Effect of single session receptive music therapy on anxiety and vital parameters in hospitalized Covid-19 patients: a randomized controlled trial

Hospitalized COVID-19 patients are vulnerable to different degrees of stress disorders as well as depression, anxiety and fear. The aim of this study was to evaluate the feasibility of introducing Music therapy on site with Covid-19 patients and investigating the immediate effects a single session has on anxiety, heart rate (HR), oxygen saturation (O2Sat) and satisfaction compared to standard care. A randomized controlled trial of 40 patients was conducted. Participants were assigned to control group (CG) or the treatment group (MG). MG received an individual single session of music therapy in presence. CG received standard care. MG and CG were subjected to identical measurements (pre-during-post) of the parameters STAI-Y, HR and O2Sat. Participants in MG were asked to fill in an optional open-ended question concerning their experience with music therapy. Significant difference in anxiety levels between scores in MG and CG (34.50 (23.25–40.00) vs 45.00(38, 25–54.00); p = 0.000) was observed. MG compared to CG had statistically significantly higher values of O2Sat (97.50 (96.25–99.00) versus 96.00 (96.00–98.00); p = 0.026). Results show the feasibility of introducing music therapy as a supporting complementary/non-pharmacological intervention on site in Covid-19 patients. A single session of music therapy improves O2Sat and can significantly reduce anxiety. Trial registration: 14/10/2021 No. NCT05077306. https://www.clinicaltrials.cov.

www.nature.com/scientificreports/ Music therapy has been shown to play a valuable role in the care of patients with serious illness, helping to address physical symptoms and psychological distress 9 . Music therapy reduces pain, improves sleep quality, decreases anxiety during mechanical ventilation, induces a relaxation response without the use of medication (Mofredj et al. 2016), lowers the respiratory rate and blood pressure, and has a possible positive impact on the use of sedatives and analgesics among mechanically ventilated patients 10 . To date, however, there are no studies of music therapy application on-site with Covid 19 patients.
The primary aim of this study was to evaluate the feasibility of introducing music therapy on site with Covid 19 patients as a supporting complementary/non-pharmacological intervention.
The secondary aim was to investigate the immediate effects a single music therapy session has on anxiety, heart rate (HR), oxygen saturation (O2Sat) and satisfaction compared to standard care.

Methods
Participants. Over a 4-week period (from 15th April to 15th May 2021), all patients with SARS-COV2, hospitalized at Covid Hospital Bari, Italy were screened in the study.
Study protocol was approved by the Hospital Ethics Committee of Bari, (no. 6841-09/04/2021) and patients signed an informed consent form.
All methods were performed in accordance with the relevant guidelines and regulations.

Design and procedure
A mixed-methods approach pre -post design was used to obtain and evaluate data in 2 areas: (1) the feasibility of delivering music therapy on site in the Codiv19 hospital and (2) the immediate effects of the intervention on patients' HR, O2Sat, anxiety and satisfaction. A sample size of 20 per group (N = 40) was calculated to achieve 80% power to detect a difference of 11.0 with a SD (standard deviation) of 12.5 in anxiety score. The sample size was calculated based on the results of a previously published trial 11 .
This study employed a patient-centered approach in which the music therapist tailored interventions to patients' individual needs in that moment (Bradt et al. 2016). An interactive relational approach of receptive music therapy (Bruscia 1998a; Grocke and Wigram 2007), supplemented by adaptation of the Bonny Method Guided Imagery and Music in the medical setting (MED-GIM) 12 was used (Bruscia and Grocke 2002) 13 .
Participants were assigned to control group (CG) or treatment group (MG) by computer simple randomization. Participants in MG received an individual bedside single session of receptive music therapy by a certified music therapist-GIM fellow in presence. Each session consisted of five parts: (1) patient assessment and verbal interaction to help the patient to disclose a concern, facilitating or encouraging a state of mind of wellbeing or enjoyment that is known to the patient 13 ; (2) creation of customized play list by music therapist following both patient's assessment and the specific music elements as pulse, mood, melodic line, dynamic change, bass line, volume, timbre, rhythm, form; (3) brief relaxation exercise to help patients find an image as the focus with a positive outcome; (4) music listening, with dialogue between patient and music therapist; (5) conclusion to validate feelings and reinforce a positive experience 14 .
Participants listened to the playlist with bone conduction headphones from Ipod® and volume was controlled by the music therapist. The music therapist tailored music for each participant was based on the results of patient assessment (Robb, Carpenter, & Burns, 2011). The music therapist used music selected from classical music of the Western tradition, pop, rock, new age, soundtrack, light jazz CG received standard care.
This study was designed in accordance with the Consolidated Standards of Reporting Trials (CONSORT) recommendation for RCTs.
Measure. State Trait Anxiety Inventory Y-1(STAI-Y1) was used to measure how the subject felt in that moment. Subjects were asked to rate the intensity of their anxious feelings on 20 items on a four point scale: not at all, somewhat, moderately so, or very much so.
MG and CG were subjected to identical measurements of the parameters STAI-Y, HR and O2Sat. STAI-Y values < 40 defined absence of anxiety, between 40 and 50 mild anxiety, 51-60 moderate anxiety, and > 59 severe anxiety.
From the variables under study the parameter ΔSTAI-Y was derived, defined as the difference between the value of STAI-Y calculated at time T2 minus the value of STAI-Y calculated at time T0. Similarly, the parameters ΔO2Sat% and Δ H.R. were calculated. ΔO2Sat% was defined as the difference between the value of O2 Sat calculated at time T2 minus the value of O2Sat at time T0. ΔHR was defined as the difference between the value of O2Sat calculated at time T2 minus the value of O2Sat at time T0. STAI Y-1 was administered 5 min before session (STAI Y-1 PRE) and 15 min after session (STAI Y-1 POST) in paper form in MG and CG.
HR and O2Sat were recorded from the bedside monitor three times: start session (T0), 10 min. (T1), end session (T2). At the same time, STAI Y-1, O2Sat and HR were recorded in CG. Participants in MG were asked to fill in an optional open-ended question concerning their experience with RMUSIC THERAPY. HR and O2SAT were collected by music therapist, STAI Y-1 and optional open-ended questions were collected by a psychologist researcher.
Statistical analysis. Continuous variables were expressed as Mean ± SD and median (IQ25, IQ 75) depending on whether they were with normal or non-normal distribution. The dichotomous or non-continuous varia- Variables with normal distribution were compared with the Student's T-test for independent samples and variables with non-parametric distribution were compared with the Mann-Whitney U test. Non-parametric analysis was carried out by means of the Friedman test, and subsequently, if a significance emerged, the two-bytwo comparison between the times was carried out with the Wilcoxon test.
To correctly classify the MG with the control group, linear canonical discriminant analysis was used to create a model that optimizes the between sample classes and within-sample class distances. The cross validated accuracy percentage (CVA, %) was calculated.
All analyzes were conducted with the SPSS 23 software. Statistical significance was assumed for p value < 0.050.

