratified block-randomized controlled study with one experimental condition (MT/PT co-treatment using live music-supported exercise) and one control condition (PT alone). Stratified block-randomization was based on sex and analgesic level. Analgesic levels were represented as a proportion of 1 morphine equivalent, with participants divided into two categories: increased (≥0.5) morphine equivalent levels and reduced (<0.5) morphine equivalent levels (see study procedures and Table 1). Blocking and randomization were performed using an online randomization tool (Dallal, 2017). Sample size was not determined using a power analysis; rather, the sample was based on available study-eligible participants. Table 1 Demographic Characteristics Demographic Control, n =16 Intervention, n =16 Mean agea 67.6 67.9 Age range 53–80 45–87 Gender  Female 12 11  Male 4 5 Starting flexionb 83.3 86.3 Comorbidity  Depression 4c 3  Anxiety 1c 1  Chronic pain 1 1  Bilateral knee pain 1 0  Fibromyalgia 1 1 Analgesics  Opioid morphine   Equivalentd    ≥0.5e 3 3    <0.5f 13 13  Celecoxib 2 2  Acetaminophen 13 7  Aspirin 1 1 Demographic Control, n =16 Intervention, n =16 Mean agea 67.6 67.9 Age range 53–80 45–87 Gender  Female 12 11  Male 4 5 Starting flexionb 83.3 86.3 Comorbidity  Depression 4c 3  Anxiety 1c 1  Chronic pain 1 1  Bilateral knee pain 1 0  Fibromyalgia 1 1 Analgesics  Opioid morphine   Equivalentd    ≥0.5e 3 3    <0.5f 13 13  Celecoxib 2 2  Acetaminophen 13 7  Aspirin 1 1 Note. The analgesia demographics are reflective of those who received analgesia within duration of most recent administration in physical therapy (PT) window (PT session within time of drug duration). This is not reflective of if participant was administered drug prior to PT window. aMean age expressed in years. bStarting flexion expressed in degrees. cOne participant with depression and anxiety. dDependent on drug duration. eRange from 0.5 to 1.27. fRange from 0 to 0.38. View Large Table 1 Demographic Characteristics Demographic Control, n =16 Intervention, n =16 Mean agea 67.6 67.9 Age range 53–80 45–87 Gender  Female 12 11  Male 4 5 Starting flexionb 83.3 86.3 Comorbidity  Depression 4c 3  Anxiety 1c 1  Chronic pain 1 1  Bilateral knee pain 1 0  Fibromyalgia 1 1 Analgesics  Opioid morphine   Equivalentd    ≥0.5e 3 3    <0.5f 13 13  Celecoxib 2 2  Acetaminophen 13 7  Aspirin 1 1 Demographic Control, n =16 Intervention, n =16 Mean agea 67.6 67.9 Age range 53–80 45–87 Gender  Female 12 11  Male 4 5 Starting flexionb 83.3 86.3 Comorbidity  Depression 4c 3  Anxiety 1c 1  Chronic pain 1 1  Bilateral knee pain 1 0  Fibromyalgia 1 1 Analgesics  Opioid morphine   Equivalentd    ≥0.5e 3 3    <0.5f 13 13  Celecoxib 2 2  Acetaminophen 13 7  Aspirin 1 1 Note. The analgesia demographics are reflective of those who received analgesia within duration of most recent administration in physical therapy (PT) window (PT session within time of drug duration). This is not reflective of if participant was administered drug prior to PT window. aMean age expressed in years. bStarting flexion expressed in degrees. cOne participant with depression and anxiety. dDependent on drug duration. eRange from 0.5 to 1.27. fRange from 0 to 0.38. View Large Participants This study received institutional review board (IRB) approvals from the authors’ university and a hospital in the southeastern United Sates where the study took place. Participant inclusion criteria were (a) admission to the inpatient rehabilitation unit for primary or revision TKA, (b) having had a unilateral procedure, and (c) at least 30 years of age. Additionally, the researcher aimed to capture each participant’s first time on the restorator. Exclusion criteria were (a) bilateral TKA, (b) cognitive impairment, active psychosis, hearing impairment, or flexion measurements above 110 (as reported by the physical therapist), and (c) excess opioid intake (see procedures for threshold). Procedures The author (H. Leonard) served as the MT clinician and principal researcher for this study (hereafter referred to as “researcher”) and conducted all aspects of the outlined study procedures. To identify study-eligible participants, the researcher called the hospital therapy gym each day for reports on new TKA evaluations. When there were new admissions/evaluations, the researcher would come the same or next day to meet with potential participants, explain the study, and obtain informed study consent and consent for video prior to the participant’s scheduled PT session. At that point, the researcher checked the participant’s medical chart to assess analgesic level, and, then based on the participant’s biological sex and analgesic level, placed the participant into one of the predetermined randomized blocks. The researcher then returned to inform the participant which group they would be in, additionally taking music preference information from those randomized to the music group (i.e., preferred music genres and performing artists). In order to determine analgesic level, the researcher accessed medical charts to screen groups for relative equality of analgesia (Adams, 2005; Gippsland Region Palliative Care Consortium Clinical Practice Group, 2017; Jabs, B. G. Richards, & F. D. Richards, 2008). The researcher used multiple equianalgesic charts to tabulate levels of morphine equivalent opioid intake per participant. The amount of opioid analgesia was adjusted based on the window of time it was taken in regard to opioid onset time, opioid duration, half-life, and timing of PT exercise after opioid intake. Since drugs have varying durations and varying frequencies in which they are taken, this was taken into account for each type of analgesia. Opioids taken by participants in this study were hydrocodone, hydromorphone, oxycodone (Percocet), and morphine. Analgesic levels were represented as a proportion of 1 morphine equivalent (see Table 1). Participants were divided into categories of increased (≥0.5) morphine equivalent levels and reduced (<0.5) morphine equivalent levels. The researcher excluded one individual from the study due to the taking of tramadol, representing a very high outlier of morphine equivalent. Both study conditions took place either in the therapy gym or in the rehabilitation library of the hospital’s rehabilitation center. Prior to each session, the researcher set up a video camera to record each session for subsequent coding (i.e., observed pain scores). Each session began with the physical therapist setting up a chair and the restorator followed by the participant sitting to start; the physical therapist measured the flexion of each participant with a goniometer, followed by a 1-min baseline period, and two intervention periods of 2 min each. At the end of each period, participants gave self-reported pain scores to the researcher. After the final period, the physical therapist again took a flexion measurement. The researcher then asked three additional questions to those in the music group (see Appendix B). The researcher set up a camera on a stand prior to the sessions for video recording. Pedaling on the restorator was differentiated from other types of cycling. This has been referred to most commonly as a “pedal exerciser” in clinical language, research literature, and in retail spaces. When exploring the research literature about pedaling or cycling, it was a challenge to be clear about what type of machine is being used. Some studies cited in the review of literature mentioned the exercise bike, which is a bike with a seat installed and may have other different features. The restorator, as the machine was referred to at the hospital, is a free standing mini cycle without an attached chair; participants were sitting in regular chairs. Generally, with exercise bikes, adjustment for adaptation for an individual’s knee flexion level can be made by raising or lowering the height of the seat on the bike. With the restorator, adjustment was made by the individual moving closer or further away from the machine with direction from the physical therapist. Patients did not typically engage with the restorator bicycle for a set time. Some would exercise for longer times than others. When piloting the