Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Effects of music therapy and music-based interventions in the treatment of substance use disorders: A systematic review

  • Louisa Hohmann ,

    Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Visualization, Writing – original draft, Writing – review & editing

    louisa.hohmann@fu-berlin.de

    Affiliations Department for Educational Sciences and Psychology, Freie Universität, Berlin, Germany, Department for Biological and Medical Psychology, University of Bergen, Bergen, Norway

  • Joke Bradt,

    Roles Methodology, Supervision, Validation, Writing – original draft, Writing – review & editing

    Affiliation Department of Creative Arts Therapies, College of Nursing and Health Professions, Drexel University, Philadelphia, United States of America

  • Thomas Stegemann,

    Roles Validation, Writing – original draft, Writing – review & editing

    Affiliation Department of Music Therapy, University of Music and Performing Arts, Vienna, Austria

  • Stefan Koelsch

    Roles Supervision, Validation, Writing – original draft, Writing – review & editing

    Affiliation Department for Biological and Medical Psychology, University of Bergen, Bergen, Norway

Effects of music therapy and music-based interventions in the treatment of substance use disorders: A systematic review

  • Louisa Hohmann, 
  • Joke Bradt, 
  • Thomas Stegemann, 
  • Stefan Koelsch
PLOS
x

Abstract

Music therapy (MT) and music-based interventions (MBIs) are increasingly used for the treatment of substance use disorders (SUD). Previous reviews on the efficacy of MT emphasized the dearth of research evidence for this topic, although various positive effects were identified. Therefore, we conducted a systematic search on published articles examining effects of music, MT and MBIs and found 34 quantitative and six qualitative studies. There was a clear increase in the number of randomized controlled trials (RCTs) during the past few years. We had planned for a meta-analysis, but due to the diversity of the quantitative studies, effect sizes were not computed. Beneficial effects of MT/ MBI on emotional and motivational outcomes, participation, locus of control, and perceived helpfulness were reported, but results were inconsistent across studies. Furthermore, many RCTs focused on effects of single sessions. No published longitudinal trials could be found. The analysis of the qualitative studies revealed four themes: emotional expression, group interaction, development of skills, and improvement of quality of life. Considering these issues for quantitative research, there is a need to examine social and health variables in future studies. In conclusion, due to the heterogeneity of the studies, the efficacy of MT/ MBI in SUD treatment still remains unclear.

Introduction

The misuse of legal and illegal substances is a significant problem in modern societies. For example, in the United States, the estimated 12-months prevalence rates for addictions in 2014 were 3.0% for alcohol and 1.9% for illicit drugs [1]. Use and misuse of alcohol and drugs are associated with a variety of health, social, and economic disadvantages for the users themselves and others (e.g., family, friends, community, environment, and country [2]). Treatment programs for patients with substance use disorders (SUD) include body detoxification, pharmaceutical, psychosocial, and psychotherapeutic treatment, and recovery management [3]. Nevertheless, only a minority of people with SUD, i.e., about 10%, receives such professional help [4]. Moreover, the treatment completion rates are low (i.e., 47% in the USA in 2006 [5]) and the relapse rates are high (40–60% [6]). Thus, there is still need to improve addiction treatment.

Standard psychological treatments mostly consist of verbal therapies such as cognitive behavior therapy, motivational interviewing, and relapse prevention [7]. In addition, complementary and alternative medical therapies are utilized to allow for creative and expressive ways to address issues. Music therapy is one of such non-mainstream therapies [8]. According to the American Music Therapy Association [9], 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”. Therefore, in this review, the term music therapy (MT) is used only for studies where music therapists were involved in the delivery of the intervention; for studies where the intervention was delivered without participation of music therapists, or their participation remains unclear, we will use the term music-based intervention (MBI). Furthermore, we include studies examining the effect of music stimuli presentation without presence of persons therapeutically guiding the interventions, which are referred to as music presentation (MP) studies.

How can MT/ MBI help patients with SUD? Compared to commonly used verbal psychological therapies, MT and MBI provide different opportunities for self-expression, cooperative group activity, imagination, and synchronized sensorimotor experience [10]. In addition to that, there is evidence of beneficial impact of MT/ MBI on mood [11,12], stress [13], self-esteem [14], motivation [15], emotional expression [16], and social cohesion [17]. Furthermore, MT/ MBIs appear to address general challenges of SUD treatment: For instance, in a study with patients with SUD and comorbid severe mental illnesses MT appreciation was associated with benefits in global functioning and motivation [15]. For patients with non-organic mental disorders and low treatment motivation positive effects of an individual three month MT program on negative symptoms, global functioning, clinical global impressions, social avoidance and vitality were reported [18]. Furthermore, for subgroups of addicted patients with special needs (e.g., women and adolescents [8]) MT/ MBI led to improvements in anxiety [19] and internal locus of control [20].

To clarify the clinical efficacy of MT/ MBIs in addiction treatment, a summary of their effects is warranted. Although there are many reports about the effects of MT/ MBI in patients with SUD in single studies, no meta-analyses are yet available on this topic. In 2008, Mays, Clark, and Gordon [21] systematically reviewed the use of MT for patients with SUD and emphasized a lack of evidence. In their review, they included five quantitative studies that greatly varied in terms of treatment settings, frequency, duration, persons guiding the session, and outcome variables. Furthermore, outcomes like drug consumption or long-term abstinence were not assessed in these studies. Therefore, the treatment effects of MT were primarily related to participants’ attitudes and emotions. In line with that, most of the MT studies in SUD treatment met the criteria of lower levels of evidence according to evidence-based practice hierarchies, indicating that high-quality research has not been conducted [22].

In this paper, we aimed to address the research question of whether MT and MBIs are clinically effective for people with substance use disorders (SUD) by reviewing the current state of research regarding this topic. Because little is known about the key outcomes affected by MT/ MBIs in patients with SUD [21], we evaluated the existing evidence to summarize the benefits of music interventions for this population.

Methods

Criteria for considering studies for this review

Types of studies.

We included all types of studies with quantitative or qualitative data assessed in a systematic way, e.g., by at least semi-structured interviews, video-taping, or questionnaires. We decided not to limit our inclusion criteria to randomized controlled trials (RCTs), even though there are many scholars who recommend focusing on this type of study for systematic reviews and meta-analyses [22,23]. We based this decision on the following rationale: (1) Silverman [22] and Mays et al. [21] emphasized the lack of RCTs available for our research question, and this is still valid at present; (2) for rare conditions and difficult clinical investigations (such as music therapy in psychiatry) the inclusion of other study types (such as case studies or case-control studies) is recommended because they may be the only available evidence [24]; (3) Furthermore, qualitative studies are useful to examine perspectives and experiences [22,25].

We also included MP studies examining the effects of music stimuli presentation on people with SUD without the presence of a music therapist or other persons therapeutically guiding the music intervention.

Types of participants.

We considered studies that included patients or clients with SUD, regardless of age, gender or comorbid disorders. Studies examining subgroups like women or adolescents were included as well. If it was unclear whether all participants suffered from SUD (e.g., a study on residents and staff members of a rehabilitation center [26]), those studies were excluded. If separate conclusions about patients with and without SUD were drawn, those studies were included.

Types of interventions.

All studies examining MT, MBI or MP were included. Articles were excluded if combined programs with music and other complementary approaches were used (e.g., combinations of art, video, music, group therapy, and individual counseling [27]) as this would not allow for the identification of separate effects of MT/ MBI/ MP.

Types of outcome measures.

Similar to Mays et al. [21], we included all outcomes. For a listing of the outcomes included in the study, see Table 1.

thumbnail
Table 1. Clusters of outcomes examined in studies about the effects of music therapy and music-based interventions for patients with substance use disorders.

https://doi.org/10.1371/journal.pone.0187363.t001

Search methods for identification of studies

First, we identified articles by conducting a literature search in the electronic databases ISI Web of Knowledge and Scopus on 1st April, 2016. We used the search term “(music therapy AND addiction) OR (music therapy AND substance use disorder) OR (music therapy AND substance abuse) OR (music therapy AND alcohol*) OR (music AND intervention AND addiction)) OR (music AND intervention AND substance use disorder) OR (music AND intervention AND substance abuse) OR (music AND intervention AND alcohol*)”. After deleting duplicate studies, we scanned the abstracts to include only articles published in English, focusing on MT/ MBI or MP and participants with SUD. Additionally, the bibliographies of the remaining records were scanned for further studies. Articles without systematic data assessment were excluded. Remaining sources were further subdivided with respect to the type of music/ intervention that was examined: (1) studies examining effects of the presentation of music stimuli without application of MT/ MBI (MP studies), (2) studies investigating one session of MT/ MBI, and (3) studies examining more than one session of MT/ MBI. With respect to category (1), for example examinations of simple listening to music without the presence of therapists or other persons guiding the session or experiments were included.

