Secondary Logo

Journal Logo


The Use of Music Intervention in Healthcare Research

A Narrative Review of the Literature

Tang, Hsin-Yi (Jean); Vezeau, Toni

Author Information
doi: 10.1097/JNR.0b013e3181efe1b1
  • Free



The purpose of this article was to provide a narrative review of literature to explore how "music therapy" has been used in healthcare research to promote healing in adult populations. Specifically, this article identified the populations, the conditions, and the types of music that have been used in healthcare research. The critique for "music therapy" has always been its lack of scientific basis. A solid theoretical framework and methodology provides a root for the intervention. This review explored how music interventions were operationalized and reported as well as their effectiveness. This article also provided specific recommendations on the use of music in healthcare research studies and suggests improvements in scholarly dissemination to further stimulate a discussion of the science behind music as a holistic intervention strategy for healing.


The use of music as a healing modality dates to ancient times (Schullian & Schoen, 1948). The first documented use of music in healing is written in the Kahun medical papyrus (ca. 1825 BCE, ancient Egypt; Standley & Prickett, 1994). In medieval times, illness was perceived as disharmony of the mind-body system, and music was used as a remedy to dispel evil spirits and sickness to restore harmony and health. By the late 18th century, practitioners evaluated patients' perceptions of music interventions and also began exploring the effects of music on the human body in terms of variables such as heart rate, blood pressure, and respiration (Standley & Prickett, 1994). Music therapy was introduced into the healthcare setting in the 20th century, when music therapists visited hospitalized war veterans and resultant improvements in patient physical and psychological health were noted. Since then, healthcare settings have incorporated music therapy as part of nonpharmacological care (American Music Therapy Association, 2005). Research findings from the 1970s through the 1990s, although not necessarily pertaining to music interventions alone, provide a foundation for a scientific model of the mind-body. Most of the research conducted during these three decades took place in the laboratory in a "pure research" context. Additional physiological responses (brain waves, cerebral blood flow, muscle tension, and peripheral sweating level; Mindlin & Evans, 2009) and biological markers, such as cortisol levels in the saliva and the blood stream that reflect sympathetic or parasympathetic activities and emotional arousal, have more recently been studied. Research describes physiological data along with subject reports on their emotional response to music, providing a more holistic grasp of the phenomena. Currently, research interest in music and healthcare has left the pure research environment to address specific healthcare concerns in more natural settings (Tang et al., 2009; Tang, Harms, & Vezeau, 2008).

Although music has been investigated in many different populations within the healthcare setting (Dileo, 2006; Standley, 1986, 2002), there has been limited review of literature analyzing the use of music as a methodology. The goal of this article was to provide specific methodology recommendations for future research. In this article, music therapy is defined as "the clinical and evidence based use of music interventions to accomplish individualized goals within a therapeutic relationship by a credentialed professional who has completed an approved music therapy program" (American Music Therapy Association, 2005).


To retrieve appropriate literature, four databases were searched for clinical randomized controlled studies: PubMed, PsycINFO, CINAHL, and Cochrane Library. A comprehensive search was undertaken using the MeSH term "music therapy" in the title field, with the following search criteria: adults 19 years and older, humans, clinical randomized controlled studies, and English. On the basis of search criteria, a total of 33 articles were retrieved from PubMed, PsycINFO, and CINAHL. Four meta-analysis articles on the use of music therapy in adult populations (schizophrenia, pain relief, depression, and dementia) were retrieved from the Cochrane Library. Fifteen articles were cross referenced in multiple databases, especially between PubMed and PsycINFO. This article focuses on the 33 studies that met the search criteria.

Studies were compared on the basis of the population studied, the size of the sample, how music was operationalized, the outcome measures, and the effectiveness of the intervention. Citations and brief descriptions of included studies are presented illustrated in Table 1.

Citations and Brief Intervention Descriptions of Retrieved Studies
Citations and Brief Intervention Descriptions of Retrieved Studies, continued
Citations and Brief Intervention Descriptions of Retrieved Studies, continued

The Review Analysis: Music Intervention in Healthcare Research

The goal of this narrative review of literature was to answer five questions:

  1. In what populations and conditions has music intervention been studied?
  2. What specific kinds of music have been used for study interventions?
  3. How have music interventions been operationalized (delivery method, frequency, dosage, duration, setting, control groups, supervision)?
  4. What metrics were used for outcome measures?
  5. Have music interventions been effective?


In What Populations and Under What Conditions Has Music Intervention Been Studied?

Music has been studied under various conditions in a multitude of populations such as various age groups and health conditions. In the 33 reviewed articles, studied populations included healthy adults (n = 2), through patients with serious illnesses and those facing major medical procedures; schizophrenia (n = 2), depression (n = 1), terminal illness (n = 2), radiation therapy (n = 2), pregnancy and delivery (n = 4), cardiovascular rehabilitation (n = 1), cardiac surgery (n = 1), myocardial infarction (n = 2), surgical procedures (n = 3), gastrointestinal procedures (n = 2), ventilator care (n = 2), stem cell transplant (n = 1), multiple sclerosis (n = 1), tension headache (n = 1), Parkinson's disease (n = 1), and dementia (n = 5). The most commonly studied population was people with dementia.

Conditions and related symptoms that interventions aimed to alleviate ranged from anxiety (n = 8), perceived stress or stress markers (n = 5), agitation or behavioral problems (n = 4), pain (n = 2), quality of life or length of hospital stay (n = 2), psychotic symptoms (n = 2), cognitive functions (n = 1), degree of sickness (n = 1), and skin symptom relief (n = 1). Of significant note, anxiety relief was more often addressed than the relief of pain or other physiological symptoms. This situation is true even in the surgical and labor delivery studies reviewed.