study, the researcher and PT staff selected 5 min for the study exercise, deciding that 8 min was too intensive. Additionally, it was also decided that a very short warm up period (no more than 10 s) was needed prior to baseline measures for each participant, because initial pain might be reflective of things such as relative stiffness or surprise. Intervention Group The intervention group participated in a MT/PT co-treatment session that used live music to support exercise. The session consisted of individualized live music (i.e., singing, paced guitar accompaniment) delivered by a credentialed music therapist to support a physical therapist-directed pedaling exercise, and PT alone, on the lower extremity restorator for a total of 5 min. The researcher, a board-certified music therapist with 7 years clinical experience at the time of the study, delivered all music interventions. There was the baseline (1 min) without music, followed by the first study interval (2 min with music), then the second study interval (2 min without music). This design was implemented in order to isolate the variable of music both between groups and within groups. The music intervention consisted of the music therapist providing live music which included singing with paced guitar accompaniment. The guitar was a Takamine GS-330S, a steel-string acoustic with a solid cedar top. Songs used during the music condition were based on individual participant music preferences taken during the consent process. The researcher either received two music options from a participant or decided on a song with the participant while in his or her hospital room (see Appendix A for songs used). In addition to patient preference, songs were selected based on tempo (i.e., moderate to fast tempo) to facilitate pedaling speed. In addition, all songs were in 4/4 or 6/8 meter, representing duple meter, which has been used in past rehabilitation studies (Kim & Koh, 2005). The music therapist informally assessed and established tempo of the song in the moment based on each participant’s baseline pedaling to match tempo and support their baseline pedaling rate of exercise during interval 1. There was therefore some flexibility for small variation of a song’s original tempo (see Discussion). Control Group The control group participated in PT on the lower extremity restorator for a total of 5 min. The baseline (1-min baseline), first study interval (2 min), and second study interval (2 min), each performed without music. Measures Measures included participants’ self-reported perceived pain, observed pain scores, and observed pedaling adherence scores. Participants were involved in repeated measurements, with data taken after each of two intervals following a baseline interval. Numeric Rating Scale.  An adapted Numeric Rating Scale (NRS) was used to measure participants’ self-reported pain. The NRS is commonly used by physicians and nurses to inform pain medicine regimes for patients (Williamson & Hoggart, 2005). While there is no data related to distribution of error for the NRS, as an intervallic measurement, it provides parametric analysis. The NRS has also shown to correlate well with other pain scales (Bijur, Latimer, & Gallagher, 2003). This consists of a patient ranking his or her pain level on a 0–10 scale, with 0 representing “no pain” and 10 “the worst pain imaginable.” The NRS is a key component that doctors and nurses use to decide both the initial and ongoing pain medicine regimen for patients (Williamson & Hoggart, 2005). Researchers studying TKA and therapeutic uses of music have made use of the NRS (Ko et al., 2016; Merle et al., 2014). Though some studies use the NRS of 0–10, the researcher chose a scale of 1–10 to match the hospital’s use of the NRS, as reported by the head staff music therapist, and as other researchers have applied as well (McCaffrey & Locsin, 2006). Since ratings were not just a “state” rating, but dynamic and centered around a procedure, the researcher chose to operationalize the way the question about “pain” was asked for this study. The researcher asked participants, “As you were pedaling, what was your perception of pain or pain experience on a 1–10 scale, with 1 being no pain and 10 being the worst pain imaginable?” Observational Coding for Pain.  For observed pain, partial-interval recording was used to assess video of each participant for nonverbal (motor, facial) and verbal indicators of pain. The target pain behaviors were adapted from previous research indicating a correlation between self-reported pain perception and observed indicators of acute pain (Puntillo et al., 2004). Observers watched videos in 15-s intervals and marked an “X” for any occurrence of “facial,” “verbal,” or “motor” indicators of pain observed during the interval. If a pain behavior from any of these three categories was marked for an interval, then that interval was counted as a pain interval. Table 2 includes operational definitions for coded pain behaviors across the three categories (facial, verbal, and motor). Baseline measures included four 15-s time intervals and each study interval included eight of these time intervals. Three trained observers (Masters-level MT students) and the researcher performed video coding and reliability, with an interrater reliability threshold of .8. The researcher created videos mimicking pain behaviors while doing the pedaling exercise to train observers. Observers were not blinded to treatment conditions. Table 2 Operation Definitions of Pain Behaviors Facial  Grimace: a sharp contortion of the face  Furrowed brow: wrinkling the brows in a negative manner  Wince: shrinking, flinching, quickly drawing back  Eyes closed: lids are closed with tightness in brow, lips, or neck  Eyes wide open: with eyebrows raised in anticipation  Looking away: in the opposite direction of pain  Grin/smile: upward curling at corners of the mouth; lips closed or showing teeth  Mouth wide open: to expose teeth and tongue  Clenched teeth: exposing slightly opened mouth Verbal  Moaning: low, soft indistinguishable sounds  Screaming: loud, piercing, shrill cry or sound  Whimpering: low, broken sobbing sounds  Crying: loud, sobbing sounds, with onset of tears  Protest words: words used to object, e.g., stop, no, don’t, I can’t  Verbal complaints: words used to describe pain, e.g., it hurts, ouch Motor  Clenched fist: act of forming a fist  Withdrawing: moving away from  Rubbing/massaging: act of touching in a back a forth movement  Defensive grabbing: reach out to grasp knee area  Pausing: struggling with a steady cycle Facial  Grimace: a sharp contortion of the face  Furrowed brow: wrinkling the brows in a negative manner  Wince: shrinking, flinching, quickly drawing back  Eyes closed: lids are closed with tightness in brow, lips, or neck  Eyes wide open: with eyebrows raised in anticipation  Looking away: in the opposite direction of pain  Grin/smile: upward curling at corners of the mouth; lips closed or showing teeth  Mouth wide open: to expose teeth and tongue  Clenched teeth: exposing slightly opened mouth Verbal  Moaning: low, soft indistinguishable sounds  Screaming: loud, piercing, shrill cry or sound  Whimpering: low, broken sobbing sounds  Crying: loud, sobbing sounds, with onset of tears  Protest words: words used to object, e.g., stop, no, don’t, I can’t  Verbal complaints: words used to describe pain, e.g., it hurts, ouch Motor  Clenched fist: act of forming a fist  Withdrawing: moving away from  Rubbing/massaging: act of touching in a back a forth movement  Defensive grabbing: reach out to grasp knee area  Pausing: struggling with a steady cycle View Large Table 2 Operation Definitions of Pain Behaviors Facial  Grimace: a sharp contortion of the face  Furrowed brow: wrinkling the brows in a negative manner  Wince: shrinking, flinching, quickly drawing back  Eyes closed: lids are closed with tightness in brow, lips, or neck  Eyes wide open: with eyebrows raised in anticipation  Looking away: in the opposite direction of pain  Grin/smile: upward curling at corners of the mouth; lips closed or showing teeth  Mouth wide open: to expose teeth and tongue  Clenched teeth: exposing slightly opened mouth Verbal  Moaning: low, soft indistinguishable sounds  Screaming: loud, piercing, shrill cry or sound  Whimpering: low, broken sobbing sounds  Crying: loud, sobbing sounds, with onset of tears  Protest words: words used to object, e.g., stop, no, don’t, I can’t  Verbal complaints: words used to describe pain, e.g., it hurts, ouch Motor  Clenched fist: act of forming a fist  Withdrawing: moving away from  Rubbing/massaging: act of touching in a back a forth movement  Defensive grabbing: reach out to grasp knee area  Pausing: struggling with a steady cycle Facial  Grimace: a sharp contortion of the face  Furrowed brow: wrinkling the brows in a negative manner  Wince: shrinking, flinching, quickly drawing back  Eyes closed: lids are closed with tightness in brow, lips, or neck  Eyes wide open: with eyebrows raised in anticipation  Looking away: in the opposite direction of pain  Grin/smile: upward curling at corners of the mouth; lips closed or showing teeth  Mouth wide open: to expose teeth and tongue  Clenched teeth: exposing slightly opened mouth Verbal  Moaning: low, soft indistinguishable sounds  Screaming: loud, piercing, shrill cry or sound  Whimpering: low, broken sobbing sounds  Crying: loud, sobbing sounds, with onset of tears  Protest words: words used to object, e.g., stop, no, don’t, I can’t  Verbal complaints: words used to describe pain, e.g., it hurts, ouch Motor  Clenched fist: act of forming a fist  Withdrawing: moving away from  Rubbing/massaging: act of touching in a back a forth movement  Defensive grabbing: reach out to grasp knee area  Pausing: struggling with a steady cycle View Large Observational Coding for Pedaling Adherence.  Whole-interval recording was used to assess “pedaling” versus “non-pedaling.” “Pedaling” was defined as an individual continuously pedaling for an entire (or whole) 15-s time interval duration, while “non-pedaling” was defined as an individual failing to pedal for any 1 continuous second during a 15-s time interval. Pedaling adherence was defined by the number of intervals a participant was pedaling (as opposed to non-pedaling) among total intervals observed. The researcher and two observers each coded 9 of 32 participant videos (28% of participant intervals) for observed pain and pedaling adherence. Data Analysis For this study, a baseline interval was included to assess whether the two groups were equivalent. Independent t-tests were used to analyze baseline measures. Baseline data for pain perception, observed pain, and pedaling adherence show that there was no difference between the groups (see Table 3). Table 3 Independent T-Test Descriptive Data for Baseline Measures Source n Mean SD t df p Self-Reported Pain  Control Group 16 6.66 2.24 0.18 30 .86  Music Group 16 6.50 2.68 Observed Paina  Control Group 16 1.79 1.68 0.33 30 .75  Music Group 16 1.61 1.58 Pedaling Adherencea  Control Group 16 3.59 0.88 0.81 30 .43  Music Group 16 3.34 0.87 Source n Mean SD t df p Self-Reported Pain  Control Group 16 6.66 2.24 0.18 30 .86  Music Group 16 6.50 2.68 Observed Paina  Control Group 16 1.79 1.68 0.33 30 .75  Music Group 16 1.61 1.58 Pedaling Adherencea  Control Group 16 3.59 0.88 0.81 30 .43  Music Group 16 3.34 0.87 Note.aMissing baseline video data from two participants (one in each group). Values were imputed from mean data. View Large Table 3 Independent T-Test Descriptive Data for Baseline Measures Source n Mean SD t df p Self-Reported Pain  Control Group 16 6.66 2.24 0.18 30 .86  Music Group 16 6.50 2.68 Observed Paina  Control Group 16 1.79 1.68 0.33 30 .75  Music Group 16 1.61 1.58 Pedaling Adherencea  Control Group 16 3.59 0.88 0.81 30 .43  Music Group 16 3.34 0.87 Source n Mean SD t df p Self-Reported Pain  Control Group 16 6.66 2.24 0.18 30 .86  Music Group 16 6.50 2.68 Observed Paina  Control Group 16 1.79 1.68 0.33 30 .75  Music Group 16 1.61 1.58 Pedaling Adherencea  Control Group 16 3.59 0.88 0.81 30 .43  Music Group 16 3.34 0.87 Note.aMissing baseline video data from two participants (one in each group). Values were imputed from mean data. View Large Pain perception, observed pain, and pedaling adherence data were analyzed using a mixed analysis of variance (ANOVA) with repeated measures on the factor of time (study interval). For pain perception, observed pain, and pedaling adherence, normality of distribution was established using a Shapiro–Wilk test, and Levene’s test for Equality of Variances was used to establish homogeneity of variances for baseline and for the study intervals. For dependent variables, assumptions were met except for homogeneity of variance for pedaling adherence. Correcting for non-homogeneity using the Greenhouse-Geisser method had no impact upon the results, so the unadjusted results are reported. An alpha level of .05 was used for all statistical comparisons. Statistical tests were facilitated using the program JASP, version 0.8.1.2. Due to the stratified block design of the study, ANOVA data were run using categorizations of analgesic level and gender as covariates. When there was a significant interaction between study interval and group, post hoc paired t-tests were used to compare outcomes between study intervals within the two groups. Analysis of interrater reliability was performed with SPSS 25.0 using intraclass correlation coefficients (ICCs). Results Sample A total of 39 participants were assessed for eligibility, with 32 participants enrolled in the study. Three individuals were excluded—two because they were put on the restorator prior to the research protocol and one due to excess opioid intake. A participant flow chart is shown in Figure 1. Four individuals declined to participate in the study due to lack of interest. Table 1 contains demographic and clinical characteristics of enrolled participants. Participants’ age ranged from 45 to 87 years, and the majority were female. There was no difference between groups. Figure 1. View largeDownload slide Participant flowchart. Figure 1. View largeDownload slide Participant flowchart. Self-Reported Pain Perception A mixed ANOVA with repeated measures on one factor (study interval) was calculated for pain perception ratings, with no significant interaction between group and study interval (see Table 4). Although between-group differences were not statistically significant, the control group reported higher mean scores for self-reported pain at time interval 1, compared with the intervention group. In interval 2, when both groups experienced no music, the control group reported lower means scores compared with the intervention group. Looking at within group change, control group self-reported mean pain scores decreased from interval 1 to interval 2, while intervention group self-reported mean pain scores increased during interval 2 when the music intervention was removed (see Table 7). Table 4 ANOVA Effects for Self-Reported Pain Perception Source SS df MS F p partial η2 Between Groups Group 5.24 1 5.24 0.43 .52 .02 Medication 12.06 1 12.06 0.98 33 .03 Gender 14.00 1 14.00 1.14 .30 .04 Error 344.57 28 12.31 Within Groups Study Interval 0.71 1 0.71 0.44 .52 .02 Study Interval × Group 5.37 1 5.37 3.31 .08 .11 Study Interval × Medication 1.79 1 1.79 1.11 .30 .04 Study Interval × Gender 0.02 1 0.02 0.01 .91 .00 Error 45.39 28 1.62 Source SS df MS F p partial η2 Between Groups Group 5.24 1 5.24 0.43 .52 .02 Medication 12.06 1 12.06 0.98 33 .03 Gender 14.00 1 14.00 1.14 .30 .04 Error 344.57 28 12.31 Within Groups Study Interval 0.71 1 0.71 0.44 .52 .02 Study Interval × Group 5.37 1 5.37 3.31 .08 .11 Study Interval × Medication 1.79 1 1.79 1.11 .30 .04 Study Interval × Gender 0.02 1 0.02 0.01 .91 .00 Error 45.39 28 1.62 Note. MS = mean of squares; SS = sum of squares. View Large Table 4 ANOVA Effects for Self-Reported Pain Perception Source SS df MS F p partial η2 Between Groups Group 5.24 1 5.24 0.43 .52 .02 Medication 12.06 1 12.06 0.98 33 .03 Gender 14.00 1 14.00 1.14 .30 .04 Error 344.57 28 12.31 Within Groups Study Interval 0.71 1 0.71 0.44 .52 .02 Study Interval × Group 5.37 1 5.37 3.31 .08 .11 Study Interval × Medication 1.79 1 1.79 1.11 .30 .04 Study Interval × Gender 0.02 1 0.02 0.01 .91 .00 Error 45.39 28 1.62 Source SS df MS F p partial η2 Between Groups Group 5.24 1 5.24 