Data collection and analysis

General preparing procedure.

A review protocol does not exist. All unique articles (i.e. duplicates removed) were listed in a table. After their abstracts were scanned, we indicated whether or not the studies met the inclusion criteria listed above. Full texts of studies that met the inclusion criteria were analyzed. The study characteristics and results were summarized in separate tables.

Many studies included similar outcomes but used different terminology. Outcomes that were very similar were clustered under one common outcome term. For example, the outcomes depression, depressiogenic thought frequency, and feeling depressed were clustered under the outcome “depression” (See Table 1 for labels and included variables). For all studies, we extracted design aspects as well as statistical data. Based on this data, we examined if meta-analytic calculations would be useful.

We used three different types of data summary: (1) a description of the effects of MT/ MBI for the quantitative studies separated by outcomes, (2) a summary of effects of MT/ MBI/ MP for the quantitative studies separated by study characteristics, (3) a summary of the topics and themes described in the qualitative studies.

We did not conduct a meta-analysis due to the following reasons. First, according to the Cochrane systematic review guidelines [23], combining studies that use different types of control conditions may lead to meaningless results. After separating the studies per type of control condition, there were too few studies per outcome to allow for meta-analysis. Second, predominantly including studies by the same authors in the same meta-analysis would violate the assumption of independence of study reports [28]. As most of the studies with similar comparison designs were conducted by Silverman [2937], there was too much dependency on the hierarchical level. A more detailed description of reasons for not conducting a meta-analysis is provided in the Results section below.

Descriptive summaries.

We aimed to give an overview of the efficacy of MT/ MBI per outcome in consideration of the quality of the studies. To this end, we created a categorization system (see Fig 1) based on an evidence-based practice (EBP) taxonomy by Melnyk and Fineout-Overholt [52] that was developed for the nursing profession. As MT and nursing contexts appear to be similarly diverse, Silverman [22] recommended the use of this taxonomy when examining EBP for MT. This hierarchy contains seven levels of evidence with (I) being the highest rank and (VII) being the lowest rank in research. The articles we collected for our review did not cover the whole range. Therefore, we refer to Melnyk and Fineout-Overholt’s following levels: (II) well-designed RCTs, (III) well-designed controlled trials without randomization, and (VI) single descriptive or qualitative study. Based on these levels, we developed four main categories for our categorization system: (1) studies without reporting all necessary statistical data to compute a meta-analysis (e.g., means, standard deviations, sample sizes), (2) studies without a control group (CG), (3) non-randomized studies with CG, and (4) RCTs. Categories (3) and (4) are further subdivided into (3a)/ (4a) studies that reported no beneficial treatment effects of MT/ MBI and (3b)/ (4b) studies that reported treatment benefits of MT/MBI compared to a CG. For an overview of the categorization procedure see Fig 1. To draw conclusions about MT/ MBI efficacy, RCTs are necessary [25]. Thus, studies fitting in categories (4a) and (4b), which are matching level (II) of Melnyk and Fineout-Overholt’s taxonomy, are categorized as high level evidence of efficacy. Categories (3a) and (3b) match level (III) in the EBP taxonomy, and categories (1) and (2) match level (VI), i.e., lower levels of evidence. Thus, the categories (1), (2) and (3a)/(3b) are referred to as of low level evidence of efficacy. Nevertheless, it is important to note that research designs other than RCTs are useful for research as well [25], so that our taxonomy of low and high level of evidence of efficacy only refers to the assessment of MT/ MBI efficacy.

thumbnail
Fig 1. Categorization procedure for quality of evidence.

CG = control group.

https://doi.org/10.1371/journal.pone.0187363.g001

For the descriptive summaries we used the following rules: We counted how many unique studies examined a certain outcome (cluster). For studies that included multiple measures (e.g. two different scales) per outcome, data from only one measure was included. This was to avoid artificially inflating the weight of single studies. Articles that reported results of two separate studies within a single publication were used more than once (e.g., [20]). If different raters (e.g. client ratings and therapist ratings) were included only client ratings were counted. Finally, for studies with repeated measures, only immediate post-intervention scores were used.

Summary of music and MT/ MBI effects.

We created separate summaries for (1) MP studies, (2) studies that investigated only one session, and (3) studies that examined the effects of more than one session of MT/ MBI. For each of these three categories a separate table including study characteristics and results was created.

Summary of qualitative articles.

Studies were read carefully, and described topics and themes were summarized in a separate table.

Results

Description of the studies

The identification process is displayed in the flow diagram (adapted from the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [53]) in Fig 2. Our database search resulted in 383 records (without duplicates), 50 of which met the inclusion criteria. The other records were excluded because (a) they were not written in English (n = 44), (b) did not include MT/ MBI as single program or MP (n = 250), or (c) did not primarily focus on SUD (n = 39). One full-text could not be obtained [54], therefore it was excluded. Five of the initially included records turned out to be book reviews and conference abstracts, thus they were excluded. Full-texts were obtained for the remaining 44 articles and an additional 16 articles were found in their references lists, resulting in a total of 60 records. Twenty-one of them were descriptive articles without structured qualitative or quantitative data and were excluded.

Thirty-nine records with systematic data collection remained. One article included two quantitative studies [20], and one consisted of both qualitative and quantitative studies [55]. Two articles reported about the same dataset [16,56], so that these results were summarized as one study. Altogether, we identified 34 quantitative studies, which were further subdivided with respect to the type of music/ MT that was examined: (1) six studies examined effects of music stimuli presentation without application of MT/ MBI, (2) thirteen studies investigated only one session of MT/ MBI, and (3) fifteen studies examined more than one session of MT/ MBI. Six records included qualitative data obtained through semi-structured interviews, structured or video-taped observations or questionnaires.

Sample and setting characteristics

The characteristics of the studies are summarized in Table 2 for studies that examined the effects of music stimuli presentation, Tables 3 and 4 for quantitative studies about MT/ MBI, and Table 5 for qualitative studies about MT/ MBI. Sample settings and characteristics are presented separately in the following for (a) MT/ MBI studies with both qualitative and quantitative data, and (b) MP studies.

thumbnail
Table 2. Characteristics and results of studies examining effects of music stimuli presentation on patients with substance use disorders.

https://doi.org/10.1371/journal.pone.0187363.t002

thumbnail
Table 3. Characteristics and results of studies examining effects of single music therapy session on patients with substance use disorders.

https://doi.org/10.1371/journal.pone.0187363.t003

thumbnail
Table 4. Characteristics and results of studies examining effects of more than one music therapy/music-based intervention session on patients with substance use disorders.

https://doi.org/10.1371/journal.pone.0187363.t004

thumbnail
Table 5. Characteristics and themes of qualitative studies about effects of mt/mbi on patients with substance use disorders.

https://doi.org/10.1371/journal.pone.0187363.t005

MT/ MBI studies.

For the majority of the studies, sessions were held in group settings, except a single-case study [40] and one study with individual application of the music-based program [41]. Most of the studies, i.e., three qualitative and 23 quantitative studies, were classified as “MT studies” (according to the music therapy definition provided in Introduction). With respect to MBIs, one study was conducted by vocal performance majors [57], one by different artists of the Council of Music [58], one by a cultural drumming teacher and a substance abuse counselor [48], one by a social worker [44], one by a counselor [41], and in three cases [15,40,55] the therapist’s background remained unclear.

Not considering the case study, sample size ranged from 8 participants [42,47] to 188 participants [49] for the quantitative studies, and from 3 participants [59] to 20 participants [55] for the qualitative studies. One quantitative [50] and one qualitative study [58] did not report sample sizes.

Six studies examined men only [40,45,50,55,59,60] and five women only [19,4143,47].

Regarding the diagnosis, many samples included various drug addictions, i.e., polydrug abuse. Other studies only focused on chemical dependency [20,42,47], alcohol [45,50,60] or inhalant abuse [40].

With respect to the age of the participants, four studies investigated adolescents only with mean ages/ age ranges between 15 and 17 years [20,42] or as a single case study with a 14-year old boy [40]. For the other studies, mean age varied from 34.4 years [16] up to 52.5 years [48]. Eleven studies [16,19,39,43,47,50,51,55,57,58,60] did not report any measure of central tendency regarding age. In 16 cases [15,16,38,39,41,4349,55,59,60] numeric age ranges were reported which varied from 21 years [44] (31–51 years) to 53 years [48] (19–71 years).

Music stimuli presentation studies.

Sample sizes ranged from 19 participants [61] to 59 participants [62].

Two studies examined men only [55,62], and three investigated both men and women. One study did not report any information about gender [63].

Regarding the diagnosis, three studies focused on alcohol addiction [6264], and the others included various drug addictions.

