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


From December 2016 to October 2017, we screened 1589 patients. Of the 117 eligible patients (7.4%), 56 (48%) consented to participate, and 52 (44%) were randomized (see Figure). We achieved a recruitment rate of 5 patients per month. Seventeen patients were randomized to PM, 17 to STM, and 18 to AC. The mean age was 57.4 years (SD, 14.2 years), and 40% of patients were African American. The mean Acute Physiology, Age, Chronic Health Evaluation II score was 21.7 (SD, 8.7). Characteristics did not differ significantly among the 3 groups at baseline (Table 1).


Study CONSORT diagram.

We randomized patients, developed playlists, and initiated the intervention within 24 hours of enrollment for all except 1 patient. Adherence to the intervention was higher in the 2 music arms than in the AC arm (Table 2). The PM and STM groups received 80% of their eligible sessions (interquartile range, 30%-90% [PM arm], 50%- 90% [STM arm]), whereas those in the AC arm received only 30% of their sessions (interquartile range, 10%-60%) (P = .02). More patients in the AC arm withdrew after 1 or more sessions (n = 8) than did so in the PM (n = 3) or the STM (n = 3) arms. Overall, 27% of patients withdrew after receiving at least 1 session (withdrawal was based on the patient's preference once they were clinically able to make decisions, on family member input, or both). Eight patients (15%) or their family members refused at least 1 intervention session during the study (1 [6%] in the PM arm, 2 [12%] in the STM arm, and 5 [28%] in the AC arm). Ten patients (4 in the PM arm, 3 in the STM arm, and 3 in the AC arm) completed an acceptability questionnaire after discharge. Among the patients, 80% rated the music enjoyable and the duration not too long, liked receiving sessions twice a day but would prefer to choose their own music, and would enroll in a similar study again. Among the surveyed patients, 90% rated the headphones comfortable and the volume appropriate. In comments, patients noted that music made them feel normal and calm. Patients rated the audiobooks poorly with regard to enjoyment and cited that as a reason for withdrawal. Adherence and acceptability did not differ between audiobooks.

We measured the effects of music on level of consciousness, delirium incidence and severity, anxiety, pain, and use of sedatives.

Patients in the STM group had more median delirium/coma-free days by day 7 than did patients in the PM and AC groups, but the difference was not statistically significant (Table 2). Similarly, the median delirium severity during the intervention period was lower in the STM group than in the other 2 groups, but again, the difference was not statistically significant (Table 2). The median Richmond Agitation-Sedation Scale scores were slightly (but not significantly) higher in the STM group than in the PM and AC groups (Table 2).

Patients in the STM group had significant increases in heart rate and diastolic blood pressure compared with patients in the PM and AC groups (Table 2). The changes in anxiety and pain scores by day 7 did not differ significantly among the 3 groups (Table 2). Other exploratory outcomes of mobility, duration of mechanical ventilation, and mortality are shown in Table 2.

Patients in the STM group received lower mean daily doses of haloperidol, opioids, propofol, and quetiapine by day 7 than did patients in the PM and AC groups, but the differences were not statistically significant (Table 3). Other medication exposures are also shown in Table 3. No adverse safety events occurred during the study.