0.43 .52 .02 Medication 12.06 1 12.06 0.98 33 .03 Gender 14.00 1 14.00 1.14 .30 .04 Error 344.57 28 12.31 Within Groups Study Interval 0.71 1 0.71 0.44 .52 .02 Study Interval × Group 5.37 1 5.37 3.31 .08 .11 Study Interval × Medication 1.79 1 1.79 1.11 .30 .04 Study Interval × Gender 0.02 1 0.02 0.01 .91 .00 Error 45.39 28 1.62 Note. MS = mean of squares; SS = sum of squares. View Large Observed Pain ICCs were used to calculate average interrater reliability for behavioral coding of observed pain. The ICC results show high reliability (.88) for observed pain. Results of the mixed ANOVA show that there was a significant interaction between study interval and group (F(1,28) = 5.80, p = .023, partial η2 = .17). Analysis is given in Table 5. Mean data are given in Table 7. Although the interaction effect for group and study interval was significant in the ANOVA, when post hoc paired t-tests were conducted to compare the two study intervals within the intervention and control groups, there were no significant differences between study interval found in either group. Figure 2 shows the interaction between group and study interval for observed pain. This figure illustrates that during interval 1, when music was present for the intervention group, intervention participants’ mean observed pain scores remained consistent from baseline. In contrast, the control group experienced increased observed pain from baseline to interval 1. During interval 2, when both the intervention and control groups experienced no music, the intervention group experienced elevated observed pain (similar to the control group in interval 1), while control group pain scores decreased greatly. Table 5 ANOVA Effects for Observed Pain Source SS df MS F p partial η2 Between Groups Group 2.66 1 2.66 0.24 .63 .01 Medication 4.54 1 4.54 0.41 .53 .01 Gender 2.61 1 2.61 0.23 .63 .01 Error 314.02 28 11.22 Within Groups Study Interval 1.17 1 1.17 0.28 .60 .01 Study Interval × Group 24.13 1 24.13 5.76 .02a .17 Study Interval ×Medication 1.74 1 1.74 4.18 1.00 .00 Study Interval × Gender 5.93 1 5.93 1.42 .24 .05 Error 116.58 28 4.16 Source SS df MS F p partial η2 Between Groups Group 2.66 1 2.66 0.24 .63 .01 Medication 4.54 1 4.54 0.41 .53 .01 Gender 2.61 1 2.61 0.23 .63 .01 Error 314.02 28 11.22 Within Groups Study Interval 1.17 1 1.17 0.28 .60 .01 Study Interval × Group 24.13 1 24.13 5.76 .02a .17 Study Interval ×Medication 1.74 1 1.74 4.18 1.00 .00 Study Interval × Gender 5.93 1 5.93 1.42 .24 .05 Error 116.58 28 4.16 Note. MS = mean of squares; SS = sum of squares. aSignificant Effect. View Large Table 5 ANOVA Effects for Observed Pain Source SS df MS F p partial η2 Between Groups Group 2.66 1 2.66 0.24 .63 .01 Medication 4.54 1 4.54 0.41 .53 .01 Gender 2.61 1 2.61 0.23 .63 .01 Error 314.02 28 11.22 Within Groups Study Interval 1.17 1 1.17 0.28 .60 .01 Study Interval × Group 24.13 1 24.13 5.76 .02a .17 Study Interval ×Medication 1.74 1 1.74 4.18 1.00 .00 Study Interval × Gender 5.93 1 5.93 1.42 .24 .05 Error 116.58 28 4.16 Source SS df MS F p partial η2 Between Groups Group 2.66 1 2.66 0.24 .63 .01 Medication 4.54 1 4.54 0.41 .53 .01 Gender 2.61 1 2.61 0.23 .63 .01 Error 314.02 28 11.22 Within Groups Study Interval 1.17 1 1.17 0.28 .60 .01 Study Interval × Group 24.13 1 24.13 5.76 .02a .17 Study Interval ×Medication 1.74 1 1.74 4.18 1.00 .00 Study Interval × Gender 5.93 1 5.93 1.42 .24 .05 Error 116.58 28 4.16 Note. MS = mean of squares; SS = sum of squares. aSignificant Effect. View Large Figure 2. View largeDownload slide Interaction between group and study interval for observed pain. Figure 2. View largeDownload slide Interaction between group and study interval for observed pain. Pedaling Adherence The ICC scores for interrater reliability were .76 for pedaling adherence. Results of a mixed ANOVA show there was no significant interaction between group and study interval. Analysis is given in Table 6. Mean data are given in Table 7. Table 6 ANOVA Effects for Pedaling Adherence Source SS df MS F p partial η2 Between Groups Group 2.42 1 2.42 2.44 .13 .08 Medication 0.34 1 0.34 0.34 .57 .01 Gender 0.87 1 0.87 0.88 .36 .03 Error 27.68 28 0.99 Within Groups Study Interval 0.04 1 0.04 0.22 .64 .01 Study Interval × Group 0.02 1 0.02 0.10 .76 .00 Study Interval × Medication 0.00 1 0.00 0.01 .91 .00 Study Interval × Gender 0.02 1 0.02 0.09 .76 .00 Error 5.08 28 0.18 Source SS df MS F p partial η2 Between Groups Group 2.42 1 2.42 2.44 .13 .08 Medication 0.34 1 0.34 0.34 .57 .01 Gender 0.87 1 0.87 0.88 .36 .03 Error 27.68 28 0.99 Within Groups Study Interval 0.04 1 0.04 0.22 .64 .01 Study Interval × Group 0.02 1 0.02 0.10 .76 .00 Study Interval × Medication 0.00 1 0.00 0.01 .91 .00 Study Interval × Gender 0.02 1 0.02 0.09 .76 .00 Error 5.08 28 0.18 Note. MS = mean of squares; SS = sum of squares. View Large Table 6 ANOVA Effects for Pedaling Adherence Source SS df MS F p partial η2 Between Groups Group 2.42 1 2.42 2.44 .13 .08 Medication 0.34 1 0.34 0.34 .57 .01 Gender 0.87 1 0.87 0.88 .36 .03 Error 27.68 28 0.99 Within Groups Study Interval 0.04 1 0.04 0.22 .64 .01 Study Interval × Group 0.02 1 0.02 0.10 .76 .00 Study Interval × Medication 0.00 1 0.00 0.01 .91 .00 Study Interval × Gender 0.02 1 0.02 0.09 .76 .00 Error 5.08 28 0.18 Source SS df MS F p partial η2 Between Groups Group 2.42 1 2.42 2.44 .13 .08 Medication 0.34 1 0.34 0.34 .57 .01 Gender 0.87 1 0.87 0.88 .36 .03 Error 27.68 28 0.99 Within Groups Study Interval 0.04 1 0.04 0.22 .64 .01 Study Interval × Group 0.02 1 0.02 0.10 .76 .00 Study Interval × Medication 0.00 1 0.00 0.01 .91 .00 Study Interval × Gender 0.02 1 0.02 0.09 .76 .00 Error 5.08 28 0.18 Note. MS = mean of squares; SS = sum of squares. View Large Table 7 Mean Scores and SDs for Outcome Measures Study Interval 1a Study Interval 2 Source Mean (SD) Mean (SD) Self-reported pain perception  Control group 5.91 (2.27) 5.56 (2.52)  Music group 4.69 (2.50) 5.44 (3.20) Observed painb  Control group 3.19 (3.21)c 2.31 (2.36)  Music group 1.56 (1.97) 3.06 (3.13) Pedaling adherenced  Control group 7.56 (0.81)c 7.44 (1.21)  Music group 8.00 (0.00) 7.81 (0.40) Study Interval 1a Study Interval 2 Source Mean (SD) Mean (SD) Self-reported pain perception  Control group 5.91 (2.27) 5.56 (2.52)  Music group 4.69 (2.50) 5.44 (3.20) Observed painb  Control group 3.19 (3.21)c 2.31 (2.36)  Music group 1.56 (1.97) 3.06 (3.13) Pedaling adherenced  Control group 7.56 (0.81)c 7.44 (1.21)  Music group 8.00 (0.00) 7.81 (0.40) Note.aIntervention interval for the music group. bFor “Observed pain,” mean scores indicate the mean number of units in which pain was exhibited. cFor “Observed pain” and “Pedaling adherence,” study intervals 1 and 2 were each 2 min (eight 15-s intervals) of PT activity, translating to a maximum unit of 8. dFor “Pedaling adherence,” mean scores indicate the mean number of units in which pedaling occurred. View Large Table 7 Mean Scores and SDs for Outcome Measures Study Interval 1a Study Interval 2 Source Mean (SD) Mean (SD) Self-reported pain perception  Control group 5.91 (2.27) 5.56 (2.52)  Music group 4.69 (2.50) 5.44 (3.20) Observed painb  Control group 3.19 (3.21)c 2.31 (2.36)  Music group 1.56 (1.97) 3.06 (3.13) Pedaling adherenced  Control group 7.56 (0.81)c 7.44 (1.21)  Music group 8.00 (0.00) 7.81 (0.40) Study Interval 1a Study Interval 2 Source Mean (SD) Mean (SD) Self-reported pain perception  Control group 5.91 (2.27) 5.56 (2.52)  Music group 4.69 (2.50) 5.44 (3.20) Observed painb  Control group 3.19 (3.21)c 2.31 (2.36)  Music group 1.56 (1.97) 3.06 (3.13) Pedaling adherenced  Control group 7.56 (0.81)c 7.44 (1.21)  Music group 8.00 (0.00) 7.81 (0.40) Note.aIntervention interval for the music group. bFor “Observed pain,” mean scores indicate the mean number of units in which pain was exhibited. cFor “Observed pain” and “Pedaling adherence,” study intervals 1 and 2 were each 2 min (eight 15-s intervals) of PT activity, translating to a maximum unit of 8. dFor “Pedaling adherence,” mean scores indicate the mean number of units in which pedaling occurred. View Large Follow-up Questions At the end of the data collection, the researcher asked each participant in the music group (n = 16) to rate “how pleasing” and “how