Among the 33 studies retrieved, 27 listed sample selection criteria. Criteria selection methodology was not stated clearly in six articles. There was no common inclusion or exclusion criteria noted among the 33 studies. Listening abilities, a vital consideration in music intervention, were reportedly screened in 10 (33%) of 33 studies. This is a critical issue for future studies because adequate subject hearing is essential to study validity. It is concerning to note that some studies neither assessed nor reported on hearing loss. It is well documented in the literature that 31.5 million people have hearing difficulties of which their care providers may or may not be aware. The rate of hearing loss in the United States has grown significantly, from 6.8% in 2000 to 9.9% in 2005 (the most current figure available). Hearing screening by primary caregivers has declined, and certain groups, such as women and those in lower economic strata, have shown increasing incidence of hearing loss since 2000 (Kochkin, 2005).

Sample sizes in articles surveyed were generally small, ranging from n = 14 to n = 236, and most studies were midsized (mean = 55, SD = 42). The power analysis and determination process for sample size was generally not reported in the studies. Gender was equally represented in most studies with some notable exceptions. For example, in studies on populations of pregnant women and in persons with dementia (Chang, Chen, & Huang, 2008; Chang & Chen, 2005; Ledger & Baker, 2007; Raglio et al., 2008; Suzuki et al., 2004), participants were either solely or primarily female, whereas there were relatively more male participants in studies on myocardial infarction patients. Although it is understandable that studies on pregnant and nursing home dementia populations focused on primarily female participants, the preponderance of male participants in the studies (Guzzetta, 1989; White, 1992) with myocardial infarction condition is unexplained.

In terms of ethnicity, Caucasian and Asian were the predominant ethnic groups, respectively, in studies done in Western and Asian countries. Black, Hispanic, and Native Americans had little to no representation in these studies. Three of these studies did not provide details on participant demographic characteristics (Browning, 2001; Cassileth, Vickers, & Magill, 2003; Svansdottir & Snaedal, 2006). Ethnicity, however, may represent an important factor of influence because ethnicity and culture may influence the outcome of music intervention, as discussed in the following section.

What Specific Kinds of Music Have Been Used As Study Interventions?

Among the retrieved studies, employed music interventions can be categorized into two types: passive (receptive) and active. The terms passive (receptive) versus active were commonly used among studies that employed a music intervention. Passive (receptive) music intervention commonly involved subjects in a resting position listening to music, whereas active music intervention was usually carried out in a group format in which subjects actively produce music or have an active role in music intervention. Active interventions in the studies reviewed included singing, instrument playing (such as drumming), group discussion of music, lyric analysis, song writing, moving in response to music, dancing, and improvisation (Bittman et al., 2001; Hilliard, 2003; Horne-Thompson & Grocke, 2008; Ulrich, Houtmans, & Gold, 2007). Among the retrieved studies, one third (11/33) were defined as "active." It should be noted that the active music intervention was often used with cognitively or mentally incapacitated populations, such as people with dementia or psychotic symptoms. The active style was also used with the terminally ill and patients having brain surgery. Of these 11 studies, only one provided a rationale to justify the decision between the active versus the passive (receptive) music intervention. A study by Ledger and Baker (2007) indicated that group music therapy was considered effective in promoting interaction and feelings of belonging and so may be more therapeutic in engaging cognitively impaired participants in group activities that involve behavioral modeling (singing, instrument playing) or large muscle movement (dancing). Similarly, an active approach to music intervention may be more therapeutic for participants who are going through end-of-life treatment or scheduled for a potentially life-threatening surgery because group intervention may serve as a support group.

Two studies are worth noting for researchers who plan to adopt an active music intervention in future studies. In Svansdottir and Snaedal (2006), a brief description was provided on how the music intervention was implemented in a group of dementia clients who often had different levels of cognition from day to day. The study design took an inclusive approach that allowed participants to engage at their own comfort level to achieve a desired therapeutic effect by varying the intensity of the music intervention. In the second study, Bittman et al. (2001) conducted a single trial experiment in healthy adults using drumming. In the report, the authors described how the drumming experience was progressively introduced to participants. For example, some participants started by listening to drum music followed by the use of shaker eggs to establish a sense of rhythm and teamwork before receiving instructions in playing the hand drum. The drumming intervention was complicated by varying beat, tempo, and volume. The thorough description included for this intervention provided informative data regarding its methodology and intervention design. In addition, this study broadened the notion of "music therapy" to include rhythm. In a study conducted by Pacchetti et al. (2000), a program that encompassed several active music intervention methods was delivered to participants with Parkinson's disease. The study indicated that a comprehensive program that stimulates different sensory pathways may promote socialization, involvement with the environment, expression of feelings, awareness, and responsiveness. The program included 10 minutes of relaxation music listening and visualization; 15 to 20 minutes of choral singing, facial expression, and breathing; 30 minutes of rhythmic movements; 30 to 40 minutes of music improvisation; and 10 minutes of conversation. Instruments such as xylophones, drums, and wood blocks were also available for participants to play freestyle. Although it is unclear whether the music was involved during the "rhythmic movement" and the "conversation" sections, the report offers a sample protocol that encompasses multiple activity methods.

In the remaining 22 studies, subjects passively listened to music. Of these, only 15 specified the type (category) of music used. Specified selections varied widely and included classical, country, New Age, easy listening, pop, rock, religious, "era [sic]," motion picture soundtrack, jazz, lullabies, nature sounds, and crystal sounds. Music was preselected by the therapist or the researcher in half of the studies, and in 11 studies, participants chose the music either with the therapist's assistance or from the selection offered by the therapist. This approach may be problematic.

