UCSF

Background: Over 4 million adults are admitted annually to intensive care units (ICU) in the US. Critically ill patients experience significant symptom burden including high rates of pain, anxiety, delirium, restlessness, dyspnea, confusion, sleep disorders, loneliness, depression, and fear. Advanced respiratory support with mechanical ventilation (MV) is the most common intervention used in critical care and is an independent risk factor for each of these co-occurring symptoms. Medications are of limited effectiveness for the management of many symptoms and also confer increased morbidity including worsening delirium, increased length of stay and the development of long-term psychological and cognitive problems. To mitigate these risks, guidelines for symptom management in the ICU include recommendations for the use of non-pharmacologic therapies such as music-based interventions. Purpose: The purpose of this dissertation study is to analyze and synthesize existing literature that measures outcomes of recorded music listening interventions (RMLIs) for the management of common symptoms in critically ill adults during MV, and to describe the effects of listening to recorded music during critical care hospitalization. Methods: This dissertation presents three manuscripts. The first is a systematic review summarizing the state of the evidence that examines RMLIs to manage common symptoms experienced by critically ill adults during MV using the 2009 PRISMA guidelines. The inclusion criteria were experimental and quasi-experimental designed studies published between January 1, 1998, and March 20, 2022. The quality of available evidence was evaluated using the Evidence Project Risk of bias tool. The second study is a grounded theory analysis of patients’ perceptions of the effects of listening to self-selected music on symptom experience during MV and critical care hospitalization. The third study examines the effects of a personally selected RMLI on ventilator-derived breathing measures of rate, depth and pattern during MV. Both prospective studies were conducted in the Surgical and Neurosurgical ICUs of a single, academic, urban, level-one trauma center and safety net hospital in San Francisco, California between August 2020, and November 2021. Results: The results of the systematic review confirm that RMLIs are effective for the treatment of anxiety and pain and also identified other common symptoms such as agitation, that may be moderated by RMLIs in adults during MV. Most studies used investigator-selected music or restricted music to a limited selection of slow tempo recordings. The broad scope of the review and heterogeneity of outcome measures confounded synthesis of the results and precluded summative recommendations for RMLIs for symptom management but highlighted important literature gaps. Specifically, few studies measured the effect of RMLIs on common symptoms such as dyspnea, confusion, delirium, sleep, loneliness, and general distress. Physiologic signs did not vary with RMLI, underscoring the lack of reliable objective instruments that measure the effects of RMLI in patients unable to self-report. Results of the qualitative study show that listening to personally selected recorded music may be beneficial for cognitive and psychological recovery in ICU. Analysis of 14 semi-structured interviews, notes, and observations revealed 6 uses of personally selected music listening in ICU: 1) Restoring consciousness; 2) Maintaining cognition; 3) Humanizing the hospital experience; 4) Providing a source of connection; 5) Improving psychological wellbeing; and 6) Resolving the problems of silence. Recorded music was rarely used to address pain or anxiety. Instead, participants reported the use of recorded music listening to address complex psychological experiences of loneliness, fear, de-situation, confusion, intrusive thoughts, threats to identity and loss of control through the use of music related memories and distraction. Listening to self-selected music provided a sense of self and a method to process trauma and grief. Additional benefits of listening to recorded music included feeling joy, pleasure, hope, resilience, and a sense of normalcy. Participants identified an aversion to being sedated as well as feelings of abandonment and restraint associated with silence. In the quantitative analysis, breathing measures theoretically associated with relaxation did not vary as hypothesized with an RMLI during both machine-controlled ventilation and spontaneous breathing modalities. In this pilot study of 16 critically ill adults during MV, analysis of breath-by-breath data did not demonstrate an effect of RMLIs on respiratory rate (breaths per minute), breath depth (tidal volume in liters or maximum pressure associated with fixed volume breaths), nor breath pattern (liters of breath per minute, breaths per liter per minute, inspiratory flow, and other indirect measures of work of breathing). The null results in this pilot were likely related to the magnitude of the variance within the small sample, high incidence of delirium, and the heterogeneity of the music selections in the intervention. Use of a noninvasive pulmonary mechanics monitor facilitated the collection of objective, rich, continuous data and is a feasible instrument that adds to the rigor of the findings. Conclusions: The morbidity associated with the symptoms experienced during MV speak to the imperative to expand the access to and use of non-pharmacologic interventions for adults during critical care hospitalization. Listening to preferred, personally selected music represents a safe, widely acceptable, equitable intervention that is likely to be of benefit for the management of common psychological and cognitive symptoms not amenable to traditional medical interventions. Future studies that use validated instruments that address clinically meaningful outcomes will provide scientific justification for the use of RMLIs and contribute to the understanding of the mechanism of action of RMLIs.