Live Music for Critically Ill Children

1. Problem definition Admission into a pediatric intensive care unit (PICU) can be a highly stressful experience, with up to 62% of children demonstrating posttraumatic stress symptoms following a PICU admission. The PICU is for many critically ill children a toxic rather than a healing environment. Children needing invasive and/or painful procedures during their PICU stay are particularly at increased risk of these sequelae, with most requiring titration of pain and sedation medications to facilitate care. Caregiver and provider concern for the unclear impact of pain and sedative medications on the developing brain and increased health risks (e.g., physiologic instability, ICU acquired weakness, withdrawal and dependence, delirium, sleep-wake cycle disturbances due to patient-ventilator asynchrony, and long-term health outcomes) are energizing interest into trials investigating ancillary approaches to comfort. It is therefore crucial to explore and validate nonpharmacologic interventions toward child comfort in the PICU to alleviate some of these hazards. One potential approach to decreasing stress and improving comfort is live-performed music therapy. Music and medicine have a long history of being entwined, going as far back as the Greek philosopher Pythagoras prescribing music to promote health. There is an increasing interest in studying music's impact on health outcomes. Among mechanically ventilated adults, music therapy decreased physiologic and psychologic responses to stress (e.g., vital signs, self-reported anxiety) and sedative use. It is important to differentiate between "music therapy" and the general provision of music, sometimes referred to as "music medicine". Both music therapy and music medicine have been found to decrease pain, increase physical, cognitive, and speech recovery, and improve quality of life. Although music therapy is provided by a credentialed therapist who uses music-based interventions for individualized goal attainment, music medicine is typically prerecorded music provided by a medical. The use of live-performed music may be more advantageous than recorded music with a critically ill population because a music therapist is trained to manipulate musical elements to facilitate the desired outcomes and can respond immediately and adequately in response to the patient's reactions. Recorded music plays continuously regardless of circumstances. Although music has been used for years in healthcare, the exact mechanisms by which it can reduce pain/anxiety are not well understood. It is known that music can modify emotional state by releasing anti-stress hormones and by activating the limbic system of the brain. According to the gate control theory of pain, distracters such as music can block certain neural pathways and diminish the amount of perceived pain. A systematic review on the use of music in mechanically ventilated adults found that music was associated with lower levels of anxiety, lower sedation requirements, improved vital signs suggesting relaxation, and improved sleep. Whether these observations can also be made in mechanically ventilated, critically ill children remains unclear as there is no literature on the efficacy of utilizing music therapy in this patient population. Objective(s): In this pilot study in mechanically ventilated, critically ill children, the investigators want to study the effects of live-music therapy on patient comfort. The investigators also want to study the effects of live-music therapy on patient and parental stress levels and parent-child interaction, use of sedato-analgesic drugs, on hemodynamics and respiration and on the occurrence of patient-ventilator synchrony.
2. Strategy Study population Mechanically ventilated children consecutively admitted to the PICU of the UMCG.

Inclusion criteria Children < 5 years old and with expected mechanical ventilatory support for at least 48 hours after inclusion. The age range is set to 5 years, as this reflects most children admitted to our PICU.

Exclusion criteria Children are excluded if they have a neurocognitive disorder, or if they are admitted post-operatively with an expected length of stay <48 hours, or a critical situation where end-of-life care is expected, or if they are deaf or suffer from any other hearing impairment, which would not allow for the Music Intervention to be administered. Also, children are excluded when the parents are unable to understand / speak Dutch.

Sample size Because the nature of the pilot study, no power analyses can be performed to determine the sample size. The sample size is therefore set at 50 patients (25 in the intervention group and 25 in the control group). The investigators will explore after 10 - 15 patients the feasibility of the intervention, determined as a combination of participation, drop-out, overstimulation (based on COMFORT-B scores), and evaluations of the intervention by parents and nurses.

Primary endpoint The main study endpoint is patient comfort measured with the Dutch Version of the COMFORT Behavior (COMFORT-B) scale.

Secondary endpoints:

• Parental stress levels
• Changes in hemodynamic variables (heart rate, blood pressure)
• Changes in respiratory parameters (respiration rate, oxygen saturation, pressure-rate product, pressure-time product)
• Daily cumulative dose of benzodiazepines and opioids
• On-demand boluses of benzodiazepines and opioids
• Number of asynchronous breaths
• DNA methylation stress genes

Randomization and blinding Patients will be randomly assigned to one of the two arms: live-music therapy or care as usual (CAU). Assignment to treatment allocation will be done through a computer-generated randomization schedule. Blinding is not possible.

Intervention to be tested: Live music therapy Upon PICU admission, children will be randomly selected for either the live-music or the care-as-usual (CAU) group. Children will be offered live-music therapy for up to two weeks, with three 30 minute sessions a week in which 10 to 20 minutes of music is provided by a trained music therapist in the presence of parents. In the sessions, the music therapist will make a tailored plan for each child to avoid overstimulation, which includes choosing the appropriate instrument, determining the child's behavioral state, and continuously monitoring the child (i.e., looking for signs of relaxation but also of overstimulation such as tension, crying movements, hiccups, yawning or frowning). The music therapist will collaborate with parents in constructing a program for the sessions. Parents will be actively involved in the sessions, to stimulate their role as caregiver and empower them. The contents of the live-music is thus constantly evaluated and adapted for the individual child.