helpful” the music was on a 5-point Likert-type rating scale. The choices for each question ranged from “not at all,” “not very,” “neutral,” “fairly,” and “very.” Each participant in the music group was also asked if they preferred the restorator exercise with or without music. All 16 participants stated that they preferred the PT exercise “with music.” Complete descriptive data for follow-up questions are given in Appendix B. Discussion This study was driven by two research questions, (1) “Does live music during restorator exercise cause a difference in (a) self-reported and (b) observed measures of pain?” and (2) “Does live MT during restorator pedaling cause a difference in patient exercise adherence?” For self-reported pain and exercise adherence, mixed ANOVA results show that there was not an interaction between group and study interval. Observed pain data from video observations show that there was a significant interaction between group and time interval. Yet, post hoc tests show no study interval effects within either group. This outcome does not give a clear indication that observed pain is reduced by live MT. However, mean score data reveal interesting trends that may warrant further investigation in subsequent trials. Figure 2 shows, after baseline, during study interval 1, there seemed to be an important difference in observed pain with the intervention group (music condition) exhibiting fewer pain behaviors than the control group. Moving to study interval 2, there was a change, with increased pain behaviors for the invention group (no music occurred during this interval) and fewer pain behaviors for the control group. Although not statistically significant, these data seem to suggest the importance of ongoing music to help manage pain. It is interesting that from interval 1 to interval 2, the MT group had heightened pain, while the control had diminishing pain. This might suggest that the music intervention participants were able to focus their attention on the music, rather than their pain; but when music was removed, their attention focused on pain (much like the control group experienced in the first interval). The data also suggest that the control group was perhaps habituating to the pain and/or anticipating the end of their pain in interval 2 (hence the downward trend in their pain scores). These findings add to existing research findings by showing that the intervention may be effective within a specific orthopedic population (i.e., TKA). In a similar study, Kendelhardt (2003) found that live MT can lead to a significant reduction of pain while using the restorator in orthopedic inpatient rehabilitation. Similar to the current study, Kendelhardt was also researcher and therapist, the study included experimental and control groups, and dependent variables included self-perception of pain and negative verbalization frequency. Important differences across these two studies is that subjects in Kendelhardt’s study were seen over multiple PT sessions; there were additional dependent variables of subjects’ exercise duration, anxiety, and rehabilitation levels; included multiple orthopedic conditions; did not seek a comprehensive dependent variable of observed pain; and live MT was contingent (music was stopped for 3 s when the subject stopped pedaling). Taken together, these studies support further exploration and development of music interventions as a nonpharmacologic, multimodal approach to pain management (Ekman & Koman, 2004; Nett, 2010). For pedaling adherence, observed data show that there was not a difference between the two groups or across time. Both groups had good adherence, with the intervention group having greatest adherence during the intervention interval (interval 1), with an important note that only during the music condition (first interval) did participants exhibit 100% pedaling adherence. Study participants indicated a preference for PT with music and indicated that MT was both helpful and pleasing. Similar to Kim and Koh (2005) and Clark and colleagues (2012), participants expressed and exhibited many verbal and nonverbal responses corresponding to enjoyment of the music, which aligns with conclusions of Clark and colleagues (2012) that familiar musical elements seem to be important for adherence. Positive participant responses in combination with full pedaling adherence during the music condition support the idea that participants were likely motivated. Also consistent with studies by Kim and Koh (2005) and Clark and colleagues (2012), physical therapists anecdotally communicated positive thoughts and comments about the effects of MT, in this case on adherence, motivation, and the overall atmosphere of the therapeutic space. While Johnson and colleagues (2001) found that the use of familiar vocal music during exercise seemed to lead toward decreased adherence and distraction from exercise, in the current study, participants were able to sing while continuing to pedal. Johnson and colleagues stated that singing and the use of familiar music might be contraindicated during rehabilitation exercise and adherence; however, in their study, the music therapist used live vocal music without instrumental accompaniment to support 14 prescribed PT rehabilitation exercises. A variety of factors may have contributed to these differential findings including the patient population (TKA patients vs. elderly), type of exercise (pedaling vs. a 20-min protocol of 14 exercises), and the use of an accompanying instrument played at tempo matched to desired rate of exercise (current study). Future studies are needed to isolate these factors, examine whether pedaling adherence is related to changes in observed pain, and determine the specific functions of music that may be responsible for changes observed in this study. During the music condition, it can be seen on the video recordings that many participants sang along with the therapist while pedaling, suggesting that the music may have helped regulate participants’ attention—focusing on the music, rather than their pain. This is consistent with neuromatrix theory which states that the stress system’s integration (incorporating the sensory and affective dimensions) with the higher cognitive systems facilitate inputs and may help explain the connection between increased attention and decreased pain during active engagement with music (Melzack, 1999; Trout, 2004). Participants also participated in song selection which likely reinforced the use of a patient-centered approach which is a recommended pain management strategy (Daveson, 2001). In summary, the music intervention may have influenced cognitive (i.e., attention regulation) and affective areas of functioning (i.e., mood), both of which have been noted as important in managing pain and influencing adherence (Ekman & Koman, 2004; McLean et al., 2010). Participants in the music group also reported the live MT intervention to be “helpful” and “pleasing” and preferred the exercises with music versus without music. Music may influence attitude regarding pain and exercise, which Holden and colleagues (2012) stated was important to adherence. Finally, it is important to discuss the possible effects of tempo on participants’ exercise adherence. It is not known if music functioned as entrainment to structure movement in this study. Descriptions of entrainment within neurologic MT research stipulate that music not merely accompany or respond to movement, but drive the movement (Thaut, 2014). While music was chosen to facilitate and structure movement, it was not conceived to explicitly drive movement in this study. Perhaps music, functioned as entrainment for some participants, was merely a rhythmic cue, or had the quality of rhythmicity, functioning in the attentional-distraction dimension, as Brown, Chen, and Dworkin (1989) described. A sample of tempo played from intervention videos show that songs were played within 20% lower or higher tempo than the original recordings. As examples, “Rolling in the Deep” by Adele was played at 110 BPM as opposed to 104 BPM in the original recording, representing a 6% tempo increase; “Take it Easy” by The Eagles was played at 155 BPM while the original