Decreasing Delirium Through Music

A Randomized Pilot Trial

Sikandar H. Khan, DO, MS; Chenjia Xu, PhD; Russell Purpura, MD; Sana Durrani, MBBS; Heidi Lindroth, PhD, RN; Sophia Wang, MD; Sujuan Gao, PhD; Annie Heiderscheit, PhD, MT-BC, LMFT; Linda Chlan, PhD, RN; Malaz Boustani, MD, MPH; Babar A. Khan, MD, MS


Am J Crit Care. 2020;29(2):e31-e38. 

In This Article


The Decreasing Delirium through Music trial demonstrates the feasibility and acceptability of an innovative, scalable music intervention among patients receiving mechanical ventilation in the ICU. In contrast to music interventions used in prior randomized controlled trials in this population, our intervention did not require the daily input of a board-certified music therapist or the participation of an awake and alert patient (music preferences could be obtained from LARs).

We found high acceptability of and adherence to both PM and STM, and we were able to deliver the intervention within 24 hours of enrollment, early during the course of mechanical ventilation. We chose to investigate preferred PM because of its familiarity, and relaxing STM (60–80 beats per minute) because of its sedative-sparing and anxiolytic effects. We chose to use audiobooks because they incorporate the spoken-word elements of PM, serving as behavioral and psychosocial controls..[27] In addition, in a pediatric study, audiobooks provided adequate distraction during radiology testing..[28] We learned, however, that audiobooks had poor acceptability and adherence among our patients, who completed only 30% of eligible sessions. This finding indicates that future study designs should avoid audiobooks as a control condition and consider noise-canceling headphones as an AC device.

Results of our secondary outcomes related to delirium, level of consciousness, exposure to sedatives and antipsychotics, and duration of mechanical ventilation may suggest a possible trend toward benefit in the STM group. Unlike the results of previous studies, heart rate and blood pressure significantly increased, and exposure to benzodiazepines was higher, in the STM group than in the PM group. These findings may be confounded by the use of inotropic and vasopressor agents and our study's small sample size (see the limitations described later). The findings also suggest that the beneficial effects of music on delirium may occur through a pathway other than physiological relaxation. Anxiety and pain scores decreased among patients in the STM and AC groups, whereas the opposite trend occurred in the PM arm. Our findings promote the need for further comparison of STM with an acceptable method of AC in the highly stimulating ICU setting.

Prior studies have suggested candidate pathways by which STM may be more effective against delirium than is PM or AC. Relaxing STM may reduce delirium by exerting a sedative-sparing effect, increasing cortical engagement and cognitive processing, and promoting entrainment of the nervous system.[29] In electroencephalographic studies, classical STM increased bihemispheric communication and neural connectivity.[30] Inability to focus or shift attention is a notable feature of delirium, and functional magnetic resonance imaging studies in patients listening to music have shown increased activity in areas of the brain involved with attention.[31] Further mechanistic models are needed to explain the neurocognitive effects of music.

Our results also suggest that implementing an STM intervention may be logistically simpler than implementing a PM intervention, without loss of acceptability. Although we did not assess the dose-response effects of music, our findings suggest that 120 min/day may provide a trend toward improved delirium outcomes; we are not certain whether the potential benefits of music require twice-daily sessions or simply 120 continuous minutes of music. In a previous randomized controlled trial we performed, anxiolytic and sedative benefits occurred after patients listened to preferred, relaxing music for a mean of 79 min/day (divided among patient-initiated listening sessions).[13] Further studies, including comparisons of continuous music versus more frequent but shorter sessions, as in our study, are needed.

Strengths of our study included the assessment process, in which research assistants were blinded to the patient's grouping; the innovative intervention design; and the prospectively collected clinical data. However, our study also has certain limitations. First, our analysis was limited by the small sample size. We nevertheless obtained valuable data regarding feasibility, acceptability, attrition, recruitment rates, and playlist design. Second, the intervention was not continued after a patient was transferred from the ICU, when they are likely to be able to interact with their music devices. Third, we did not adjust data related to physiological stress for doses of vasopressors or inotropic agents, nor did we collect such data continuously. Finally, only those patients who survived the hospitalization and were able to be reached by telephone completed the acceptability questionnaire.

In this study, we found that both PM and STM (classical music) were acceptable to severely ill patients and feasibly delivered in the ICU, whereas audiobooks were not acceptable to patients. Further research is needed in order to test the efficacy of music and determine its mechanisms of action in managing delirium through the use of nonpharmacological means.