The literature suggests that an individual's musical preference is a strong influencing factor on intervention outcome. In one study with two intervention groups, Experimental Group 1 received New Age music, whereas Experimental Group 2 self-selected music from classical, country, pop, and dance categories. The group that was allowed to self-select music achieved a significantly better outcome than the group that was assigned New Age music. The control group (unadjusted ambient sounds of the operating room) showed the least change (Leardi et al., 2007).

It is a critical challenge for researchers to determine how music is selected when designing a study that involves music as the primary intervention. When deciding on the music selection, the researcher should consider both individual subject characteristics as well as study purpose. A semistructured approach that allows participants to voice their music preference under the framework established by researchers may be an appropriate research strategy. Examples of how this may be achieved can be reviewed in Clark et al. (2006) and Hanser and Thompson (1994). Both provide in-depth descriptions on how individual music preferences were taken into consideration. Therapists or researchers met with participants to discuss music preferences before a decision was made. This approach allowed participants autonomy in music selection while protecting intervention integrity.

Cultural background is also recognized as highly associated with individual music preference. In 2004, Lai conducted a study to examine the relationship between music preference and relaxation in elderly Taiwanese (mean age = 65.7 years, SD = 5.2 years). Heart rate, respiration, and finger temperature were measured to index the degree of relaxation. Participants were asked to rate their preference among six types of music: Chinese orchestral, western harp, western piano, slow jazz, western synthesizer, and western orchestral. Although results showed no differences in degree of relaxation among music type, Chinese orchestral music was requested most often. A particularly positive example in which participant backgrounds and music selection preferences are noted is in Shiraishi (1997), who clarifies the decision-making process of a music intervention by allowing the cross-referencing of participant music preferences and ethnicity. In addition to individual preference and cultural influences, age is another factor to be considered when designing a music intervention study. Gibbons (1977), a well-cited early researcher of music theory in relation to the older people, indicated that popular music was the most therapeutic category of music for adults at 25 years of age.

Many of the studies included in this review describe how several types of music were made available from which participants could choose. However, few reported on frequencies for the ultimate music selection, which would assist in understanding study results and with replication. A study that provided helpful information on music selection was that of Smith, Casey, Johnson, Gwede, and Riggin (2001), who studied 19 male subjects (mostly Caucasian, aged 39 to 78 years, mean age = 62.2 years) who were receiving medical radiation therapy. Participants were allowed to choose music from several categories, which they continued with throughout the intervention. Thirty-seven percent of the participants chose country and western, 21% chose big band, 16% preferred easy listening, 11% selected Spanish, 11% chose religious music, and 5% chose classical music. This demonstrates a wide range of music preference. The authors indicated that big band music was the most requested category. However, when participants were told that big band music was not available, country and western was consistently the second choice of participants. Although it can be considered a study limitation that the participants' first choice of music was not made available, the authors should be applauded for reporting the statistics and including the information on music preference, in that it demonstrates that subject groups may have unanticipated and broad preferences.

A relatively strong depth of data on participant music selections was provided in six studies, which identified the music selection categories and the specific music sound tracks used (Andrada et al., 2004; Bruer, Spitznagel, & Cloninger, 2007; Chang et al., 2008; Mandel, Hanser, Secic, & Davis, 2007; McKinney, Antoni, Kumar, Tims, & McCabe, 1997; Shiraishi, 1997). Such provided helpful information to understand study results and to replicate intervention.

Conceptual and theoretical frameworks were generally lacking in most scientific reports to explain music selection and other aspects of study methodologies. A notable exception that provides a useful model for a conceptual framework is a study by Chang et al. (2008). The researchers took an alternative approach to establish a theoretical foundation for study intervention (Chang et al., 2008). The team conducted a pilot study with two cases to test the effectiveness of the selected music from a physiologic perspective. Researchers noted that most music chosen had tempos in the range of 60 to 80 bpm, which mimicked the adult human heart rate and facilitated relaxation more readily. A statistical association between heart rate and relaxation, however, was not provided in the article. Although the term "mind-body" was not highlighted, the rationale of music providing a rhythmic rate of 60 to 80 bpm was also adopted in two other studies (Sendelbach, Halm, Doran, Miller, & Gaillard, 2006; White, 1992).

Researchers have also found people tend to favor music with a tempo in the range of 70 to 100 bpm (Dowling & Harwood, 1986; Rosenfeld, 1985). Although the link between the music tempo preference and the heart rate was not identified at the time, the range of 70 to 100 cycles/minute is similar to the range of adult heart rates in a state of physical rest. In the late 1980s, researchers started to explore the correlation between personal tempo such as heart rhythm and preferred music tempo. The research findings of Iwanaga (1995a, 1995b) suggest that subjects preferred music tempo in harmonic relation to their own heart rate. Iwanaga clearly laid out a direction for scientific development in this area with useful references.

Furthermore, a study not included in this review by LeBlanc, Colman, McCrary, Sherrill, and Malin (1988) on 926 students ranging from third grade in elementary school to university showed that younger participants tended to prefer a faster tempo. Similar findings were cited in Buchanan's report in his review of Walter's (1983) classic work (Buchana, 1988).

The preference in music tempo is not only individually unique and age related but also influenced by external factors such as mental activities. A study by Holbrook and Anand (1990) reported that a preference for faster music tempo was associated with elevated heart rate after mental arithmetic task execution. These findings seem to support the concept that individuals intuitively prefer a music tempo that is in synchronization with their current biological rhythm. It was noted that listening to music with a tempo of 50 to 60 bpm, which is at the low-range heart rates for most of adults, may promote relaxation and help achieve homeostasis (Hoffman, 1997). This may provide the strongest rationale for subject self-selection within a study.