recording is 139 BPM—a 12% increase. Regarding a song played with two different participants, “On the Road Again” by Willie Nelson has an original tempo of 111 BPM. It was played live once at 104 BPM and again at 101 BPM, representing a 6% and 9% decrease, respectively. In future studies, researchers might investigate the mechanism related to rhythmic interventions that influence attention and the motor system but may not be defined as entrainment such as reflecting, matching, or synchronization with movement. In this study, the researcher used observation of baseline speed to determine tempo for the music, working to synchronize strumming with pedaling, but questions remain about how tempo might be more explicitly determined during pedaling or if there is a specific point in the cycle rotation where the downbeat should fall. Limitations Findings should be interpreted in light of study limitations which include the absence of a power calculation to determine sample size, use of a blinded assessor to gather self-report pain and satisfaction scores, and the involvement of the researcher in behavioral coding for observed pain scores. The absence of an a priori power calculation to inform sample size limits our ability to know whether the study was adequately powered to detect clinically meaningful differences. The use of non-blinded evaluator introduced risk for biased responses from participants for self-reported pain and satisfaction. Removing the therapist from the administration of evaluations in future studies would help to address this limitation. Interpretation of results for observed pain measures must be considered alongside the relatively high standard deviations of mean score data (see Table 7). This shows that some participants had large swings in pain across time. There were also some who experienced little pain throughout. There were missing values (one participant in each group) for objective baseline measures. This was due to video failure. SPSS 25 was used to identify the mean values to replace missing values. Though this study did not have a very large sample, the negative impact of these missing values should be limited due to having effected just one participant from each group. Additionally, baseline measures were used only to establish group equivalence. The independent samples t-tests found that groups were equal. The researcher used a 1–10 NRS in the study due to the use of it at the hospital where the research was facilitated. However, traditionally, the NRS is an 11-point scale from 0 to 10 and much of the research uses this. This most likely affected the analysis of data for the self-reported measure of pain. The use of the 0–10 scale may have accounted for additional subtleties in pain differences among participants in the study and it has been validated in past research. The fact that the music intervention was one-instance for a narrow time frame is also a limitation. It may be important for studies to be replicated in both single and multiple sessions or for interventions of longer duration. Though the study compares live music to a control group, there is not a comparison with a recorded music group. One could wonder if recorded music would have yielded similar or better results. As stated, this study might have benefitted from an increased sample size, as between-group data show trends toward significance with a larger sample size. Outcomes of both pain perception and pedaling adherence seem to suggest that a larger sample might trend toward statistical significance. Finally, future studies may want to consider the addition of a standardized adherence tool such as the SIRAS (Clark et al., 2012). However, the use of whole-interval recording allowed an effective and feasible objective measurement. Unlike Johnson and colleagues (2001), the measure in this study did seem to appropriately capture participants’ responses accurately. Suggestions for Future Research A fully powered replication study would offer more definitive evidence about the use of live, preferred music with guitar accompaniment to diminish pain and improve adherence during lower extremity pedaling exercise in patients with TKA. People who have undergone TKA, as well as other orthopedic populations who receive PT, must often engage in repetitive exercises as part of rehabilitation. To increase the sample size of future studies, investigators may want to consider expanding enrollment to a broader population of orthopedic patients, and perhaps other repetitive PT exercises. In addition, much orthopedic rehabilitation happens in outpatient settings or home-based exercises. Since outcomes related to functioning and recovery with knee replacements may be observed via follow-through for months and years, a study of adherence or motivation for exercises at home might be beneficial. A brief questionnaire was administered to obtain information about whether participants preferred exercise with or without music, along with ratings of the music interventions helpfulness and pleasantness. A mixed-methods study design that includes qualitative participant interviews would provide important information about participants’ actual experiences with the intervention—offering greater understanding about how the music was experienced during exercise and in turn, may illuminate how music is functioning to diminish pain and improve adherence. The stratification process using analgesic levels seemed to be an effective way for balancing groups in this study and is recommended for future studies when possible. Finally, there is the question of whether recorded music would be as effective as live music in this and other similar clinical applications. If the study can be adequately powered, investigators may want to consider adding recorded music condition using a three-group design. However, starting with the live music condition first and isolating the content responsible for change may be important to address prior to adding a second intervention condition. Conclusions Physical therapists and other medical professionals continue to seek ways to address pain and facilitate adherence in orthopedic rehabilitation. Research has indicated that music might be an effective nonpharmacologic component within a multimodal analgesic approach. Results from this study support further investigation of live, preferred music with guitar accompaniment to diminish observed measures of pain and improve adherence during rehabilitation exercise, specifically lower extremity pedaling exercise. These findings are similar to results from Kendelhardt (2003), and together offer preliminary data that can be used to guide the use of music in this specific area of rehabilitative exercise and PT co-treatment. Adequately powered studies are needed to provide more definitive evidence, along with isolation of specific factors that may be responsible for change. The author would like to acknowledge the therapy staff at Tallahassee Memorial Healthcare Inpatient Rehabilitation Center, especially Jim Tebo, Denean Sykes, Sue Smith, April Nobles, Sheree Porter, and Kurt Gray. The author would also like to thank the patients who participated in the study. Additionally, appreciation goes to the music therapy colleagues and professors who helped to guide and support his study, especially Dr. Jayne Standley, Miriam Hillmer, Elaine Kong, Elisa Halliley, and Laura Meehan. This study was completed in partial fulfillment of the degree of Doctor of Philosophy at Florida State University. APPENDIX A Songs Used in the Study Song Genre Artist Reference Never Too Much R&B Luther Vandross The Girl From Impanema Bossa Nova Antonio Carlos Jobim Fly Me To the Moon Jazz Frank Sinatra Back When (2×) Country Tim McGraw New York, New York Traditional Pop Frank Sinatra I Walk the Line/Ring of Fire Country Johnny Cash Take It Easy (2×) Classic Rock The Eagles Blue Suede Shoes 50s Rock Elvis Presley I Heard It Through the Grapevine R&B Marvin Gaye On the Road Again (2×) Country Willie Nelson Rolling in the Deep Pop Adele The Way You Do the Things You Do R&B The Temptations Sitting on the Dock of the Bay R&B Otis Redding Song Genre Artist Reference Never Too Much R&B Luther Vandross The Girl