Regarding the age, mean age ranged from 31.1 years [61] to 43.1 years [64]. Two studies did not report any measures of central tendency [55,63] and one reported a median age of 46.4 years [62]. Age ranges (when reported) differed only slightly from 28 years [65] (20–47) to 33 years [64] (27–59).

Results of quantitative MT/ MBI studies separated by outcomes

For an overview of the efficacy of MT/ MBI per outcome (cluster) in consideration of the quality of the studies see Fig 3. Studies were classified according to the categorization scheme presented in Fig 1. None of the studies met the criteria of categories (3a) and (3b), i.e., studies with CG without randomization, so that these categories are not represented in Fig 3. In the following section, we will describe the results in more detail.

thumbnail
Fig 3. Descriptive overview of quantitative studies examining the effects of music therapy and music-based interventions on different outcomes.

Studies with effect or no effect compared to control group (CG) were classified as of high level evidence of efficacy (black and dark grey bars). Studies reporting insufficient statistical data to conduct meta-analyses and without CG were classified as of low level evidence of efficacy (light grey bars).

https://doi.org/10.1371/journal.pone.0187363.g003

Motivation.

For the effect of MT/ MBI on variables related to this cluster (motivation, treatment eagerness, change readiness), 10 results were collected, and eight of them (80%) represent high level evidence of efficacy. For Silverman [32,34] who examined different motivational constructs within the same samples only motivation scores were used. All studies except one reported all statistical data and only one included pilot data without a CG [15]. In 37.5% of studies of high level evidence of efficacy (3/8), i.e. 30% of all studies (3/10), beneficial effects of MT/ MBI were found. All RCTs except one [38] were conducted by Silverman [29,3136], and they differed widely with respect to CG designs and scales, so a meta-analysis was not conducted.

Depression.

This outcome was examined in 11 studies including Howard [42] which reported on two separate samples. Five studies report insufficient statistical data and four were non-controlled studies, so that only 27% of the results (3/11) were categorized as of high level evidence of efficacy. None of the RCTs found benefits of MT/ MBI compared to CG. A meta-analysis was not conducted due to the different CG designs.

Enjoyment.

All three data sources of high level evidence of efficacy (i.e., 50% of all studies regarding this outcome) were reported by Silverman [29,31,32], so that we did not conduct a meta-analysis. Three further studies of low level evidence of efficacy [16,46,47] did not report all necessary statistical data. One out of three RCTs, i.e., 17% of all results (1/6), reported a positive effect of MT on enjoyment.

Withdrawal and craving.

We decided to cluster these outcomes as the variables are closely linked. Patients in states of withdrawal often experience craving, and consumption of the substance may immediately and effectively reduce the symptoms [37]. Silverman [30,32,37] conducted three different RCTs examining craving and/or withdrawal in patients addicted to various drugs (e.g., alcohol, heroine, prescription drugs and cocaine). None of the studies showed beneficial effects of MT compared to different CG. A meta-analysis was not conducted because all results were reported by the same author.

MT helpfulness.

Forty percent of the results (2/5) were of high level evidence of efficacy, comparing MT to group verbal therapy, and both RCTs were conducted by Silverman [31,32], so a meta-analysis was not appropriate. The lack of statistical details prohibited inclusion in meta-analysis for two further studies [44,47], and another study was a non-controlled study [15]. All in all, 50% of studies of high level evidence of efficacy, i.e., 20% of all studies (1/5) regarding helpfulness were in favor of MT/ MBI.

Locus of control (LOC).

All three studies considering LOC were RCTs, and two of them [20] (i.e., 67%) found positive effects of MBI/ MT. A meta-analysis was not conducted due to different CG designs.

Participation.

For this category, many different constructs regarding the quantitative assessment of patients’ engagement and participation were subsumed, so six data sources were identified: Only one study included a CG [29] and did not identify benefits of MT. Furthermore, 50% of all data sources (3/6) did not report all statistical data [16,49,50] and 33% (2/6) were non-controlled studies [15,48].

Coping skills.

Only one study out of three studies (33%) for this outcome, was of low level evidence of efficacy, i.e., a case study not including a CG [40]. None of the RCTs [36,38] found benefits of MT for coping skills.

Constructs examined without studies of high level evidence of efficacy.

For five outcome clusters, namely anxiety, medical symptoms, anger, sadness, and stress, no RCTs could be identified, so conclusions about efficacy cannot be drawn.

Follow-up investigations.

Only one RCT assessed follow-up scores regarding depression, enjoyment, perceived effectiveness and being clean [31] and did not identify differences between group verbal therapy and MT groups one month after intervention completion.

Conclusion.

For at least eight categories of outcomes, studies of high level evidence of efficacy, i.e., RCTs, were identified. The descriptive summaries suggest that there is evidence for benefits of MT/ MBI compared to different control groups (CGs), especially for the variable locus of control (67% positive effects compared to CG). Additionally, regarding perceived helpfulness of the intervention, half of the RCTs reported higher values for MT compared to CG. For motivation and enjoyment there were inconsistent results, and more than half of the studies of high level evidence of efficacy did not identify statistically significant improvement for MT/ MBI participants. Regarding depression, withdrawal/ craving, participation, and coping skills none of the RCTs reported benefits for MT. Studies examining anxiety, medical symptoms, anger, sadness, and stress were all of low level evidence of efficacy, so that their results can only serve as a base for further research giving hints to constructs that should be evaluated with RCTs.

Results of quantitative studies separated by study characteristics

We now describe effects of MP, MT and MBI considering study characteristics according to the following categories: (1) effects of music in general, (2) effects of one session of MT/ MBI, and (3) effects of more than one session of MT/ MBI. Because most of the studies were not RCTs, conclusions about MT efficacy cannot be drawn. Thus, the summaries include descriptions of clinical effectiveness, i.e. the effects in clinical practice [67].

Studies comparing MT methods (e.g., lyric analysis and songwriting [46]) did not identify significant differences between the interventions, so that the methods are not differentiated in the following. With respect to the nomenclature, we noticed that regarding mood there is still no consensus, as mood, feelings, and emotions are often used interchangeably. For instance, Jones [46] refers to the terms “feelings and emotions” (p. 100), only to eventually assess “mood” using a visual analogue mood scale. Thus, due to the heterogeneity of the nomenclature used in the studies, it was not possible to differentiate these terms properly.

Effects of music presentation (MP).

Six studies examined the impact of music on patients with SUD without therapeutical involvement of an interventionist (see Table 2). The following effects of listening to music were reported: Short and Dingle [61] examined the impact of sad, happy, and relaxing songs on arousal in patients with SUD and a healthy control group (CG). Whereas the participants of the CG indicated different degrees of arousal and pleasantness for the three tracks, the SUD patients rated the three pieces of classical music equally arousing and pleasant. Furthermore, their degree of craving was linked to the personal relevance of songs: The patients reported increased craving after listening to a track associated with their substance use, whereas afterwards, listening to a track associated with abstinence resulted in decreased craving. These results indicate less emotional variations in SUD patients and a direct impact of music on relapse related variables. Furthermore, Fritz et al. [65] reported a strong context dependency of music effects. They conducted a musical feedback intervention with listening to a jointly self-produced music piece or a commercial track. Self-produced music showed positive effects on mood and locus of control (LOC) only when it was presented prior to the commercial music production. Jansma et al. [64] examined the effect of mood states on alcohol cue reactivity. They induced depressive mood by presentation of depressive music or distressed mood by negative feedback following a high performance task. Alcohol cue reactivity was present, but did not differ between negative or neutral mood states. Nevertheless, the patients were less irritated and more satisfied after depressive mood induction compared to distressing mood induction.

With respect to more abstract outcomes, patients with alcohol dependency perceived colors with greater intensity after listening to music compared to people (patients and staff of a therapeutic community) without exposure to music [63]. Similarly, music during LSD therapy was associated with colors, geometric designs, and past events. Between groups with and without music exposure, there was no difference in LSD experience. Nevertheless, only listening to familiar music appeared to have an effect on general music preference [62].

Additionally, there was experimental evidence for positive effects of music listening over a longer period of time. For members of a therapeutic community for drug users, music listening before falling asleep was related to increased sleep quality and mood on the following day as well as to decreased drop-out rate during a one-month-intervention [55].

Effects of one MT/ MBI session.

