Pain is a common symptom in critically ill adults, both in patients able and unable to self-report [1]. Guidelines recommend the use of a multimodal approach to pain management to reduce opioid use and optimize pain relief [1]. Music has been suggested as a nonpharmacologic intervention in acute and chronic care settings, but little is known about its efficacy and feasibility in the intensive care unit (ICU) [2-8]. We conducted a systematic review and meta-analysis of randomized controlled trials (RCTs) to establish the efficacy of music in the adult ICU. Music interventions of 20 to 30 minutes were effective to reduce pain by almost 2 points on a 0-10 numeric rating scale (NRS) in ICU patients able to self-report [9]. However, the effect of music on pain in ICU patients unable to self-report remains unknown. In a previous review, some studies reported that family members of ICU patients expressed their interest in participating in the music selection process and in the pain management of their loved ones [10]. Therefore, music interventions for patients unable to self-report could involve the participation of family members based on their intimate knowledge of the patient and their music preferences.

Current recommendations for music interventions in postoperative patients are to provide music in the range of 60-80 beats per minute (bpm) [11]. However, most ICUs do not have access to music therapists, who have the expertise to provide personalized music within this range. Thus, there is a need to develop an easy-to-use music intervention that produces individualized music playlists with a tempo of 60-80 bpm for ICU patients. Music streaming services allow for the music selection of specific tempo ranges and should be explored as a simple means to provide a more accessible music intervention in the adult ICU.

Studies conducted thus far have mainly focused on postsurgical patients, mechanically ventilated, and able to communicate despite the fact that many patients are likely to be unable to communicate during their ICU stay [9,12,13]. Therefore, RCTs conducted until now on the effect of music to reduce pain in critically ill adults have limited generalizability to the entirety of the ICU population, despite the knowledge that all patients can experience pain and could benefit from this nonpharmacological pain management intervention.

Another limitation in previous RCTs analyzed in the systematic review relates to sample size [9]. In 4 RCTs, no sample size calculation was reported [14-17]. In 3 RCTs, the sample size was calculated based on outcomes other than pain [18-20]. In another 3 studies, the calculated sample size required was not attained for various feasibility issues attributed to “slow” or “difficult” recruitment (due to time limit, refusal rate due to randomization or family visits, and narrow inclusion criteria), or withdrawal of ICU adult patient participants who did not like the music chosen for them [21-23]. In 1 study, it was unclear whether the sample size represented the number of ICU adult patient participants or the number of observations [24]. Inadequate power may explain why 11 of the 18 RCTs found a significant pain reduction while 7 did not. Therefore, there is a need to evaluate the feasibility of research methods by conducting a pilot RCT prior to evaluating the efficacy of any new music intervention in the adult ICU.

This study aims to (1) evaluate the acceptability and feasibility of a new patient-oriented music intervention (POMI) to reduce pain in ICU patients (primary objective); (2) evaluate the feasibility of conducting a crossover pilot RCT for intervention testing in the adult ICU (primary objective); and (3) examine the preliminary efficacy of the POMI (secondary objective).

A single-blind 2×2 crossover pilot RCT was selected for this study protocol, where patients undergo a sequence of 2 intervention periods: the POMI and the control intervention (CTL: headphones or pillow without music), with an allocation ratio of 1:1.

A single-blind 2×2 crossover pilot RCT is being conducted to evaluate the acceptability, feasibility, and preliminary efficacy of the POMI. Recruitment is planned to occur over a period of 6 months during which we aim to enroll a total of 24 patient participants. As shown in Figure 1, each participating patient is randomly assigned to 1 sequence (sequence 1 or 2) of 2 intervention periods: the POMI and the CTL (headphones or pillow without music). Patients in sequence 1 receive the POMI during the first intervention period, followed by the CTL in the second intervention period; and patients in sequence 2 receive the CTL first, followed by the POMI. Each 20-30–minute intervention period is provided before a bed turning procedure that is planned as part of the participating patient’s usual care by the nursing staff. There is a 4-hour minimum washout period between both intervention periods, with data collected during the day and evening. Data collection begins as soon as possible following recruitment, always in coordination with patient care. Therefore, the first period of data collection may occur in the daytime or in the evening.

Recruitment is conducted by the first author (MRL), who is introduced to the eligible patient and family member by the nurse caring for the eligible patient. If the eligible candidate is interested in hearing about the study, the student researcher meets with them at the ICU bedside to present the study, provides a copy of the informed consent form, and answers any question. The eligible candidate is then given time to think about whether they are interested in the study. ICU nursing staff (ie, nurses and orderlies) involved in the turning procedure during the study are invited individually prior to the scheduled bed turning procedure. Eligible nursing staff are given the informed consent form and can choose to participate at their convenience. Eligible criteria are detailed in Table 1.

For the equal allocation of both groups, 24 opaque envelopes were prepared in advance by an independent member of the research team, with the use of a digitally generated list [30]. Once patients or their representatives consent to participate and agree to randomization, patient participants are categorized as either able to self-report or unable to self-report, producing 2 strata. Within each stratum, patients are randomized to either sequence 1 or sequence 2, following a permuted block randomization to ensure balance within each stratum. A block size of 4 was used (3 blocks per stratum) for a sample size of 24 ICU adult patient participants (in each block, 2 patient participants will be assigned to sequence 1 and 2 patient participants will be assigned to sequence 2, and the ordering will be random).