Results
Forty subjects were randomized into two homogeneous subgroups of 20 patients each. Patients were divided into the 2 groups in order to have statistically similar values of age, sex and clinical parameters P/F ratio and use of NIV/CPAP. Patients in MG compared to CG had comparable values of age, sex, P/F ratio (300.56 ± 101.89 vs 267.40 ± 94.65; p = 0.293), use of CPAP or NIV (15% vs 20%; p = 0.500) (Table1).

Discriminant analysis. The derived parameters ΔSTAI-Y and ΔO2
Sat were used to verify how much the improvement in anxiety and oxyhemoglobin saturation were able to discriminate the group undergoing music therapy compared to the control group. As can be seen from Fig. 2, 90% of the subjects were correctly classified within the two groups starting from the parameters ΔSTAI-Y and ΔO2Sat. There is no overlap between the two figures as can be seen graphically (Fig. 3).

Discussion
This study supports the feasibility of introducing music therapy on site in Covid-19 patients as a supporting complementary/non-pharmacological intervention. Results show that a single session of receptive music therapy improves O2Sat and can significantly reduce anxiety.
All patients in MG completed the session of receptive music therapy and engaged openly with the music therapist, with interest and receptivity. No drop-out were registered. Despite the difficulties in speaking and interacting with patients due to personal protective equipment, music therapist was able to complete each session in safely and with acceptable discomfort. The receptive music therapy intervention did not prevent standard health procedures and did not hinder the work of health personnel, who proved to be available and collaborative.
Results show that MG patients, compared to CG, not only had significantly lower anxiety values in T2 than T0, but 70% had no more anxiety, and 30% had low anxiety values. Considering that anxiety can impose harmful effects on the course of recovery and overall well-being of Covid-19 patients, the significant reduction we found in anxiety suggests that music therapy was particularly helpful and highlights how the receptive music therapy intervention can support and contain the numerous stressful factors to which these patients are subjected.
Anxiety and stress affect and arouse sympathetic nervous system with numerous adverse responses including arterial and venous constriction, myocardial stimulation, and bronchoconstriction 15 . Music could have a positive effect on the para-sympathetic system, resulting in a relaxation response characterized by alpha brain wave frequency on the electroencephalogram, and physiologically manifesting itself as a state of muscular relaxation with regular deep breathing and lowered heart rate 16 . These physiological manifestations could be the reason why in MG there was a greater increase in O2Sat in T2 and T1 than T0. This data is important if we consider that increased breathing difficulty and fatigue were two of the most common symptoms in Covid-19 patients.
Previous studies on music therapy found improvements in anxiety, pain, quality of life, small positive effects on heart rate, respiratory rate, blood pressure, fatigue 17 and positive changes on emotional state, social interaction, and spiritual well-being 18,19 . Furthermore, when compared to pharmacological sedation, music therapy is cost-effective, has no apparent risks and can provide patients and families with physical, emotional, and cultural benefits 20 . However, no studies have yet been conducted with Covid-19 patients in such a complex and difficult setting, both for patients and for the therapists themselves.
The emotional perception of music and its associated physical effects seem to involve catecholamines, but to be engaged in an music therapy session is first and foremost an experience with music and music therapist, and cannot be reduced to physical mechanisms alone. Its therapeutic potential must principally implicate other aspects characteristic of this experience-attitudes, expectations, affects, the imagination, memory, bodily selfawareness- 21,22 . Receptive music therapy technique used in this study, together with direct interaction with music therapist, helped the patient, stimulated by the music, to find and connect to internal resources of confidence, to cope with the present stress 13 and to facilitate psychodynamic responses. www.nature.com/scientificreports/ Using a music medicine intervention, patients would have been subjected to passive listening to pre-recorded music provided to the patient by a nurse or other medical staff, without being guided and accompanied to listen to the music. They would not have been able to share their experience with anyone and they would be alone again. Furthermore, music therapy intervention allows to adapt and modify play lists in real time, based on the verbal and non-verbal responses that patients have while listening.