MT/ MBIs typically include more methods than simply listening to music [68]. Effects of quantitative studies examining single MT sessions (mostly lyric analysis, songwriting or improvisation) are summarized in Table 3. Most of them were conducted in detoxification centers with a short duration of stay between three and five days. Compared to a verbal therapy CG, MT participants showed similar measures of change readiness, depression, sobriety [31], client-rated working alliance [29], LOC [30], treatment eagerness, drug avoidance self-efficacy [34], craving [32], and withdrawal symptoms [30]. Silverman compared MT groups to wait-list CGs with pretest only, and found no differences regarding craving and withdrawal [37]. Positive effects of MT vs. group verbal therapy were found for therapist-rated working alliance [29], comfort [31], and motivational variables: MT participants had higher realization that aspects of change can be better than the status quo and more active changes [32]. In line with that, MT groups showed increased problem recognition, desire for help, treatment readiness, and total motivation compared to a wait-list CG with pretest only [33,35]. Furthermore, Silverman [36] found higher motivation to reach and maintain sobriety for participants of educational MT compared to patients receiving education without music or a music game. In the same study, treatment eagerness and knowledge of coping skills or triggers did not differ between groups. In three other studies, similar motivation scores between MT groups and verbal therapy or pretest CG were identified [29,32,34], indicating that the effects of single MT sessions on motivational aspects are not coherent. Regarding perceived enjoyment and helpfulness, the results were not consistent as well [29,31,32].

Other studies with single sessions for data analysis were conducted in an inpatient non-medical detoxification unit [46], an in- and outpatient rehabilitation unit [16], an inpatient dual diagnosis treatment unit [51] and an inpatient gender-specific residential program [19]. All these studies reported beneficial effects on mood: For instance, 65% of the participants showed a positive mood change [16]. More specifically, a great amount of the participants reported decreased anxiety [19], anger, and sadness [51], and or an increase in acceptance, enjoyment, happiness, and joy [46]. Furthermore, 87.5% of the participants used MT for mood regulation [16]. Nevertheless, one study found no differences between pre- and posttest regarding anxiety and depression [46].

Effects of multiple MT/ MBI sessions on mood.

Effects of studies examining more than one session are summarized in Table 4. Awareness, expression, and change of emotions are often mentioned as important intended therapy goals [50]. Therefore, five studies in inpatient settings [38,39,42,45,47] and five studies in outpatient settings [40,41,43,44,49] examined treatment effects on mood and emotions. Generally, MT participation was associated with positive mood changes [49], and the scores for perceived enjoyment and effectiveness of MT were almost at the maximum [47]. With respect to negative emotions, MT was linked to reduced anger, depression, stress, and anxiety [40,41,4345]. Two RCTs identified beneficial effects of MT regarding therapist-reported, but not self-reported depression scores [38,39].

Effects of multiple MT/ MBI sessions on other outcomes.

MT and MBI also affected other psychological variables: Adolescents with chemical dependency completing MT showed increased internal LOC compared to a wait-list CG engaging in alternative activities [20]. Results regarding motivation and coping skills were not clear: While in one RCT similar levels for both variables after standard treatment (CG) and additional Guided Imagery and Music (GIM) therapy were reported [38], a single case study found improved coping skills and motivation [40]. This patient had also reduced psychiatric symptoms after the MT intervention. In line with this finding, a cultural-based drumming treatment was associated with improved psychiatric and medical status in Native Americans [48]. In a non-randomized pilot study conducted in an inpatient treatment for dually diagnosed people with SUD and mental illness, Ross et al. [15] examined relationships between MT variables, psychiatric symptoms, general functioning, aftercare appointment, and motivation measured by the Stages of Change, Readiness and Treatment Eagerness Scale (SOCRATES). They found positive associations between MT appreciation and global functioning during hospital stay. Therapist appreciation was positively related to changes in global functioning and the Taking Steps subscale of SOCRATES measuring active changes. Furthermore, cross-sectional analyses at discharge revealed associations between MT appreciation and Taking Steps as well as between therapist appreciation and the Ambivalence and Taking Steps subscales of SOCRATES. Although MT variables did not directly predict improvement in psychiatric symptoms, the number of attended sessions was positively related to aftercare appointment in a following outpatient program within one week after hospital discharge. With regard to long-term effects beyond the hospital treatment, MT was also associated with sobriety and reduced substance use in another study [48]. As this pilot sample consisted of a small number of Native Americans not involved in inpatient settings, it remains unclear whether the results are transferrable to other populations. Nevertheless, MT was associated with beneficial behavioral aspects like high involvement, attendance and on-task behavior in several studies [42,49,50]. These findings suggest that MT and MBI may be important tools for recovery in line with the participants’ subjective evaluations of treatment effects and perceived helpfulness.

Qualitative studies

Six qualitative studies examined and described the participants’ reactions, attitudes, and subjective associations in the context of MT and MBI. In four studies, the patients’ behavior during the session was recorded using video-tapes [55,59] therapist’s notes [60,66] and lyric analysis [66]. Some authors conducted semi-structured interviews [57,59] or used questionnaires that were analyzed qualitatively [58]. Four general themes were identified: Firstly, music served as a tool for expression of thoughts and feelings. Secondly, in all qualitative studies the role of music and MT/ MBI for group interaction, cohesion, and relationships to others, including the therapist [60] was emphasized. Thirdly, MT/ MBI were related to the learning of skills regarding music [57], problem solving [66], and social interaction [59]. Finally, MT/ MBIs were associated with benefits for health and quality of life [59]. In line with the quantitative data, the behavioral observations revealed high engagement and involvement of the participants [55,59,66].

Discussion

In order to address the research question whether music therapy (MT) and music-based interventions (MBIs) are clinically efficient for people with substance use disorders (SUD), we obtained a systematic collection of articles resulting in 34 quantitative and six qualitative studies. Regarding MT/MBI efficacy, we used a descriptive approach to summarize the efficacy evidence of quantitative studies. Furthermore, we summarized effects of exposure to music stimuli, MT and MBIs to describe findings regarding effectiveness. In the following, we discuss these effects, focusing on motivation and on findings regarding the four main themes identified in qualitative analyses. Furthermore, we discuss the quality of the studies. Taken together, the studies do not show clear common effects. Additionally, only few studies have assessed outcomes related to substance use even though such outcomes are critical for treatment success. Thus, variables such as long-term sobriety need to be examined in future studies. Possible mechanisms that may contribute to positive effects of MT/MBI remain to be investigated and specified as well.

Effects of music stimuli presentation

There is evidence for the direct impact of listening to music on emotions and craving without application of MT/ MBI [61]. In addition, frequent listening to relaxing tracks had a beneficial effect on sleep, mood, and treatment completion [55]. Neuro-imaging studies have demonstrated that music listening engages many brain structures important for cognitive, emotional, and sensorimotor processing [69], in particular the mesocorticolimbic system [70,71]. Positive short-term effects on variables like craving may reflect benefits for mental health even on a neurobiological level [72].

Short-term effects of single MT/ MBI sessions

Apart from the general impact of music stimuli presentation, participation in single MT sessions may result in additional short-term effects. Those are important to examine because many patients with SUD attend detoxification treatments with a low frequency of therapy sessions [3]. Single MT sessions appear to be as effective as single verbal therapy sessions for various psychological outcomes (e.g., withdrawal, LOC, craving, client-rated working alliance, and depression), and there were higher scores for MT for comfort [31], therapist-rated working alliance [29], and some aspects of change readiness [32]. These findings support the use of MT in short-term treatments for SUD. Results regarding enjoyment, helpfulness, and motivation differed between studies [29,31,32], although these aspects may be especially important in short-time interventions. As they may be related to positive therapeutic experiences, these factors may facilitate the participation in additional interventions. Importantly, the only RCT with follow-up assessment did not find any beneficial effects of single MT sessions on depression, enjoyment, perceived effectiveness and sobriety [31] after a one-month period. Additional longitudinal analyses of single session effects are necessary.

Effects of MT/ MBI on motivation

Lack of motivation is a crucial problem in the treatment of SUD [73], and beneficial effects of MT and MBI on motivation were commonly described [74,75]. Music itself is motivating and empowering for many people and it has been suggested that engagement in music making may lead to enhanced internal change motivation [76]. High rates for on-task behavior and engagement reported in qualitative and quantitative studies included in this review support this assumption [55]. Ten studies quantitatively assessed motivation, and eight of them were RCTs. Most of them investigated single sessions [2936], and two included longer interventions [15,38]. Despite the positive qualitative reports of patients, not all of these studies identified significant benefits for MT/ MBI. Silverman reported higher treatment and sobriety motivation after MT compared to a wait-list CG with pretest only [33,35], whereas others identified no differences compared to verbal therapy or pretest [32,34]. Different results may be due to different study designs, comparisons or measurement instruments. For instance, a Likert scale for the assessment of motivation revealed similar ratings across groups, whereas the use of a multidimensional scale resulted in higher scores for experimental group than CG in the same sample [32]. In line with that, most studies with Likert scales did not identify benefits for MT groups [32,34,36], whereas the use of some multidimensional instruments revealed significant treatment effects [32,33,35]. There is actually no consistent definition for motivation in the context of research on addiction [77]. Therefore, it is difficult to find an adequate outcome measure capturing all relevant aspects and fitting to the treatment setting. For instance, Silverman [33] examined treatment motivation and readiness with the Circumstances, Motivation, and Readiness Scales for Substance Abuse Treatment [78] and did not identify benefits for MT. The use of this instrument as a clinical assessment tool is not recommended [77] because it was originally developed in the context of a therapeutic community. It is, at this point, not possible to claim that issues with instrument selection are related to incongruence of findings; however, this is certainly an issue worthy of further investigation in future studies.