How Has The Music Intervention Been Operationalized?

In operationalizing music as a variable, the following areas must be addressed: delivery method, frequency/dosage/duration, research setting and controls for environmental factors, use of control groups, and supervision by the research team.

Delivery Method

Although the retrieved studies consistently identified music as the intervention, there were wide variations as to how music intervention was operationalized. As indicated in the previous section, intervention types may be categorized into two formats, namely, passive (receptive) and active. Consistent with the previous given definition, passive (receptive) music interventions were often implemented through headphones privately. Active music interventions were usually delivered in a group format in an open space in a room.

Although two studies did not provide information on how the intervention was delivered, over half of the studies (19/33, 58%) indicated that the intervention was delivered by trained music therapists; a smaller number were implemented by researchers (9/33, 27%), health providers such as nurses (1/33, 3%), and massage therapists or occupational therapists (2/33, 6%). Four (12%) of 33 retrieved studies reported that the intervention was delivered on a self-administered basis after initial instruction.

Frequency, Dosage, and Duration

The intervention frequency was highly variable and may be categorized into three types: (a) one time, single dose; (b) multiple sessions over a short term; and (c) multiple sessions over a long term. In one-time, single-dose studies, session dosing depended either on the length of medical procedure (Andrada et al., 2004; Chang et al., 2008; Chlan, 1998; Chlan, Evans, Greenleaf, & Walker, 2000; Hilliard, 2003; Leardi et al., 2007; McRee, Noble, & Pasvogel, 2003) or on another parameter determined by researchers or therapists.

There was wide variation in dosing, and the rationale for dosing was not specified in any of the reviewed studies. The reported one-time single-dose intervention ranged from 25 to 88 minutes. The 30-minute session was the most common dosing intervention (Table 1).

Among multiple-session short-term studies, the number (frequency) of sessions implemented was either not identified or reportedly ranged between 2 and 30 sessions, with an approximate mean of 10 sessions across the sampled studies. The length of each session was either not reported or ranged from 20 to 120 minutes/session, with an approximate average of 48 minutes across sampled studies. There was also a wide variation in the percentage of interventions that were completed as designed in the reviewed studies. The actual dosing that was delivered, when reported, ranged from completion rates as low as 21% up to 100%. In other words, some participants only finished 21% of the intended intervention sessions. The rigor of study results relies on intervention completeness. Lack of completeness leads to lack of clarity in actual effectiveness of music intervention.

Study duration was either not reported (n = 1) or ranged between 1 day and 12 weeks. There were few multiple-session long-term studies; one had 30 to 45 minutes/session for approximately 45 sessions over a 1-year period (Ledger & Baker, 2007). Another study had 8 to 10 sessions with unspecified session dosing and follow-ups each 3 months for 1 year (Schmid & Aldridge, 2004). Overall, the rationale of how decisions were made in terms of intervention frequency, dosing, and duration was not clearly reported across all sampled studies.

Setting or Control for Environmental Factors

In general, few details were given on how the environment was controlled to promote the intended intervention effect. Very few studies described the setting in which the intervention took place, which makes evaluation of these studies difficult. Among the retrieved studies, reports by Chlan (1998) and Wong, Lopez-Nahas, and Molassiotis (2001) effectively addressed environmental factors. Both studies used music intervention to promote relaxation for ventilator-dependent patients in the intensive care units. In the two studies, environmental measures such as dim light and the use of privacy curtains were reported. Logically, environmental factors, if not well controlled, can act as confounding variables, again resulting in lack of confidence in the study effectiveness. The two studies cited above provided a good example of how to address the intervention setting.

Use of Control Groups

Among the retrieved studies, treatment for control groups (28/33, 85%) was often described as "no treatment," "resting," or "standard/routine care," especially in the inpatient setting. In some studies, the control group was on the "wait list" before receiving the intervention that was delivered to the experimental group in the study. In one study, a crossover (switch group) design was used, and the control group viewed a movie as the intervention in lieu of the music intervention (Bruer et al., 2007). Other protocols used in the control group included "volunteer visit" (Horne-Thompson & Grocke, 2008) or "reading newspaper and magazine" in the same environment with exposure to the typical ambient sounds in the setting (Bittman et al., 2001).


An interesting phenomenon was observed in the findings of retrieved longitudinal studies. The direct supervision from the researcher or therapist seemed to be an influential factor for adherence to the study protocol in self-administered music interventions and in subject outcomes. Horne-Thompson and Grocke (2008) noted that participants who received direct supervision from the researcher had a higher completion rate. A similar finding was found in another study in which the intervention group who received a home visit attained a significantly better outcome than the intervention group that received weekly telephone visits (Shiraishi, 1997). This could be an extension of the Hawthorne effect or could indicate that more intervention was delivered with a higher completion rate, leading to an improved outcome.

What Metrics Were Used for Outcome Measures

Outcome measures in the 33 retrieved articles involve two types: psychological and physiological or biological. The psychological measures in these studies represent both subjective and objective data. However, most psychological measures in the retrieved studies were subjective, which was based on participant self-reporting. Overall, most of the subjective psychological outcome measures data indexed degree of relaxation, anxiety or stress level, or perceived quality of life. The State-Trait Anxiety Inventory was the most commonly used scale for anxiety measurement. The State-Trait Anxiety Inventory was used in 10 (30%) of 33 retrieved studies, followed in frequency by the Profile of Mood States. Remaining measures included depression scales and quality of life measures. On the other hand, six studies used objective (observational) measures because of participants' impaired cognition. In these studies, the outcome index evaluated the degree of agitation and behavioral change in people with dementia. Psychological measurements in the retrieved studies were generally appropriate for the factors that the researchers attempted to index.