From Impanema Bossa Nova Antonio Carlos Jobim Fly Me To the Moon Jazz Frank Sinatra Back When (2×) Country Tim McGraw New York, New York Traditional Pop Frank Sinatra I Walk the Line/Ring of Fire Country Johnny Cash Take It Easy (2×) Classic Rock The Eagles Blue Suede Shoes 50s Rock Elvis Presley I Heard It Through the Grapevine R&B Marvin Gaye On the Road Again (2×) Country Willie Nelson Rolling in the Deep Pop Adele The Way You Do the Things You Do R&B The Temptations Sitting on the Dock of the Bay R&B Otis Redding View Large Song Genre Artist Reference Never Too Much R&B Luther Vandross The Girl From Impanema Bossa Nova Antonio Carlos Jobim Fly Me To the Moon Jazz Frank Sinatra Back When (2×) Country Tim McGraw New York, New York Traditional Pop Frank Sinatra I Walk the Line/Ring of Fire Country Johnny Cash Take It Easy (2×) Classic Rock The Eagles Blue Suede Shoes 50s Rock Elvis Presley I Heard It Through the Grapevine R&B Marvin Gaye On the Road Again (2×) Country Willie Nelson Rolling in the Deep Pop Adele The Way You Do the Things You Do R&B The Temptations Sitting on the Dock of the Bay R&B Otis Redding Song Genre Artist Reference Never Too Much R&B Luther Vandross The Girl From Impanema Bossa Nova Antonio Carlos Jobim Fly Me To the Moon Jazz Frank Sinatra Back When (2×) Country Tim McGraw New York, New York Traditional Pop Frank Sinatra I Walk the Line/Ring of Fire Country Johnny Cash Take It Easy (2×) Classic Rock The Eagles Blue Suede Shoes 50s Rock Elvis Presley I Heard It Through the Grapevine R&B Marvin Gaye On the Road Again (2×) Country Willie Nelson Rolling in the Deep Pop Adele The Way You Do the Things You Do R&B The Temptations Sitting on the Dock of the Bay R&B Otis Redding View Large APPENDIX B Follow-up Questions for the Music Group Number of Responses Percentage How pleasingwas the music? Not at all 0 0 Not very 0 0 Neutral 0 0 Fairly 1 6 Very 15 94 How helpfulwas the music? Not at all 0 0 Not very 0 0 Neutral 1 6 Fairly 1 6 Very 14 88 Did you prefer PTwith musicor without music? With music 16 100 Without music 0 0 Number of Responses Percentage How pleasingwas the music? Not at all 0 0 Not very 0 0 Neutral 0 0 Fairly 1 6 Very 15 94 How helpfulwas the music? Not at all 0 0 Not very 0 0 Neutral 1 6 Fairly 1 6 Very 14 88 Did you prefer PTwith musicor without music? With music 16 100 Without music 0 0 Note. PT = physical therapy. View Large Number of Responses Percentage How pleasingwas the music? Not at all 0 0 Not very 0 0 Neutral 0 0 Fairly 1 6 Very 15 94 How helpfulwas the music? Not at all 0 0 Not very 0 0 Neutral 1 6 Fairly 1 6 Very 14 88 Did you prefer PTwith musicor without music? With music 16 100 Without music 0 0 Number of Responses Percentage How pleasingwas the music? Not at all 0 0 Not very 0 0 Neutral 0 0 Fairly 1 6 Very 15 94 How helpfulwas the music? Not at all 0 0 Not very 0 0 Neutral 1 6 Fairly 1 6 Very 14 88 Did you prefer PTwith musicor without music? With music 16 100 Without music 0 0 Note. PT = physical therapy. View Large References Adams K. S . ( 2005 ). The effects of music therapy and deep breathing on pain in patients recovering from gynecologic surgery in the PACU (Doctoral dissertation). Florida State University, Tallahassee, Florida. Allred K. K. D. , Byers J. F. , & Sole M. L . ( 2010 ). The effect of music on postoperative pain and anxiety . Pain Management Nursing , 11 , 15 – 25 . doi: https://doi.org/10.1016/j.pmn.2008.12.002 Google Scholar Crossref Search ADS PubMed Artz N., Elvers K. T., Lowe C. M., Sackley C., Jepson P., & Beswick A. D . ( 2015 ). Effectiveness of physiotherapy exercise following total knee replacement: Systematic review and meta-analysis . BMC Musculoskeletal Disorders , 16 , 15 . doi: https://doi.org/10.1186/s12891-015-0469-6 Google Scholar Crossref Search ADS PubMed Babos M. B. , Grady B. , Wisnoff W. , & McGhee C . ( 2013 ). Pathophysiology of pain . Disease-a-Month , 59 , 330 – 358 . doi: https://doi.org/10.1016/j.disamonth.2013.05.003 Bakkum A . ( 2015 ). A randomised control trial for the restoration of functional ability in patients post total knee arthroplasty: A comparison of eccentric versus concentric cycling ergometry (Doctoral dissertation). University of Cape Town , Cape Town, South Africa . Bennell K. L., Dobson F., & Hinman R. S . ( 2014 ). Exercise in osteoarthritis: Moving from prescription to adherence . Best Practice & Research. Clinical Rheumatology , 28 , 93 – 117 . doi: https://doi.org/10.1016/j.berh.2014.01.009 Google Scholar Crossref Search ADS PubMed Bijur P. E., Latimer C. T., & Gallagher E. J . ( 2003 ). Validation of a verbally administered numerical rating scale of acute pain for use in the emergency department . Academic Emergency Medicine: Official Journal of the Society for Academic Emergency Medicine , 10 , 390 – 392 . doi:10.1111/j.1553-2712.2003.tb01355.x Google Scholar Crossref Search ADS PubMed Boldt S . ( 1996 ). The effects of music therapy on motivation, psycho logical well-being, physical comfort, and exercise endurance of bone marrow transplant patients . Journal of Music Therapy , 33 , 164 – 188 . doi: https://doi.org/10.1093/jmt/33.3.164 Google Scholar Crossref Search ADS Bradt, J. (2010). The effects of music entrainment on postoperative pain perception in pediatric patients. Music and Medicine, 2, 150–157. doi:10.1177/1943862110369913 Brown C. J. , Chen A. C. , & Dworkin S. F . ( 1989 ). Music in the control of human pain . Music Therapy , 8 , 47 – 60 . doi: https://doi.org/10.1093/mt/8.1.47 Google Scholar Crossref Search ADS Cepeda M. S. , Carr D. B. , Lau J. , & Alvarez H . ( 2006 ). Music for pain relief . Cochrane Database Systematic Review , 2 , CD004843 . doi: https://doi.org/10.1002/14651858.CD004843.pub2 Clark I. N., Baker F., & Taylor N. F . ( 2012 ). The effects of live patterned sensory enhancement on group exercise participation and mood in older adults in rehabilitation . Journal of Music Therapy , 49 , 180 – 204 . doi:10.1093/jmt/49.2.180 Google Scholar Crossref Search ADS PubMed Dahlen L., Zimmerman L., & Barron C . ( 2006 ). Pain perception and its relation to functional status post total knee arthroplasty: A pilot study . Orthopedic Nursing , 25 , 264 – 270 . doi:10.1097/00006416-200607000-00009 Google Scholar Crossref Search ADS PubMed Dallal G . ( 2017 ). Randomization plans . Retrieved 10 January, 2019 , from http://www.randomization.com Daveson B. A . ( 2001 ). Music therapy and childhood cancer: Goals, methods, patient choice and control during diagnosis, intensive treatment, transplant and palliative care . Music Therapy Perspectives , 19 , 114 – 120 . doi:10.1093/mtp/19.2.114 Google Scholar Crossref Search ADS Dileo C. , & Bradt J . ( 2005 ). Medical music therapy: A meta-analysis & agenda for future research . Cherry Hill, NJ : Jeffrey Books . Ekman E. F. , & Koman L. A . ( 2004 ). Acute pain following musculoskeletal injuries and orthopaedic surgery mechanisms and management . The Journal of Bone & Joint Surgery , 86 , 1316 – 1327 . Google Scholar Crossref Search ADS Elmallah R. K., Cherian J. J., Pierce T. P., Jauregui J. J., Harwin S. F., & Mont M. A . ( 2016 ). New and common perioperative pain management techniques in total knee arthroplasty . The Journal of Knee Surgery , 29 , 169 – 178 . doi: https://doi.org/10.1055/s-0035-1549027 Google Scholar Crossref Search ADS PubMed Elvir-Lazo O. L., & White P. F . ( 2010 ). The role of multimodal analgesia in pain management after ambulatory surgery . Current Opinion in Anaesthesiology , 23 , 697 – 703 . doi: https://doi.org/10.1097/ACO.0b013e32833fad0a Google Scholar Crossref Search ADS PubMed Fredenburg H. A. , & Silverman M. J . ( 2014 ). Effects of music therapy on positive and negative affect and pain with hospitalized patients recovering from a blood and marrow transplant: A randomized effectiveness study . The Arts in Psychotherapy , 41 , 174 – 180 . doi: https://doi.org/10.1016/j.aip.2014.01.007 Google Scholar Crossref Search ADS Gippsland Region Palliative Care Consortium Clinical Practice Group . ( 2017 ). Opioid conversion guidelines . Gippsland Region Palliative Care Consortium . Retrieved 10 January, 2019 from http://www.grpcc.com.au/wp-content/uploads/2015/05/GRPCC-Opioid-Conversion-Guidelines-FINAL-2015.pdf