Prochaska and DiClemente [79] argued that behavior change always occurs as process with different stages of change, so that differentiating aspects of motivation regarding these stages might be useful. Considering this, beneficial effects of MT on problem recognition, desire for help, treatment readiness, and overall motivation were reported [35]. Furthermore, there might be a benefit of therapeutic use of music compared to solely music engagement without therapeutic context as MT participants showed higher motivation scores than patients playing a music game instead [36]. Nevertheless, there were no differences for treatment eagerness in the same study, suggesting that there is need to differentiate between the motivational variables. More RCTs that use the same outcome measures and use the same control group interventions are needed to draw further conclusions.

Examining more than one session of MT, K. M. Murphy [38] did not identify benefits in motivation for patients with an additional GIM intervention compared to those with standard program only. Because this study did not include a sufficient amount of participants (N = 16), long-term effects on motivation should be systematically examined in larger samples in more detail.

Effects of MT/ MBI on mood and emotions

In many studies, MT/ MBI had beneficial effects on mood and emotions, i.e., positive mood change, decreased negative emotions, e.g., anxiety, depression, and anger, and increased positive feelings, e.g., enjoyment and happiness. This is in line with the importance of MT for the expression and regulation of feelings, as identified in our qualitative analyses. MT provides opportunities for the exploration and expression of feelings without drugs and appears to be a non-threatening intervention [80]. Therapist-selected songs as well as songs written or selected by the participants themselves contain aspects related to feelings [33]. Many music therapy studies have demonstrated that songs may be used as a verbal and nonverbal tool for the exploration of feelings [12,16,46,81,82]. Jones reported that over the course of lyric analysis and songwriting interventions, emotional expression appears to increase, and suggests that positive mood changes may have a positive influence on further treatment-related variables such as therapeutic alliance [46]. Additionally, support by other group members may facilitate emotional expression [50]. Nevertheless, it should be noted that for many emotional variables (e.g., anxiety, anger, sadness) RCTs are needed to assess MT/ MBI efficacy.

Effects of MT/ MBI on skills and locus of control

Qualitative analyses suggested that MT/ MBI provide opportunities to learn in various areas. Many patients with SUD have poor psychosocial skills, which improved over the course of MT/ MBI [46,59]. M. Murphy [83] has suggested that music, as part of the participants’ everyday life, is adaptable to low levels of psychosocial functioning, and group interventions may be helpful in reducing social isolation. According to Ghetti [76], in group music therapy sessions, the therapist structures the active music making purposefully to enable group interaction in a non-threatening atmosphere. Successful group interactions in music making may help to develop social and problem solving skills. Furthermore, discussion of lyrics of popular songs can help enhance understanding of the individual’s dynamics regarding substance use and may lead to the development of more healthy coping strategies [76]. Only very few studies examined effects of MT/ MBI on cognitive abilities quantitatively. In contrast to the findings reported in qualitative studies, in RCTs no positive effects of MT on coping abilities could be identified [36,38], and also no effects of MBI on cognitive functioning were reported [48]. In contrast to that, a single case study showed enhanced coping after individual sessions [40]. However, these studies differed with respect to many variables, e.g., age, drugs, MT/ MBI methods, and duration so conclusions regarding treatment effect cannot be drawn at this time.

As internal change motivation is a critical aspect for the treatment of addictions, effects of MT/ MBI on locus of control (LOC) were examined as well. After a single session, MT participants did not differ regarding LOC compared to a verbal therapy group [30]. Furthermore, in an experimental setting examining effects of music stimuli presentation, increased internal LOC depended on the context [65], but after longer MT interventions, enhanced internal LOC was identified [20]. These results suggest that MT/ MBI may lead to increased internal LOC over time. When patients experience that their own abilities and actions determine what happens [84] during MT/ MBIs, this may be transferred to life outside the therapy setting and result in better outcomes of addictions’ therapy in the long term [85]. Typically in MT, music experiences are carefully structured and supported by the music therapist to enhance the potential for positive experiences by the patient [86]. This may lead to positive effects of MT on factors such as self-esteem [87] or self-efficacy [88]. However, it is important to acknowledge that asking patients to engage in music making may lead to some anxiety and insecurity as well for some patients, as has been reported in studies outside of the SUD population [89]. However, no studies to date have directly examined the relationship between mastery in music therapy and long-term treatment outcomes for patients with SUD. More research is necessary to explore this possible mechanism.

MT/ MBI effects on group interaction and relationships

Positive group dynamics were identified as important motivators in all qualitative studies. Over the course of the intervention, behavioral observations revealed increased exchange and cohesion [57,59,66]. Nevertheless, in their study with offenders in a substance abuse/mental illness treatment program, Gallagher and Steele [49] reported that 53% of their participants were “not sociable” (p. 121). For planning of the sessions, clinicians need to keep in mind that many patients with SUD have poor social skills. However, none of the quantitative studies in our review systematically examined group-related variables, so future research should examine social skills or aspects like group cohesion. Summarizing studies with respect to the outcome cluster, participation reveals a lack of studies of high level evidence of efficacy regarding this topic as well.

Regarding working alliance between therapist and patients, beneficial effects from the therapist’s perspective were identified quantitatively [29] as well as qualitatively [60]. By contrast, patients attending MT did not perceive a better working alliance compared to a verbal therapy CG [29]. This is in line with previous studies identifying weak reliability between therapist-rated and patient-rated working alliance in drug treatment [90]. Regarding the relationship between different perspectives of working alliance and therapeutic success, results are not consistent: Some studies found stronger relationships between the counsellor’s/ therapist’s view and success [9094], whereas in other studies the patient’s view was identified as a more important predictor [95] or both measures were only weakly correlated with success [96]. Furthermore, levels of working alliance had different effects on outcome for different types of therapies [97]. These inconsistent results indicate that working alliance may be more complex and depend on many aspects. As most of the studies emphasized the importance of the therapist’s view, especially ratings at early time points after starting the therapy [98] as examined by Silverman [29], working alliance should be examined in further MT studies.

MT/ MBI effects on quality of life and health

In many studies, MT and MBI were associated with a great amount of perceived enjoyment and also reported to enhance quality of life and improve health [59]. In line with this, longer MBI were related to positive psychiatric and medical outcomes [40,48]. Nevertheless, these investigations were conducted in very specific settings, so that there is still a lack of studies examining health-related and long-term variables in common treatments for SUD. Especially, variables related to substance use are understudied. Furthermore, all studies examining medical symptoms were categorized as of low level evidence of efficacy in our descriptive summaries. Thus, high quality evidence has not been conducted.

Study quality and methodological recommendations

Our descriptive summaries considered the quality of the identified studies and revealed that in the last years, since the review of Mays et al. in 2008, more RCTs were conducted. Thus, for outcomes like motivation, depression, enjoyment, withdrawal and craving, perceived helpfulness, working alliance, and locus of control studies of high level evidence of efficacy already exist. Nevertheless, we did not calculate meta-analyses due to study heterogeneity or because similar variables were only examined by the same author. Furthermore, across all studies included in our descriptive approach, still only 38% (25/65) were RCTs, and especially for mood variables and long-term abstinence, high quality research has not been conducted. Due to the low quality of most of the studies, in the end, strong key outcomes cannot be substantiated.

It is important to consider that in studies that examine the impact of group interventions, the independence of observations, a common assumption for standard statistical tests, may have been violated because of interactions between group members. This may have resulted in biased standard errors and erroneous inference [99].

In Table 6, methodological recommendations are summarized that are aimed at helping to overcome issues in future research. Most importantly, studies should investigate long-term outcomes such as abstinence and use randomized controlled trial designs. In order to reduce problems related to the independence of observations, hierarchical analyses taking into account the group structure of the data or cluster randomization should be applied. However, designing and executing of cluster randomized trials is difficult because for example larger sample sizes are needed or recruitment bias could occur [100].

