Effectiveness of Music Therapy on Utilization Outcomes among Surgical Patients with Mental Health Conditions: A Propensity-Score Matched Cohort Study

Abstract

Background:

Music interventions can be beneficial for reducing post-operative pain and opioid exposure, but few studies have investigated music therapy (MT)’s real-world effects on utilization outcomes. This study examined characteristics associated with receiving MT and its impacts on length of stay (LOS) and opioid utilization among surgical patients with mental health and/or substance use disorders (MSUD).

Methods:

A retrospective cohort study was conducted of 4239 inpatients (267 MT, 3972 usual care) with MSUD who underwent surgery between August 2020 and July 2023. A propensity score for receiving MT accounting for sociodemographics, pain, hospital services, and multiple comorbidities was used to create a matched cohort. The effects of MT on log-transformed LOS and mean morphine milligram equivalents (MME)/day were estimated.

Results:

Among patients receiving MT, the median (IQR) time to MT referral was 80.43 (39.27–179.31) hours (h). Patients received 3 (2–4) MT interventions, representing exposure on 18.18% (11.32%−29.81%) of admitted days. Covariates associated with receiving MT (adjusted odds ratio [95% CI]) included receiving palliative care (3.57 [2.02, 6.25]) and having a trauma/stress disorder (2.93 [2.00, 4.24]), anxiety (1.51 [1.11, 2.06]), heart failure (1.63 [1.05, 2.51]), a more resource-intensive admission (1.55 [1.08, 2.23]), and more opioids administered by 48 h (1.19 [1.04, 1.37]). In the analysis of 251 matched pairs, MT’s estimated effect on log-transformed outcomes was 1.42 [1.27, 1.58] for LOS and 0.95 [0.75, 1.19] for MME/day.

Conclusion:

Patients with MSUD who undergo surgery and receive MT are more medically complex and thus may have prolonged LOS compared with those receiving usual care. Receiving MT on 18.18% of admitted days does not meaningfully impact opioid utilization within this population. Findings inform factors that may influence the decision to refer MT. LOS effects should be interpreted with caution given the inability to account for some factors that influence LOS.