Holden M. A., Nicholls E. E., Young J., Hay E. M., & Foster N. E . ( 2012 ). Role of exercise for knee pain: What do older adults in the community think ? Arthritis Care & Research , 64 , 1554 – 1564 . doi: https://doi.org/10.1002/acr.21700 Google Scholar Crossref Search ADS PubMed Jabs D., Richards B. G., & Richards F. D . ( 2008 ). Quantitative effects of tumescent infiltration and bupivicaine injection in decreasing postoperative pain in submuscular breast augmentation . Aesthetic Surgery Journal , 28 , 528 – 533 . doi: https://doi.org/10.1016/j.asj.2008.07.005 Google Scholar Crossref Search ADS PubMed Johnson G., Otto D., & Clair A. A . ( 2001 ). The effect of instrumental and vocal music on adherence to a physical rehabilitation exercise program with persons who are elderly . Journal of Music Therapy , 38 , 82 – 96 . doi: https://doi.org/10.1093/jmt/38.2.82 Google Scholar Crossref Search ADS PubMed Kendelhardt A. R . ( 2003 ). Effect of live music on exercise duration, negative verbalizations, and self-perception of pain, anxiety, and rehabilitation levels of physical therapy patients (Unpublished master’s thesis). Florida State University , Tallahassee, FL . Kim S. J., & Koh I . ( 2005 ). The effects of music on pain perception of stroke patients during upper extremity joint exercises . Journal of Music Therapy , 42 , 81 – 92 . doi: https://doi.org/10.1093/jmt/42.1.81 Google Scholar Crossref Search ADS PubMed Ko Y. L. , Chang S. Y. , Lee Y. T. , & Lin P. C . ( 2016 ). The effects of music listening and education on reducing pain from total joint arthroplasty . International Journal of Studies in Nursing , 1 , 29 . doi: https://doi.org/10.20849/ijsn.v1i1.98 Google Scholar Crossref Search ADS Kremers H. M. , Larson D. R. , Crowson C. S. , Kremers W. K. , Washington R. E. , Steiner C. A. , ... & Berry D. J . ( 2015 ). Prevalence of total hip and knee replacement in the United States . The Journal of Bone Joint Surgery , 97 , 1386 – 1397 . doi: https://doi.org/10.2106/JBJS.N.01141 Google Scholar Crossref Search ADS PubMed Kurtz S. M . ( 2007 ). Projections of primary and revision hip and knee arthroplasty in the United States from 2005 to 2030 . Journal of Bone and Joint Surgery, American Volume , 89 , 780 – 785 . doi: https://doi.org/10.2106/JBJS.F.00222 Lamplot J. D., Wagner E. R., & Manning D. W . ( 2014 ). Multimodal pain management in total knee arthroplasty: A prospective randomized controlled trial . The Journal of Arthroplasty , 29 , 329 – 334 . doi: https://doi.org/10.1016/j.arth.2013.06.005 Google Scholar Crossref Search ADS PubMed Lee J. H . ( 2016 ). The effects of music on pain: A meta-analysis . Journal of Music Therapy , 53 , 430 – 477 . doi: https://doi.org/10.1093/jmt/thw012 Google Scholar Crossref Search ADS PubMed Lingard E. A., Katz J. N., Wright E. A., & Sledge C. B. ; Kinemax Outcomes Group . ( 2004 ). Predicting the outcome of total knee arthroplasty . The Journal of Bone and Joint Surgery. American Volume , 86-A , 2179 – 2186 . doi:10.2106/00004623-200410000-00008 Google Scholar Crossref Search ADS PubMed Madson A. T., & Silverman M. J . ( 2010 ). The effect of music therapy on relaxation, anxiety, pain perception, and nausea in adult solid organ transplant patients . Journal of Music Therapy , 47 , 220 – 232 . doi: https://doi.org/10.1093/jmt/47.3.220 Google Scholar Crossref Search ADS PubMed McCaffrey R., & Locsin R . ( 2006 ). The effect of music on pain and acute confusion in older adults undergoing hip and knee surgery . Holistic Nursing Practice , 20 , 218 – 224. doi:10.1097/00004650-200609000-00002 . Google Scholar Crossref Search ADS PubMed McLean S. M. , Burton M. , Bradley L. , & Littlewood C . ( 2010 ). Interventions for enhancing adherence with physiotherapy: A systematic review . Manual Therapy , 15 , 514 – 521 . doi: https://doi.org/10.1016/j.math.2010.05.012 Google Scholar Crossref Search ADS PubMed McLeod W. D. , & Blackburn T. A . ( 1980 ). Biomechanics of knee rehabilitation with cycling . The American Journal of Sports Medicine , 8 , 175 – 180 . doi: https://doi.org/10.1177/036354658000800306 Google Scholar Crossref Search ADS PubMed Meier W., Mizner R. L., Marcus R. L., Dibble L. E., Peters C., & Lastayo P. C . ( 2008 ). Total knee arthroplasty: Muscle impairments, functional limitations, and recommended rehabilitation approaches . The Journal of Orthopaedic and Sports Physical Therapy , 38 , 246 – 256 . doi: https://doi.org/10.2519/jospt.2008.2715 Google Scholar Crossref Search ADS PubMed Melzack R . ( 1999 ). From the gate to the neuromatrix . Pain , 82 , S121 – S126 . doi: https://doi.org/10.1016/S0304-3959(99)00145-1 Google Scholar Crossref Search ADS Melzack R., & Wall P. D . ( 1965 ). Pain mechanisms: A new theory . Science (New York, N.Y.) , 150 , 971 – 979 . Google Scholar Crossref Search ADS PubMed Merle C. , Brendle S. , Wang H. , Streit M. R. , Gotterbarm T. , & Schiltenwolf M . ( 2014 ). Multidisciplinary treatment in patients with persistent pain following total hip and knee arthroplasty . The Journal of Arthroplasty , 29 , 28 – 32 . doi: https://doi.org/10.1016/j.arth.2013.05.004 Google Scholar Crossref Search ADS PubMed Moucha C. S., Weiser M. C., & Levin E. J . ( 2016 ). Current strategies in anesthesia and analgesia for total knee arthroplasty . The Journal of the American Academy of Orthopaedic Surgeons , 24 , 60 – 73 . doi: https://doi.org/10.5435/JAAOS-D-14-00259 Google Scholar Crossref Search ADS PubMed Nett M. P . ( 2010 ). Postoperative pain management . Orthopedics , 33 ( Suppl. 9 ), 23 – 26 . doi: https://doi.org/10.3928/01477447-20100722-60 Google Scholar Crossref Search ADS PubMed Puntillo K. A., Morris A. B., Thompson C. L., Stanik-Hutt J., White C. A., & Wild L. R . ( 2004 ). Pain behaviors observed during six common procedures: Results from Thunder Project II . Critical Care Medicine , 32 , 421 – 427 . doi: https://doi.org/10.1097/01.CCM.0000108875.35298.D2 Google Scholar Crossref Search ADS PubMed Rastogi R. R . ( 2007 ). A cross-sectional look at patient concerns in the first six weeks following primary total knee arthroplasty . Health and Quality of Life Outcomes , 5 , 48 . doi: https://doi.org/10.1186/1477-7525-5-48 Google Scholar Crossref Search ADS PubMed Sloman R., Rosen G., Rom M., & Shir Y . ( 2005 ). Nurses’ assessment of pain in surgical patients . Journal of Advanced Nursing , 52 , 125 – 132 . doi: https://doi.org/10.1111/j.1365-2648.2005.03573.x Google Scholar Crossref Search ADS PubMed Strassels S. A., Chen C., & Carr D. B . ( 2002 ). Postoperative analgesia: Economics, resource use, and patient satisfaction in an urban teaching hospital . Anesthesia and Analgesia , 94 , 130 – 137. doi:10.1213/00000539-200201000-00025 . Google Scholar PubMed Thaut C. P . ( 2014 ). Patterned sensory enhancement (PSE) . In M. H. Thaut & V. Hoemberg (Eds.), Handbook of neurologic music therapy (pp. 106 – 115 ). Oxford University Press, New York, NY . Tighe P., Buckenmaier C. C. 3rd , Boezaart A. P., Carr D. B., Clark L. L., Herring A. A. , … Reisfield G. M . ( 2015 ). Acute pain medicine in the United States: A status report . Pain Medicine (Malden, Mass.) , 16 , 1806 – 1826 . doi: https://doi.org/10.1111/pme.12760 Google Scholar Crossref Search ADS PubMed Trout K. K . ( 2004 ). The neuromatrix theory of pain: Implications for selected nonpharmacologic methods of pain relief for labor . Journal of Midwifery & Women’s Health , 49 , 482 – 488 . doi: https://doi.org/10.1016/j.jmwh.2004.07.009 Google Scholar Crossref Search ADS PubMed Williamson A., & Hoggart B . ( 2005 ). Pain: A review of three commonly used pain rating scales . Journal of Clinical Nursing , 14 , 798 – 804 . doi: https://doi.org/10.1111/j.1365-2702.2005.01121.x Google Scholar Crossref Search ADS PubMed © American Music Therapy Association 2019. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Live Music Therapy During Rehabilitation After Total Knee Arthroplasty: A Randomized Controlled Trial
JF - Journal of Music Therapy
DO - 10.1093/jmt/thy022
DA - 2019-02-16
UR - https://www.deepdyve.com/lp/oxford-university-press/live-music-therapy-during-rehabilitation-after-total-knee-arthroplasty-tb6OagYeAU
SP - 61
VL - 56
IS - 1
DP - DeepDyve
ER -