If in the clinical context randomization is not possible, studies should at least include control groups as reference frameworks. In within subjects designs aimed at examining pre to post MT/MBI intervention improvements in functioning, one needs to consider that the statistical regression to the mean may be an explanation for the patients’ improvement. Including a control group may solve this issue. Studies of low level evidence of efficacy can be useful for generating hypotheses, getting information about subjective experiences, exploring effects on individual levels, or assessing the ecological validity of treatments [25]. Thus, we also included them in our review, but in 50% of these non-RCT studies (20/40), the results were reported without sufficient statistical information. Furthermore, across all studies, reports about characteristics of intervention, studies and participants varied widely, so that giving a transparent overview and comparing the studies regarding these aspects was difficult. In addition to that, only few studies reported standardized effect sizes [3137,39,41], so the effects of MT/ MBI could hardly be interpreted and compared across studies. Therefore, we recommend the inclusion of reports that clearly describe characteristics of intervention, studies and participants, including diagnostic criteria, transparent information about statistical procedures, and all necessary statistical data (including effect sizes) according to the guidelines of the Task Force on Statistical Inference [101] in the articles. In addition to that, as described in the paragraphs above, high-quality research for outcomes related to skills, group interaction and relationships has not been conducted although these aspects are important topics mentioned in qualitative research. Thus, future studies should investigate variables such as cognitive abilities, group cohesion or medical symptoms among others. Measurement instruments for the same outcomes widely varied across studies (e.g., Likert scales vs. standardized tests) and they mostly captured different aspects, so comparisons were difficult. Therefore, in future research authors should use the same standardized measurement instruments that are suitable for the addiction and music therapy context. Furthermore, the researcher often acted also as music therapist and collected the data which may lead to procedural bias (such as Rosenthal effect [102]) or response bias in the data. It also remains unclear whether effects are due to the music therapy or the person of the music therapist. To reduce these tendencies, we recommend the inclusion of external researchers for data collection and analysis.

Conclusions

There is still no consensus regarding the effects of music therapy (MT) and music-based interventions (MBI) for patients with substance use disorders (SUD). Previous reviews [21,22] highlighted the need for more randomized controlled trials (RCTs) regarding long-term outcomes like maintenance of sobriety. The current literature includes additional RCTs, but most of them focused on short-term effects after single sessions in detoxification units. One RCT examined sobriety after a one-month period without significant differences between a single session of MT or group verbal therapy [31]. The only study examining abstinence after more than one session was conducted with one specific ethnic group without inpatient participants [48]. Therefore, future studies should include long-term investigations and follow-up measurements, in particular regarding variables related to substance use. Due to the great fluctuation in SUD treatments, planning of these studies may be a challenge. However, reduction of substance use and abstinence are critical aspects regarding the success of addictions’ treatment, so evaluations of treatment effects for these outcomes are necessary for future investigations. MT/ MBI appeared to be effective in the regulation of emotions and subjective outcomes, as also indicated by qualitative analyses. Nevertheless, the quantitative studies in our review were very diverse which was one important reason for not conducting meta-analyses. As MT/ MBI are commonly and specifically used in the treatment of groups and subgroups with SUD, e.g., women or adolescents [8], it is important to examine its efficacy and effectiveness in these specific populations as well. However, these results may not be generalizable across general SUD settings. Additionally, it is important to be aware that music can also trigger relapse (e.g. if the music is associated with substance abuse [61]), and that, therefore, music has to be used with great care in SUD patients.

Regarding limitations of the current review it must be noted that collecting the characteristics of the studies was particularly difficult because of missing information. We did not consider the patients’ additional diagnoses and treatment options, methods or specific therapy goals. These topics could be included in future reviews to provide additional insights in characteristics of effective MT/ MBI/ MP. Due to the small number of MT studies, separations regarding these aspects are currently not useful. Whereas this systematic review summarizes the available evidence in terms of treatment efficacy, it does not provide information about potential mechanisms of action of MT/ MBI for SUD. Furthermore, a methodological review of MT/ MBI/ MP and SUD studies may be warranted in the future. For example, studies could be codified regarding methodological strengths and weaknesses to make further methodological recommendations with respect to the investigation of concrete outcomes.

From a methodological point of view, future studies examining the efficacy of MT/ MBI/ MP for patients with SUD should include RCTs, so that meta-analytic calculations will be possible. Regarding content and outcome variables, future studies should consider including outcomes related to the qualitative findings as well as variables related to substance abuse so that a comprehensive picture of the efficacy of MT/ MBI/ MP can be drawn. In addition, we urgently need mechanistic studies that identify and examine the impact of potential treatment mediators and moderators. Additionally, the effects on problem solving, cognitive, and coping abilities and the role of MT/ MBI/ MP for different stages of motivation should be clarified. Furthermore, effects of the interventions on long-term medical and psychiatric outcomes, treatment retention and completion should be examined, while considering additional moderating and mediating variables like MT appreciation. Based on these findings, implications for future MT/ MBI as independent or adjunctive treatment programs for SUD can be developed. As individual preferences regarding music and MT as well as group dynamics appeared to be important for the success of MT [21], careful group composition and selection of materials are necessary. All in all, due to its high acceptance, flexibility, easy accessibility and low costs, MT/ MBI provide opportunities for SUD treatment for various groups in various settings. Nevertheless, its efficacy and effectiveness have to be evaluated more systematically and should focus on further long-term outcomes.

Supporting information

S1 Table. Prisma 2009 checklist for the review about music therapy and music-based interventions in the treatment of substance use disorders.

https://doi.org/10.1371/journal.pone.0187363.s001

(PDF)