Children in the CAU group will not receive live-music therapy.

Data acquisition Baseline data will be acquired at PICU admission to characterize the study population, including age (in weeks), gender, patient history, PICU admission indication, and the pediatric risk of mortality (PRISM) IV score to characterize disease severity.

To study the primary and secondary endpoints, the following data is collected:

• Biological parameters: Biological parameters including heart rate, respiratory rate, blood pressure and transcutaneous measured oxygen saturation (SpO2), and esophageal pressure are constantly monitored during the live-music therapy sessions, being standard care. Additionally, abdominal and brain tissue oxygen saturation will be registered, using near-infrared spectroscopy (NIRS).These data will be extracted from the monitor, in the hour before the session, during the session and for the hour after every session. Pressure-rate (PRP) and pressure-time product (PTP) are calculated as mean change in esophageal pressure during respiratory effort over the measurement period multiplied by the respiratory rate (PRP) or by inspiratory time (PTP).
• Comfort B scale: The COMFORT-B scale will be derived before, during and after each therapy session. It is currently widely used, because it is the best tool to assess sedation levels in critically ill children. The scale consists of six behavioral indicators: alertness, calmness, respiratory response or crying, physical movement, muscle tone, and facial tension. Each item is assessed based on a 5-point Likert scale, from 1 (no distress) to 5 (severe distress). The total score ranged from 6 to 30 points, and each patient needs to be observed for 2 minutes. The cutoff points of the COMFORT-B scale are 10 and 23 where less than or equal to 10 represents no undersedation and greater than or equal to 23 represents under sedated. Thus, the higher the score, the lower the comfort and the more the distress.
• DNA methylation of stress-related genes. Candidate genes that are interesting and involved in early stress related responses are mentioned in the appendix. The investigators will non-invasively collect saliva samples before the start and at the end of the study. The investigators have previously developed a unique protocol to measure DNA methylation of human DNA. DNA will be isolated, undergo a bisulphite conversion to ensure that information on methylation status is preserved and then amplified using a polymerase chain reaction (PCR). DNA methylation will be measured by locus-specific pyrosequencing. The investigators will develop a pyrosequencing assay if not available, for our candidate genes, and will use the Pyromark Q24 software to determine the methylation percentage of individual CpG sites. This will give information on methylation without insights into genomic variations or mutations, avoiding the risk to yield unwanted genetic information.
• Parental stress level: Before and at the end of the study, the investigators will ask parents to fill in the Parental Stressor Scale PICU (PSS-PICU) and Beck's Depression Inventory. The PSS-PICU is a 37 item instrument originally developed by Carter and is used to assess the stressors of PICU. It has 3 broad areas - personal family, situational and environmental stressors. It is further classified into 7 subscales with each subscale having multiple items. Each item is assessed on a 5 point Likert scale with 0 - not experienced, 1 - not stressful to 5 - extremely stressful. Each subscale score is mean of item scores and final total score is mean of seven subscale scores. The Beck's Depression Inventory will be used to assess depression symptoms in the mother. This will serve as a proxy of the parent-child interaction, because the investigators know that mothers with depressive symptoms are more prone to insecure attachment with their child. The Beck's Depression Inventory has been adapted to the Dutch situation.
• Daily cumulative dose of benzodiazepines and opioids: the investigators will calculate the daily cumulative dose of benzodiazepines and opioids on the day the live-music session is scheduled.
• Percentage of asynchronous breaths: the investigators will record pressure and flow signals from the mechanical ventilator. These signals will be visually inspected from the presence of ineffective triggering, double triggering, auto-triggering, trigger delay, flow asynchrony, delayed termination, premature termination, and expiratory asynchrony. The investigators will then calculate the percentage of asynchronous breaths as the total number of asynchronous breaths divided by the total number of breaths.

Statistical analyses All calculations will be carried out using the statistical programs SPSS version 25.0 (IBM Corp, Armonk, New York, USA) or R version 3.5.1. A value of p<0.05 will be considered in all analyses. The investigators do not expect missing data for the participants during PICU stay, as a medical researcher will be responsible for inclusion and data collection of all children. However, in the unforeseen circumstance that data is missing, the investigators will exclude these children from analyses. Regarding early childhood outcomes, we have a high participation rate (80%) at our department, and the medical researcher will contact the parents that have not yet filled out the questionnaires. The primary study endpoint will be analyzed with generalized estimating equations (since the data are repeated measurements), adjusting from randomization arm. The secondary endpoints will be analyzed using the appropriate statistical tests depending on the distribution of the variables and whether they are single of repeated measures. In addition, the investigators will perform linear and logistic regression analyses, for continuous and dichotomous variables, respectively to adjust for potential confounding. These multivariable analysis will at least include age, gender, and disease severity as covariates. The investigators will also consider potential group differences regarding admission diagnosis diagnoses and therapies, for example sedative medications.