In addition to the psychological measure, physiological or biological measures were also used in 13 (39%) of 33 retrieved studies. A biological measure is defined in this article as a method that involves body fluids (e.g., blood, saliva), and a physiological measure is defined as nonbiohazard, noninvasive method such as blood pressure or brain waves. Physiological or biological measures were generally appropriate and reflective of stress responses.

Six of 33 retrieved studies used biological stress markers to evaluate the intervention effectiveness. The most commonly used biological maker was cortisol level (Bittman et al., 2001; Leardi et al., 2007; McKinney et al., 1997; McRee et al., 2003; Suzuki et al., 2004). Six of 33 retrieved studies used physiological markers such as vital signs (Chlan, 1998; Guzzetta, 1989; Mandel et al., 2007; McRee et al., 2003; Sendelbach et al., 2006; White, 1992; Wong et al., 2001) and cerebral blood flow (Lazaroff & Shimshoni, 2000) as outcome measures. Among the 13 studies that used biological or physiological measurements, 12 also used some form of psychological method to obtain combined measurements.

Most single-dose studies used a before and after design. The outcome measure was taken once after an intervention. For mid- to long-term studies, the exit outcome measurement ranged from 1 day to 1 year postintervention, depending on the study design. There is an important observation to be noted here. Suzuki et al. (2004) reported that biological measures such as endocrine level were sensitive to circadian rhythm. Sample collection time and timings were well documented in their report. In the two studies with biological or physiological measures, data were continuously measured every 5 minutes throughout the intervention (Chlan, 1998; Wong et al., 2001). These continuous data allow for an enhanced overall view of treatment effectiveness and also provide information about trend change over the course of intervention sessions.

Has Music Intervention Been Effective?

The effectiveness of the music intervention was reportedly successful in most of the retrieved studies. All sampled studies except one (Smith et al., 2001) reported on the efficacy of the music intervention. There was consistency in the intervention efficacy in terms of both psychological and physiological or biological measures in short- and long-term interventions when the participants continued to adhere to the intervention.

Although the effectiveness of the music intervention appears to be promising, as reported in the retrieved studies, Cochrane meta-analyses resulted in different conclusions. There are four meta-analysis reports on music intervention in adult populations. The clinical focus areas include schizophrenia, pain relief, depression, and dementia:

  1. The Cochrane review of the music therapy for schizophrenia or schizophrenia-like illness indicated that music intervention as an addition to standard care appeared to improve patient mental state. However, the intervention completion rate was either not reported or reportedly low among the studies. The Cochrane review suggested that future research should explore dosing and the long-term effects of the music intervention in the schizophrenia condition (Gold, Heldal, Dahle, & Wigram, 2005).
  2. In pain relief studies, it was reported that the music intervention helped reduce pain. However, the Cochrane review commented that the degree of pain relief was not large enough to conclude clinical significance. The effectiveness of music intervention in pain relief is therefore unclear (Cepeda, Carr, Lau, & Alvarez, 2006).
  3. The Cochrane review on the effectiveness of music intervention in depression showed that although four (80%) of five retrieved studies reported significant reductions in depressive symptoms, a meta-analysis was unable to be conducted due to the significant variations in methodology among sampled studies. More randomized controlled studies with strong methodologies that adopt a variety of intervention formats were recommended (Maratos, Gold, Wang, & Crawford, 2008).
  4. Among the four conditions, the Cochrane review reserved its strongest criticisms for dementia studies. The review indicated that it was difficult to determine the effectiveness of the music intervention in relieving dementia-related behavioral and cognitive symptoms because of the poor quality of the research methodologies in the retrieved studies (Vink, Birks, Bruinsma, & Scholten, 2004). Specifically, the major critique in the review is that information about the randomization process was not clearly reported, and the psychometric integrity (reliability and validity) of the measurement tools was lacking in most of the studies included in the meta-analyses, which made evaluation of the outcomes impossible. One review suggested the use of Consolidated Standards of Reporting Trials (CONSORT) guidelines for future study reporting (Vink et al., 2004). The CONSORT guideline was established in 2001, with a goal of improving reporting quality in randomized controlled trial studies. The CONSORT checklist prompts authors to provide complete and thorough information on the study design, intervention application, data analysis, and interpretation (Moher, Schulz, & Altman, 2001).

The finding of this article was consistent with Cochrane meta-analyses regarding potentially problematic intervention designs and the lack of clear reporting in the studies, especially with dementia patients. For example, the crossover design was used in some dementia studies, but the washout period between two interventions was not clearly described, or it was not taken into consideration in data analysis. It was concluded that the method of statistical analysis was a concern in a majority of the dementia studies reviewed (4/5, 80%; Vink et al., 2004). In addition, some outcome measures in these studies used observational data that were indexed by the caregivers. Outcome measure psychometrics, for example, interrater reliability, were not clearly identified. In summary, the Cochrane review could not comment on the effectiveness of the music intervention because of the lack of scientific integrity, especially in studies of dementia.