References

  1. 1. Center for Behavioral Health Statistics and Quality. Results from the 2014 National Survey on Drug Use and Health: Detailed tables [Internet]. 2015. Available: http://www.samhsa.gov/data/sites/default/files/NSDUH-DetTabs2014/NSDUH-DetTabs2014.htm#tab5-14b
  2. 2. Nutt DJ, King LA, Phillips LD. Drug harms in the UK: A multicriteria decision analysis. The Lancet. 2010;376: 1558–1565.
  3. 3. United Nations Office on Drugs and Crime. International standards for the treatment of drug use disorders [Internet]. 2016. Available: http://projectoptions.org/wp-content/uploads/2016/03/IntStandards.for_.Treatment.DrugDisorders.2.pdf
  4. 4. Substance Abuse and Mental Health Services Administration. Results from the 2013 National Survey on Drug Use and Health: Summary of national findings [Internet]. Rockville, MD: Substance Abuse and Mental Health Services Administration; 2014. Report No.: NSDUH Series H-48, HHS Publication No. (SMA) 14–4863. Available: http://www.samhsa.gov/data/sites/default/files/NSDUHresultsPDFWHTML2013/Web/NSDUHresults2013.pdf
  5. 5. Substance Abuse and Mental Health Services Administration. Treatment Episode Data Set (TEDS): 2006 discharges from substance abuse treatment services [Internet]. Rockville, MD: Substance Abuse and Mental Health Services Administration; 2009. Report No.: DASIS Series: S-46, DHHS Publication No. (SMA) 09–4378. Available: http://wwwdasis.samhsa.gov/teds06/TEDSD2k6_508.pdf
  6. 6. McLellan AT, Lewis DC, O’Brien CP, Kleber HD. Drug dependence, a chronic medical illness: Implications for treatment, insurance, and outcomes evaluation. JAMA. 2000;284: 1689–1695. pmid:11015800
  7. 7. Jhanjee S. Evidence based psychosocial interventions in substance use. Indian J Psychol Med. 2014;36: 112–118. pmid:24860208
  8. 8. Aletraris L, Paino M, Edmond MB, Roman PM, Bride BE. The use of art and music therapy in substance abuse treatment programs. J Addict Nurs. 2014;25: 190–196. pmid:25514689
  9. 9. American Music Therapy Association. Music therapy and mental health [Internet]. 2006. Available: http://www.musictherapy.org/assets/1/7/MT_Mental_Health_2006.pdf
  10. 10. Malchiodi CA. Expressive therapies: History, theory and practice. In: Malchiodi CA, editor. Expressive therapies. New York: Guilford Press; 2005. pp. 1–15.
  11. 11. Maratos A, Gold C, Wang X, Crawford M. Music therapy for depression [Internet]. The Cochrane Collaboration, editor. Cochrane Database of Systematic Reviews. Chichester, UK: John Wiley & Sons, Ltd; 2008. https://doi.org/10.1002/14651858.CD004517.pub2 pmid:18254052
  12. 12. Shuman J, Kennedy H, DeWitt P, Edelblute A, Wamboldt MZ. Group music therapy impacts mood states of adolescents in a psychiatric hospital setting. Arts Psychother. 2016;49: 50–56.
  13. 13. Pelletier CL. The effect of music on decreasing arousal due to stress: A meta-analysis. J Music Ther. 2004;41: 192–214. pmid:15327345
  14. 14. Sharma M, Jagdev T. Use of music therapy for enhancing self-esteem among academically stressed adolescents. Pak J Psychol Res. 2012;27: 53–64.
  15. 15. Ross S, Cidambi I, Dermatis H, Weinstein J, Ziedonis D, Roth S, et al. Music therapy: A novel motivational approach for dually diagnosed patients. J Addict Dis. 2008;27: 41–53. pmid:18551887
  16. 16. Baker FA, Gleadhill LM, Dingle GA. Music therapy and emotional exploration: Exposing substance abuse clients to the experiences of non-drug-induced emotions. Arts Psychother. 2007;34: 321–330.
  17. 17. Gooding LF. The effect of a music therapy social skills training program on improving social competence in children and adolescents with social skills deficits. J Music Ther. 2011;48: 440–462. pmid:22506299
  18. 18. Gold C, Mössler K, Grocke D, Heldal TO, Tjemsland L, Aarre T, et al. Individual music therapy for mental health care clients with low therapy motivation: Multicentre randomised controlled trial. Psychother Psychosom. 2013;82: 319–331. pmid:23942318
  19. 19. Gardstrom SC, Diestelkamp WS. Women with addictions report reduced anxiety after group music therapy: A quasi-experimental study. Voices World Forum Music Ther. 2013;13.
  20. 20. James MR. Music therapy values clarification: A positive influence on perceived locus of control. J Music Ther. 1988;25: 206–215.
  21. 21. Mays KL, Clark DL, Gordon AJ. Treating addiction with tunes: A systematic review of music therapy for the treatment of patients with addictions. Subst Abuse. 2008;29: 51–59. pmid:19042198
  22. 22. Silverman MJ. Applying levels of evidence to the psychiatric music therapy literature base. Arts Psychother. 2010;37: 1–7.
  23. 23. Higgins JPT, Green S, editors. Cochrane handbook for systematic reviews of interventions [Internet]. The Cochrane Collaboration; 2011. Available: www.handbook.cochrane.org
  24. 24. Fisher CG, Wood KB. Introduction to and techniques of evidence-based medicine. Spine. 2007;32: S66–S72. pmid:17728684
  25. 25. American Psychological Association. Evidence-based practice in psychology. Am Psychol. 2006;61: 271–285. pmid:16719673
  26. 26. Hammer SE. The effects of guided imagery through music on state and trait anxiety. J Music Ther. 1996;33: 47–70.
  27. 27. Gottheil E, Rieger JA, Farwell B, Lieberman DL. An outpatient drug program for adolescent students: Preliminary evaluation. Am J Drug Alcohol Abuse. 1977;4: 31–41. pmid:612189
  28. 28. Stevens JR, Taylor AM. Hierarchical dependence in meta-analysis. J Educ Behav Stat. 2009;34: 46–73.
  29. 29. Silverman MJ. The effect of lyric analysis on treatment eagerness and working alliance in consumers who are in detoxification: A randomized clinical effectiveness study. Music Ther Perspect. 2009;27: 115–121.
  30. 30. Silverman MJ. The effect of a lyric analysis intervention on withdrawal symptoms and locus of control in patients on a detoxification unit: A randomized effectiveness study. Arts Psychother. 2010;37: 197–201.
  31. 31. Silverman MJ. Effects of music therapy on change and depression on clients in detoxification. J Addict Nurs. 2011;22: 185–192.
  32. 32. Silverman MJ. Effects of music therapy on change readiness and craving in patients on a detoxification unit. J Music Ther. 2011;48: 509–531. pmid:22506302
  33. 33. Silverman MJ. Effects of group songwriting on motivation and readiness for treatment on patients in detoxification: A randomized wait-list effectiveness study. J Music Ther. 2012;49: 414–429. pmid:23705345
  34. 34. Silverman MJ. Effects of music therapy on drug avoidance self-efficacy in patients on a detoxification unit: A three-group randomized effectiveness study. J Addict Nurs. 2014;25: 172–181. pmid:25514686
  35. 35. Silverman MJ. Effects of lyric analysis interventions on treatment motivation in patients on a detoxification unit: A randomized effectiveness study. J Music Ther. 2015;52: 117–134. pmid:25701046
  36. 36. Silverman MJ. Effects of educational music therapy on knowledge of triggers and coping skills, motivation, and treatment eagerness in patients on a detoxification unit: A three-group cluster-randomized effectiveness study. Korean J Music Ther. 2015;17: 81–101.
  37. 37. Silverman MJ. Effects of a single lyric analysis intervention on withdrawal and craving with inpatients on a detoxification unit: A cluster-randomized effectiveness study. Subst Use Misuse. 2016;51: 241–249. pmid:26800444
  38. 38. Murphy KM. The effects of group guided imagery and music on the psychological health of adults in substance abuse treatment. Temple University. 2008.
  39. 39. Albornoz Y. The effects of group improvisational music therapy on depression in adolescents and adults with substance abuse: a randomized controlled trial. Nord J Music Ther. 2011;20: 208–224.
  40. 40. Oklan AM, Henderson SJ. Treating inhalant abuse in adolescence: A recorded music expressive arts intervention. Psychomusicology Music Mind Brain. 2014;24: 231–237.
  41. 41. Yun SH, Gallant W. Evidence-based clinical practice: The effectiveness of music-based intervention for women experiencing forgiveness/grief issues. J Evid-Based Soc Work. 2010;7: 361–376. pmid:21082467
  42. 42. Howard AA. The effects of music and poetry therapy on the treatment of women and adolescents with chemical addictions. J Poet Ther. 1997;11: 81–102.
  43. 43. Cevasco AM, Kennedy R, Generally NR. Comparison of movement-to-music, rhythm activities, and competitive games on depression, stress, anxiety, and anger of females in substance abuse rehabilitation. J Music Ther. 2005;42: 64–80. pmid:15839734
  44. 44. Gallant W, Gorey KM, Holosko MJ, Siegel S. The use of music in group work with out-patient alcoholic couples: A pilot investigation. Groupwork. 1997;10: 155–174.
  45. 45. Hwang E-Y, Oh S-H. A comparison of the effects of music therapy interventions on depression, anxiety, anger, and stress on alcohol-dependent clients: A pilot study. Music Med. 2013;5: 136–144.
  46. 46. Jones JD. A comparison of songwriting and lyric analysis techniques to evoke emotional change in a single session with people who are chemically dependent. J Music Ther. 2005;42: 94–110. pmid:15913388
  47. 47. Silverman MJ. Music therapy and clients who are chemically dependent: a review of literature and pilot study. Arts Psychother. 2003;30: 273–281.
  48. 48. Dickerson DL, Annon JJ, Hale B, Funmaker G. Drum-Assisted recovery therapy for Native Americans (DARTNA): Results from a pretest and focus groups. Am Indian Alsk Native Ment Health Res Online. 2014;21: 35–58.