Due to this, our intervention was made up of 5 closely related steps. The strength of this approach was in systematically and specifically combining imagery and selected music by music therapist and patients. The listening experience involved thoughts, feelings, emotions and all the senses-visual, auditory, kinesthetic, gustatory, olfactory-.
As the surrounding environment was filled with artificial light and was distracting, patients were invited to close their eyes and to keep in touch with the images of their focus. All the patients who were involved found a focus with images from nature (sea, meadow, forest). The music prepared and selected by the music therapist then began and continued for about 15 to 20 min. While listening to the music, a dialogue between the patient and the music therapist was possible, encouraging patient's self-exploration. The patient's images that emerged while listening were guided, rather than controlled, by the music therapist. In the 5 step, the patient was able to validate his/her feelings, reinforcing the positive experience.
While listening to the first music track, many patients began to cry, then became reflective, sharing feelings, thoughts, and memories 9 . Patients reported how crying helped them release stress, anxiety and how they felt safe and secure crying along with the music. None of them stopped listening before the session ended. Within this specific approach to music, the presence of the music therapist could provide support, validating the patients' www.nature.com/scientificreports/ own emotional states through active listening, attunement, and verbal acknowledgment of feelings expressed. Music and images allowed patients to live an experience beyond that of the ward, as if they were in another place, away from the worries and tiredness of the present situation. The music played the role of the co-therapist. In this complex period of emergency, care and human needs, when Healthcare workers are called to respond to the pandemic both clinically and humanely, complementary and interdisciplinary therapies, such as music therapy can offer a valid support, facilitating the expression of emotions and memories and strengthening patients' self-awareness, social connection, and sense of personal support 22 . Thanks to this individual and customized intervention, patients in MG were able to express and communicate their feelings, shifting from an anxious and worried state, to feeling calm and content after the session. By the end of the session, many patients appeared  (Table 2).

Limitations
It was not possible to enroll a greater number of participants and no placebo was used in the control group. Blinding of participants and researcher is not present. Allocation concealment was not done. Any long-term effects of our single receptive music therapy intervention were not evaluated. The duration of the anxiety-reducing effects cannot be determined. Since all of the participants were enrolled from the same hospital, their demographic characteristics might be somewhat similar and might not reflect the characteristics of the broader population. It is also possible that patients' responses could have been biased due to the attention they were receiving from the researcher. Reports of better performance in post intervention questionnaires could have been biased and the Hawthorne effect cannot be excluded.

Conclusions
This study provides important preliminary data that support and encourage the integration of music therapy into clinical practice. Despite the difficulties of working conditions (PPE) and despite the absence in literature of experiences of music therapy on site with covid patients, this study demonstrates the feasibility of introducing music therapy as an important form of support provided in unexpected and unpredictable extreme situations, when applied appropriately and judiciously. The positive feedback of patients and the indication of a positive clinical impact would suggest the importance of further investigation into how music therapy can promote comfort and relaxation, lower anxiety, affect physiological outcomes and enhance psychosocial support (Suppl Information).

Figure 3.
Discriminating analysis between the music therapy group and the control group. STAI-y State-Trait Anxiety Inventory-y form, TO time at the start of the session, T1 time during the session, T2 time at the end of the session, ΔSTAI-Y: difference between the value of STAI-Y calculated at time T2 minus the value of ΔSTAI-Y calculated at time T0; ΔO2 Sat%: difference between the value of O2 Sat calculated at time T2 minus the value of O2 Sat at time T0. The cross validate value able to discriminate the music therapy group compared to the control group starting from parameters ΔSTAI-Y and Δ O2 Sat% was 90%.