Music intervention has been mostly studied in psychiatric conditions and mental health-related symptoms in the clinical setting. There is a lack of homogeneity in sample representation among the retrieved studies. Unexplained gender imbalance was noted in two of the cardiovascular studies. In addition, demographic information, such as ethnicity and culture, is not always fully reported. Further, it is unclear whether subjects had access to a full intervention. In particular, studies did not provide clarity on the issue of whether their subjects could hear the music intervention adequately. In some studies, the cognitive ability of participants to respond to music remained unclear. Although some studies listed hearing and cognition in the inclusion or exclusion criteria, the actual screening process was not clearly documented. While the clinical contexts in these studies were well described and deemed appropriate for a music intervention, the same cannot be said of the subjects themselves because of the lack of demographic data and data on subject functionality relevant to their ability to hear, to understand, and to respond to music.

The method of the music intervention in terms of an active versus passive (receptive) approach is an important consideration, and the decision-making process around this issue was not fully discussed in most of the retrieved studies. There is little rationale given for the methodology chosen. The reasoning undergirding music selection was generally lacking, except for a few notable exceptions. This is highly problematic because the literature suggests that what a subject may find relaxing or choose for themselves as potentially relaxing, for example, may vary widely from that which researchers may select. There was a fairly broad variety of music selection for a widely heterogeneous sample. Therefore, it becomes difficult to identify the rigor in the methodology if a clear case of matching a music intervention to subjects has not been made clear.

One may ask the significance of having consistency in music selection either within the same study or across studies. Perhaps the goal may be to take extreme care when matching music to subjects. The type of music may only matter when the music suits the subjects and the design is nested within the conceptual framework. In addition, researchers may accomplish the reasoning for music selection using a sound conceptual framework and interdisciplinary research, adding rigor to methodological choices.

No study in this review reported interdisciplinary collaboration between the arenas of science and the humanities that intersect with music as a phenomenon of study, such as music theory and music application. Music specialists, when used in a study, were from the field of therapy rather than from areas of academics, research, or theory, which may have particular insight into the human response elements to music. This is an unexplored and a potentially rich area for future study.

Regarding operationalization of music interventions within the studies under review, there was general consistency in the 30-minute intervention dosing, albeit, once more, minus rationale. The music intervention was conducted in a variety of settings with different intervention durations across retrieved studies. It is unknown how intervention duration affects the efficacy of interventions used because efficacy was noted in both one-time single treatment and multiple-session longitudinal studies. Sustained results following the end of the study were generally missing; these data would seem to be critical to reporting on the effectiveness of an intervention. The lack of detailed information on intervention design leads to problems with understanding results and intervention replication.

Intervention completion rates were either highly variable or not reported. Again, this impairs study result interpretation. Treatment adherence may be improved by direct supervision. Studies reviewed reported that high levels of supervision resulted in higher rates of completion.

Many studies focused on reporting outcome measures and did not provide adequate descriptions of interventions. This may have been editorially beyond the control of researchers. Although a research report should highlight the outcome measure, a theoretical framework underpinning the intervention should be provided, especially in newer areas of research such as music intervention. Simply, it is unknown at this time if music is effective due to the necessity of increased scientific stringency.

Only 36% of retrieved studies noted the use of combined measures. It is recommended to use both psychological and physiological or biological measures to increase concurrent or criterion validity. Although individual studies reported intervention efficacy, meta-analysis (Vink et al., 2004) faulted the lack of scientific rigor in the studies, making judgment of efficacy premature.


With respect to the preliminary work done, the authors make the following recommendations to enhance research using music as an intervention within a healthcare context:

  1. Subject demographics are integral to an intervention study. First, full demographics, inclusive of age, culture, and ethnicity, should be assessed and documented. Such data should be considered in relation to the sampling plan for the study and for the decision making on music selection.
  2. In addition to focusing on psychological symptoms, the researchers should consider appropriate physiological foci such as heart rate and brain waves for additional measurement. When studying such variables as stress, anxiety, and depression, there is ample literature to support the selection of appropriate physiological (such as vital signs) and biological measures (such as serum cortisol level). This may add concurrent and criterion validity to the study.
  3. Hearing and cognition screening should be carefully planned before the study intervention and be clearly documented, especially when studying populations wherein deficit rates may be higher, as in the older people or those patients on medications that affect cognition. Researchers should be mindful that hearing loss is not a visible deficit and may have stigma attached that could inhibit a subject's disclosure. Testing all subjects is recommended as a part of selection and inclusion procedures.
  4. Method selection in terms of active versus passive intervention approach must be based on a clear conceptual framework, consistent with what is known about the condition and the population within the intervention context.
  5. Choice of music and decision on operationalization can now be enhanced by current technology, and researchers should consider self-selection as a viable option. Technology using Internet music retrieval sources as well as portable music players is relatively inexpensive, simple, and familiar. Researchers should be clear if they are testing a specific piece of music as the object of study (unwise given this review) or consider the study of self-selected music deemed by the subjects appropriate for research intent.
  6. Researchers should be mindful of the historically low completion rates in some music studies.
  7. Interdisciplinary work with experts who understand music from multiple perspectives, even that external to healthcare research, is encouraged, for example, with experts in academics who study the theoretical foundations of music.
  8. Consistent use of CONSORT guidelines should help future research report to enhance the clarity of study results and support efforts at study replication.