  49. 49. Gallagher LM, Steele AL. Music therapy with offenders in a substance abuse/mental illness treatment program. Music Ther Perspect. 2002;20: 117–122.
  50. 50. Dougherty KM. Music therapy in the treatment of the alcoholic client. Music Ther. 1984;4: 47–54.
  51. 51. Gardstrom SC, Bartkowski J, Willenbrink J, Diestelkamp WS. The impact of group music therapy on negative affect of people with co-occurring substance use disorders and mental illnesses. Music Ther Perspect. 2013;31: 116–126.
  52. 52. Melnyk BM, Fineout-Overholt E. Making the case for evidence-based practice. In: Melnyk BM, Fineout-Overholt E, editors. Evidence-based practice in nursing and healthcare: A guide to best practice. New York: Lippincott, Williams, & Wilkins; 2005. pp. 3–24.
  53. 53. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. BMJ. 2009;339: b2535. pmid:19622551
  54. 54. Kotarba JA, Williams ML, Johnson J. Rock music as a medium for AIDS intervention. AIDS Educ Prev. 1991;3: 47–49. pmid:2036290
  55. 55. Abdollahnejad MR. Music therapy in the Teheran Therapeutic Community. Ther Communities Int J Ther Communities. 2006;27: 147–158.
  56. 56. Dingle G, Gleadhill L, Baker F. Can music therapy engage patients in group cognitive behaviour therapy for substance abuse treatment? Drug Alcohol Rev. 2008;27: 190–196. pmid:18264881
  57. 57. Liebowitz M, Tucker MS, Frontz M, Mulholland S. Participatory choral music as a means of engagement in a veterans’ mental health and addiction treatment setting. Arts Health. 2015;7: 137–150.
  58. 58. Zanker A, Glatt MM. Experiments with music in a mental hospital. Monatsschrift Für Psychiatr Neurol. 1956;131: 215–225.
  59. 59. Rio R. Adults in recovery: A year with members of the choirhouse. Nord J Music Ther. 2005;14: 107–119.
  60. 60. Eagle CT. Music and LSD: An empirical study. J Music Ther. 1972;9: 23–36.
  61. 61. Short AD, Dingle GA. Music as an auditory cue for emotions and cravings in adults with substance use disorders. Psychol Music. 2015; 559–573.
  62. 62. Thayer Gaston E, Eagle CT. The function of music in LSD therapy for alcoholic patients. J Music Ther. 1970;7: 3–19.
  63. 63. Nerad J, Neradová L. Chromesthetic music perception by alcohol dependent persons. Int J Psychophysiol. 1989;7: 334–335.
  64. 64. Jansma A, Breteler MH, Schippers GM, De Jong CA, Van Der Staak CP. No effect of negative mood on the alcohol cue reactivity of in-patient alcoholics. Addict Behav. 2000;25: 619–624. pmid:10972455
  65. 65. Fritz TH, Vogt M, Lederer A, Schneider L, Fomicheva E, Schneider M, et al. Benefits of listening to a recording of euphoric joint music making in polydrug abusers. Front Hum Neurosci. 2015;9. pmid:26124713
  66. 66. Baker FA, Dingle GA, Gleadhill LM. “Must be the ganja”: Using rap music in music therapy for substance use disorders. In: Hadley S, editor. Therapeutic uses of rap and hip-hop. Hoboken, NJ: Taylor and Francis; 2012. pp. 321–336.
  67. 67. Seligman MEP. The effectiveness of psychotherapy: The consumer reports study. Am Psychol. 1995;50: 965–974. pmid:8561380
  68. 68. Silverman MJ. A descriptive analysis of music therapists working with consumers in substance abuse rehabilitation: Current clinical practice to guide future research. Arts Psychother. 2009;36: 123–130.
  69. 69. Koelsch S. A neuroscientific perspective on music therapy. Ann N Y Acad Sci. 2009;1169: 374–384. pmid:19673812
  70. 70. Chanda ML, Levitin DJ. The neurochemistry of music. Trends Cogn Sci. 2013;17: 179–193. pmid:23541122
  71. 71. Zatorre RJ. Musical pleasure and reward: mechanisms and dysfunction. Ann N Y Acad Sci. 2015;1337: 202–211. pmid:25773636
  72. 72. Koelsch S, Stegemann T. The brain and positive biological effects in healthy and clinical populations. In: MacDonald R, Kreutz G, Mitchell L, editors. Music, health, and well-being. Oxford: OUP; 2012. pp. 436–456.
  73. 73. Buehringer G, Behrendt S. Störungen durch Substanzkonsum: Eine Einführung [Substance use disorders: An introduction]. In: Wittchen H-U, Hoyer J, editors. Klinische Psychologie & Psychotherapie. Berlin: Springer; 2011. pp. 697–714. Available: http://link.springer.com/chapter/10.1007/978-3-642-13018-2_33
  74. 74. American Music Therapy Association. Music therpy interventions in trauma, depression, and substance abuse: Selected references and key findings [Internet]. 2008. Available: http://musictherapy.org/assets/1/7/bib_mentalhealth.pdf
  75. 75. Winkelman M. Complementary therapy for addiction: “Drumming out drugs.” Am J Public Health. 2003;93: 647–651. pmid:12660212
  76. 76. Ghetti CM. Incorporating music therapy into the harm reduction approach to managing substance use problem. Music Ther Perspect. 2004;22: 84–90.
  77. 77. Groshkova T. Motivation in substance misuse treatment. Addict Res Theory. 2010;18: 494–510.
  78. 78. De Leon G, Melnick G, Kressel D, Jainchill N. Circumstances, Motivation, Readiness, and Suitability (The CMRS Scales): Predicting retention in therapeutic community treatment. Am J Drug Alcohol Abuse. 1994;20: 495–515. pmid:7832182
  79. 79. Prochaska JO, DiClemente CC. Transtheoretical therapy: Toward a more integrative model of change. Psychother Theory Res Pract. 1982;19: 276–288.
  80. 80. Lin S-T, Yang P, Lai C-Y, Su Y-Y, Yeh Y-C, Huang M-F, et al. Mental health implications of music: Insight from neuroscientific and clinical studies. Harv Rev Psychiatry. 2011;19: 34–46. pmid:21250895
  81. 81. Magee WL, Davidson JW. The effect of music therapy on mood states in neurological patients: A pilot study. J Music Ther. 2002;39: 20–29. pmid:12015809
  82. 82. Cassileth BR, Vickers AJ, Magill LA. Music therapy for mood disturbance during hospitalization for autologous stem cell transplantation. Cancer. 2003;98: 2723–2729. pmid:14669295
  83. 83. Murphy M. A self-help group experience for substance abuse patients. Music Ther. 1983;3: 52–62.
  84. 84. Rotter JB. Generalized expectancies for internal versus external control of reinforcement. Psychol Monogr Gen Appl. 1966;80: 1–28.
  85. 85. James MR. Implications of selected social psychological theories on life-long skill generalization: Considerations for the music therapist. Music Ther Perspect. 1987;4: 29–33.
  86. 86. Gardstrom SC, Hiller J. Song discussion as music psychotherapy. Music Ther Perspect. 2010;28: 147–156.
  87. 87. Porter S, McConnell T, McLaughlin K, Lynn F, Cardwell C, Braiden H-J, et al. Music therapy for children and adolescents with behavioural and emotional problems: A randomised controlled trial. J Child Psychol Psychiatry. 2017;58: 586–594. pmid:27786359
  88. 88. Bradt J, Norris M, Shim M, Gracely EJ, Gerrity P. Vocal music therapy for chronic pain management in inner-city African Americans: A mixed methods feasibility study. J Music Ther. 2016;53: 178–206. pmid:27090149
  89. 89. Bradt J, Potvin N, Kesslick A, Shim M, Radl D, Schriver E, et al. The impact of music therapy versus music medicine on psychological outcomes and pain in cancer patients: a mixed methods study. Support Care Cancer. 2015;23: 1261–1271. pmid:25322972
  90. 90. Auerbach SM, May JC, Stevens M, Kiesler DJ. The interactive role of working alliance and counselor-client interpersonal behaviors in adolescent substance abuse treatment. Int J Clin Health Psychol. 2008;8: 617–629.
  91. 91. Connors GJ, Carroll KM, DiClemente CC, Longabaugh R, Donovan DM. The therapeutic alliance and its relationship to alcoholism treatment participation and outcome. J Consult Clin Psychol. 1997;65: 588–598. pmid:9256560
  92. 92. Dundon WD, Pettinati HM, Lynch KG, Xie H, Varillo KM, Makadon C, et al. The therapeutic alliance in medical-based interventions impacts outcome in treating alcohol dependence. Drug Alcohol Depend. 2008;95: 230–236. pmid:18329827
  93. 93. Ilgen MA, McKellar J, Moos R, Finney JW. Therapeutic alliance and the relationship between motivation and treatment outcomes in patients with alcohol use disorder. J Subst Abuse Treat. 2006;31: 157–162. pmid:16919743
  94. 94. Meier PS, Donmall MC, McElduff P, Barrowclough C, Heller RF. The role of the early therapeutic alliance in predicting drug treatment dropout. Drug Alcohol Depend. 2006;83: 57–64. pmid:16298088
  95. 95. Cook S, Heather N, McCambridge J. The role of the working alliance in treatment for alcohol problems. Psychol Addict Behav. 2015;29: 371. pmid:25961147
  96. 96. Long CG, Williams M, Midgley M, Hollin CR. Within-program factors as predictors of drinking outcome following cognitive-behavioral treatment. Addict Behav. 2000;25: 573–578. pmid:10972448
  97. 97. Barber JP, Luborsky L, Gallop R, Crits-Christoph P, Frank A, Weiss RD, et al. Therapeutic alliance as a predictor of outcome and retention in the National Institute on Drug Abuse Collaborative Cocaine Treatment Study. J Consult Clin Psychol. 2001;69: 119–124. pmid:11302268
  98. 98. Knuuttila V, Kuusisto K, Saarnio P, Nummi T. Early working alliance in outpatient substance abuse treatment: Predicting substance use frequency and client satisfaction: Early working alliance and treatment effectiveness. Clin Psychol. 2012;16: 123–135.
  99. 99. Eid M, Gollwitzer M, Schmitt M. Statistik und Forschungsmethoden. 4th ed. Weinheim: Beltz; 2015.
  100. 100. Puffer S, Torgerson DJ, Watson J. Cluster randomized controlled trials. J Eval Clin Pract. 2005;11: 479–483. pmid:16164589
  101. 101. Wilkinson L, Task Force on Statistical Inference. Statistical methods in psychology journals: Guidelines and explanations. Am Psychol. 1999;54: 594–604.
  102. 102. Rosenthal R, Jacobson L. Pygmalion in the classroom. Urban Rev. 1968;3: 16–20.