We identified 365 records through a database search. After removing duplicates and irrelevant topics, we found 51 articles concerning the topic and assessed the abstracts for eligibility. We excluded the abstract of 27 articles. Finally, we reviewed 24 eligible articles, which were in full-text format and different clinical design (review, clinical study, case study, and self-description). Eight more articles did not meet inclusion criteria and were excluded (cancer population n = 3; no chronic pain population n = 3; children population n = 1; acute pain population n = 1). At the end, we included 16 articles in the narrative review. The selection process of the referred articles is presented in Fig. 1. Table 1 shows characteristics and results of selected arts therapy studies in chronic pain management (visual art, music, written expression, and dance/movement therapy).

Selection process of referred articles

We described the results only narratively because the studies found and included in the narrative review were not uniform in terms of study design, population, intervention, nor tools to measure pain and emotional state or other variables.

Thematically, we divided arts therapy into visual art therapy, music, dance/movement therapy, and written expression area.

Of the 16 studies found, eight dealt with visual art therapy [4, 20, 31, 33,34,35, 50, 51], one with music therapy [40], four with dance/movement therapy [42, 43, 49, 60], and three with written expression [45,46,47]. The representation of individual types of arts therapy was not uniform; the least in the field of music was one and the most in the field of visual art therapy were eight. Regarding the type of study design in the included 16 studies, there were five literature reviews [4, 20, 40, 49, 51], four nonrandomized clinical studies [31, 34, 42, 60], three self-reported evaluation and case studies [33, 35, 43], and four randomized designs [45,46,47, 50]. Overall, with regard to the type of design in the included studies, no type of study design (literature review, nonrandomized clinical study, self-reported evaluation and case study, randomization in design) predominated.

Six studies were supported by grants, specifically they were from the Foundation for Art and Healing [20], Johnson and Johnson and the Society for Arts in Healthcare [50], FEDER POCI-01–0145-FEDER-007746 [43], Blue Cross Blue Shield of Michigan Foundation; Arthritis Foundation; NIH grant R01 AR049059 [46], American Psychological Association; Stony Brook University; NIH Grant R01HD39753 [47] and Marian Chace Foundation, Columbia 170333-6754; Rowan University School of Osteopathic Medicine, Stratford, NJ; KIDaF Arts and Culture Foundation, Seoul, Korea [60]. Most of the included articles, ten, stated non funding [4, 31, 33,34,35, 40, 42, 45, 49, 51].

The majority of studies included in the narrative review, namely 12, were from the USA [20, 31, 33, 35, 40, 45,46,47, 49,50,51, 60]. Another smaller number, four studies, were conducted in Canada [4], Sweden [42], Israel [34], and Portugal [43]. Variables that were measured and targeted for analysis in 16 included studies of arts therapy were following: pain intensity [4, 20, 31, 34, 35, 40, 42, 43, 45,46,47, 50, 51, 60], emotional distress, mood, or wellbeing [4, 20, 31, 33, 34, 42, 43, 45,46,47, 49,50,51, 60], social support [20] and quality of life (QoL) [20, 35]. The included articles were published between 2004 and 2020 in this layout: 2004 [51], 2005 [45], 2006 [40, 46], 2008 [47], 2009 [33, 50], 2010 [20, 42], 2011 [4], 2012 [35], 2014 [34], 2017 [60], 2018 [31], 2019 [49], and 2020 [43].

Only one study [43] out of all 16 included in the review was registered, namely by registration in ClinicalTrials.gov Identifier: NCT03493308 [43].

Of the total number of 16 included arts therapy studies, 12 were related to chronic nonmalignant diseases and chronic nonmalignant pain [4,