Andrada, L. C., Vidal, A., Aguilar-Tablada, C., Reina, G., Silva, L., Guinaldo, R., et al. (2004). Anxiety during the performance of colonoscopies: Modification using music therapy. European Journal of Gastroenterology & Hepatology, 16(12), 1381-1386.
American Music Therapy Association. (2005). What is music therapy. Retrieved October 21, 2009, from
Bittman, B. B., Berk, L. S., Felten, D. L., Westengard, J., Simonton, O. C., Pappas, J., et al (2001). Composite effects of group drumming music therapy on modulation of neuroendocrine-immune parameters in normal subjects. Alternative Therapies in Health and Medicine, 7(1), 38-47.
Browning, C. A. (2001). Music therapy in childbirth: Research in practice. Music Therapy Perspectives, 19(2), 74-81.
Bruer, R. A., Spitznagel, E., & Cloninger, C. R. (2007). The temporal limits of cognitive change from music therapy in elderly persons with dementia or dementia-like cognitive impairment: A randomized controlled trial. Journal of Music Therapy, 44(4), 308-328.
Buchana, J. C. (1988). An exploratory study of preschool children's synchronization of a selected rhythmic activity with music set at their heart rates. Unpublished dissertation, University of South Florida.
Cassileth, B. R., Vickers, A. J., & Magill, L. A. (2003). Music therapy for mood disturbance during hospitalization for autologous stem cell transplantation: A randomized controlled trial. Cancer, 98(12), 2723-2729.
Cepeda, M. S., Carr, D. B., Lau, J., & Alvarez, H. (2006). Music for pain relief. Cochrane Database of Systematic Reviews, 2, CD004843.
Chang, M. Y., Chen, C. H., & Huang, K. F. (2008). Effects of music therapy on psychological health of women during pregnancy. Journal of Clinical Nursing, 17(19), 2580-2587.
Chang, S. C., & Chen, C. H. (2005). Effects of music therapy on women's physiologic measures, anxiety, and satisfaction during cesarean delivery. Research in Nursing and Health, 28(6), 453-461.
Chlan, L. (1998). Effectiveness of a music therapy intervention on relaxation and anxiety for patients receiving ventilatory assistance. Heart & Lung, 27(3), 169-176.
Chlan, L., Evans, D., Greenleaf, M., & Walker, J. (2000). Effects of a single music therapy intervention on anxiety, discomfort, satisfaction, and compliance with screening guidelines in outpatients undergoing flexible sigmoidoscopy. Gastroenterology Nursing, 23(4), 148-156.
Clark, M., Isaacks-Downton, G., Wells, N., Redlin-Frazier, S., Eck, C., Hepworth, J. T., et al (2006). Use of preferred music to reduce emotional distress and symptom activity during radiation therapy. Journal of Music Therapy, 43(3), 247-265.
Dileo, C. (2006). Effects of music and music therapy on medical patients: A meta-analysis of the research and implications for the future. Journal of the Society for Integrative Oncology, 4(2), 67-70.
Dowling, W. J., & Harwood, D. L. (1986). Music cognition. New York: Academic Press.
Gibbons, A. C. (1977). Popular music preferences of elderly people. Journal of Music Therapy, 14(4), 180-189.
Gold, C., Heldal, T. O., Dahle, T., & Wigram, T. (2005). Music therapy for schizophrenia or schizophrenia-like illnesses. Cochrane Database of Systematic Reviews, 2, CD004025.
Guzzetta, C. E. (1989). Effects of relaxation and music therapy on patients in a coronary care unit with presumptive acute myocardial infarction. Heart & Lung, 18(6), 609-616.
Hanser, S. B. (1990). A music therapy strategy for depressed older adults in the community. The Journal of Applied Gerontology, 9(3), 283-298.
    Hanser, S. B., & Thompson, L. W. (1994). Effects of a music therapy strategy on depressed older adults. Journals of Gerontology, 49(6), P265-P269.
    Hilliard, R. E. (2003). The effects of music therapy on the quality and length of life of people diagnosed with terminal cancer. Journal of Music Therapy, 40(2), 113-137.
    Hoffman, J. (1997). Tuning in to the power of music. RN, 60(6), 52-54.
    Holbrook, M. B., & Anand, P. (1990). Effects of tempo and situational arousal on the listener's perceptual and affective response to music. Psychology of Music, 18, 150-162.
    Horne-Thompson, A., & Grocke, D. (2008). The effect of music therapy on anxiety in patients who are terminally ill. Journal of Palliative Medicine, 11(4), 582-590.
    Iwanaga, M. (1995a). Harmonic relationship between preferred tempi and heart rate. Perceptual and Motor Skills, 81(1), 67-71.
    Iwanaga, M. (1995b). Relationship between heart rate and preference for tempo of music. Perceptual and Motor Skills, 81(2), 435-440.
    Kochkin, S. (2005). MarkeTrak VII: Hearing loss population tops 31 million people. Hearing Review, 16-25.
    Lai, H. L. (2004). Music preference and relaxation in Taiwanese elderly people. Geriatric Nursing, 25(5), 286-291.
    Lazaroff, I., & Shimshoni, R. (2000). Effects of medical resonance therapy music on patients with psoriasis and neurodermatitis: A pilot study. Integrative Physiological and Behavioral Science, 35(3), 189-198.
    Leardi, S., Pietroletti, R., Angeloni, G., Necozione, S., Ranalletta, G., & Del Gusto, B. (2007). Randomized clinical trial examining the effect of music therapy in stress response to day surgery. British Journal of Surgery, 94(8), 943-947.
    LeBlanc, A., Colman, J., McCrary, J., Sherrill, C., & Malin, S. (1988). Tempo preferences of different age music listeners. Journal of Research in Music Education, 36(3), 156-168.
    Ledger, A. J., & Baker, F. A. (2007). An investigation of long-term effects of group music therapy on agitation levels of people with Alzheimer's disease. Aging and Mental Health, 11(3), 330-338.
    Mandel, S. E., Hanser, S. B., Secic, M., & Davis, B. A. (2007). Effects of music therapy on health-related outcomes in cardiac rehabilitation: A randomized controlled trial. Journal of Music Therapy, 44(3), 176-197.
    Maratos, A. S., Gold, C., Wang, X., & Crawford, M. J. (2008). Music therapy for depression. Cochrane Database of Systematic Reviews, 1, CD004517.
    McKinney, C. H., Antoni, M. H., Kumar, M., Tims, F. C., & McCabe, P. M. (1997). Effects of guided imagery and music (GIM) therapy on mood and cortisol in healthy adults. Journal of Health Psychology, 16(4), 390-400.
    McRee, L. D., Noble, S., & Pasvogel, A. (2003). Using massage and music therapy to improve postoperative outcomes. AORN Journal, 78(3), 433-442, 445-457.
    Mindlin, G., & Evans, J. R. (2009). Brain music treatment: A brain/music interface In T. H. B. Budzynski, J. R. Evans, & A. Abarbanel (Eds.), Introduction to quantitative EEG and neurofeedback: Advanced theory and applications (2nd ed., pp. 225-238). New York: Academic Press.
    Moher, D., Schulz, K. F., & Altman, D. (2001). The CONSORT statement: Revised recommendations for improving the quality of reports of parallel-group randomized trials. Journal of the American Medical Association, 285(15), 1987-1991.
    Pacchetti, C., Mancini, F., Aglieri, R., Fundaro, C., Martignoni, E., & Nappi, G. (2000). Active music therapy in Parkinson's disease: An integrative method for motor and emotional rehabilitation. Psychosomatic Medicine, 62(3), 386-393.
    Raglio, A., Bellelli, G., Traficante, D., Gianotti, M., Ubezio, M. C., Villani, D., et al (2008). Efficacy of music therapy in the treatment of behavioral and psychiatric symptoms of dementia. Alzheimer Disease and Associated Disorders, 22(2), 158-162.
    Rosenfeld, A. H. (1985). Music, the beautiful disturber. Psychology Today., 19(12), 48-56.
    Schmid, W., & Aldridge, D. (2004). Active music therapy in the treatment of multiple sclerosis patients: A matched control study. Journal of Music Therapy, 41(3), 225-240.
    Schullian, D., & Schoen, M. (1948). Music and medicine. New York: Henry Schuman, Inc.
    Sendelbach, S. E., Halm, M. A., Doran, K. A., Miller, E. H., & Gaillard, P. (2006). Effects of music therapy on physiological and psychological outcomes for patients undergoing cardiac surgery. Journal of Cardiovascular Nursing, 21(3), 194-200.
    Shiraishi, I. M. (1997). A home-based music therapy program for multi-risk mothers. Music Therapy Perspectives, 15(1), 16-23.
    Smith, M., Casey, L., Johnson, D., Gwede, C., & Riggin, O. Z. (2001). Music as a therapeutic intervention for anxiety in patients receiving radiation therapy. Oncology Nursing Forum, 28(5), 855-862.
    Standley, J. M. (1986). Music research in medical/dental treatment: Meta-analysis and clinical applications. Journal of Music Therapy, 23(2), 56-122.
    Standley, J. M. (2002). A meta-analysis of the efficacy of music therapy for premature infants. Journal of Pediatric Nursing, 17(2), 107-113.
    Standley, J. M., & Prickett, C. A. (1994). Research in music therapy: A tradition of excellence. New York Silver Spring.
    Suzuki, M., Kanamori, M., Watanabe, M., Nagasawa, S., Kojima, E., Ooshiro, H., et al (2004). Behavioral and endocrinological evaluation of music therapy for elderly patients with dementia. Nursing and Health Sciences, 6(1), 11-18.
    Svansdottir, H. B., & Snaedal, J. (2006). Music therapy in moderate and severe dementia of Alzheimer's type: A case-control study. International Psychogeriatrics, 18(4), 613-621.
    Talwar, N., Crawford, M. J., Maratos, A., Nur, U., McDermott, O., & Procter, S. (2006). Music therapy for in-patients with schizophrenia: Exploratory randomized controlled trial. The British Journal of Psychiatry, 189, 405-409.
      Tang, H. Y., Harms, V., Speck, S., Vezeau, T. & Jesurum, J. (2009). Effects of audio relaxation program for blood pressure reduction in older adults. European Journal of Cardiovascular Nursing, 8, 329-336.
      Tang, H. Y., Harms, V., & Vezeau, T. (2008). An audio relaxation tool for blood pressure reduction in older adults. Journal of Geriatric Nursing, 29(6), 392-401.
      Ulrich, G., Houtmans, T., & Gold, C. (2007). The additional therapeutic effect of group music therapy for schizophrenic patients: A randomized study. Acta Psychiatrica Scandinavica, 116(5), 362-370.
      Vink, A. C., Birks, J. S., Bruinsma, M. S., & Scholten, R. J. (2004). Music therapy for people with dementia. Cochrane Database of Systematic Reviews, 3, CD003477.
      Walworth, D., Rumana, C. S., Nguyen, J., & Jarred, J. (2008). Effects of live music therapy sessions on quality of life indicators, medications administered and hospital length of stay for patients undergoing elective surgical procedures for brain. Journal of Music Therapy, 45(3), 349-359.
        White, J. M. (1992). Music therapy: An intervention to reduce anxiety in the myocardial infarction patient. Clinical Nurse Specialist, 6(2), 58-63.
        Wong, H. L., Lopez-Nahas, V., & Molassiotis, A. (2001). Effects of music therapy on anxiety in ventilator-dependent patients. Heart & Lung, 30(5), 376-387.

        music therapy; music intervention; clinical randomized controlled studies; review of literature

        Copyright © 2010 by the Taiwan Nurses Association.