The CTL consists of providing a 20-30–minute period without music either while wearing headphones (Bose, QuietComfort 35) or with the head resting on a music pillow (MusiCure, hospital-grade), which is consistent with the mode of delivery for the POMI. Adult patients admitted to the ICU and able to self-report can choose 1 of the 2 modes of delivery. Patients unable to self-report are given the music pillow.

To allow for any pain-reducing effect of the POMI to dissipate before the CTL, a washout period of at least 4 hours is scheduled between both intervention periods, based on previous data on the duration of the analgesic effect of music on pain [32].

Video will be recorded for the duration of the intervention period and at each pain assessment time point. These video recordings will be used to have an independent member of the research team evaluate: (1) the Critical-Care Pain Observation Tool (CPOT) scores for each patient participant (to ensure blinding of the research team member interrater to sequence allocation) and (2) the intervention fidelity.

Ethics approval was submitted in July 2021 and approved in December 2021 (Project #2022-3005). Participation in this research project is voluntary and ongoing for all participants. Participants are free to refuse to participate and may withdraw from this research study at any time, without having to give a reason, and without any consequence to them now or in the future. The participant’s decision not to participate in the study, or to withdraw from it, will have no impact on the quality of care and services to which they are otherwise entitled. Participants are free to refuse to answer any question and remain in the study. Participants are free to refuse to be video recorded.

For a patient who is unable to consent to participate in the study, a family member representative will provide the written consent on behalf of the patient. From this time and until the patient participant is discharged from the ICU, the student researcher will follow up on the ICU adult patients who were unable to consent to determine if they regain the ability to consent for themselves. In the case where a patient participant who was previously unable to consent regains the ability to consent at any time, before or after the intervention, the student researcher will present the research project and the information and consent form to allow the ICU adult patient to make an informed decision regarding their participation in the study.

The data collected from a participant as part of this study, excluding the video recordings, could be used for future research projects related to this study only with the participant’s explicit permission. The results of the research study, excluding video recordings, may be presented at conferences, published in specialized journals or be the subject of scientific discussions, or be used for teaching purposes. No identifying information will be published in any way.

At any time, participants have the right to consult their study file in order to verify the information gathered and to have it corrected if necessary. All study data will be stored safely for 10 years, after which time they will be permanently destroyed by being either shredded or permanently deleted from the server.

Patients and families will be offered the hyperlink to their playlist as well as a paper version of the list of songs played as part of the study. Nursing staff participants will be offered a US $20 gift certificate to compensate for their time spent participating in the study.

One strength of this study is the evaluation of the intervention acceptability from multiple sources: ICU patients, families, and nursing staff. This will allow us to gain access to the different perspectives of the various stakeholders and acquire a more comprehensive understanding of the overall acceptability of the intervention. The crossover design is another strength as it allows each patient participant to be their own control, thus reducing confounding factors that are usually present in between-subject designs [50,51]. Moreover, the crossover design will allow for patient participants able to self-report to share their preference between the POMI and the CTL because they will experience both interventions. Information about participant preference will add rich qualitative data beyond the quantitative comparison of preliminary efficacy [41,51]. The crossover design is relevant to use during procedures that are planned within a short period of time (eg, molar extractions at 2 different times) [50]. In addition to intervention comparisons and patient participant preferences, we will be able to determine the pain and distress differences individually (numerically) and compute the proportion of patient participants for whom the treatment was effective in reducing pain by more than 1 point on a 0-10 NRS, for example [51]. The crossover design will also enable us to describe whether any reduction in pain or distress was qualified as meaningful by the patient participant (for those able to self-report). These data could later contribute to the understanding of the minimally clinically significant difference in procedural pain in the adult ICU population [51]. Knowing that patient participants will receive both interventions as part of the crossover design may allow them to think more objectively about pain levels in each intervention period (thus minimizing the placebo effect) compared to if they participated in a parallel-group trial, in which the patient participants’ hope to receive the intervention may influence their pain rating [50]. Finally, blinding patient participants unable to self-report will reduce the risk of bias based on these participants’ expectations of music efficacy to reduce pain.

The Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials recommendations state that the use of a crossover trial at an earlier stage study, followed by confirmation of the results in a larger parallel-group study, is an efficient approach, as long as a washout period is carefully planned to minimize or avoid carryover effects [50]. To allow for any pain-reducing effect of the POMI to dissipate before the CTL, there will be a washout period of at least 4 hours between both intervention periods, based on previous data on the duration of the analgesic effect of music on pain [32]. Due to the sample size and to evaluate the time effect, the Friedman and Wilcoxon rank sum tests will be used when analyzing pain-related data for each subgroup [51]. To avoid response shifts due to the crossover design, baseline pain intensity will be measured for each intervention period (T₀ in Figure 1). Because blinding will be difficult with patient participants receiving both the POMI and the CTL, there is a risk of ascertainment bias for patient participants. To address this, patient participants able to self-report will be asked about their perception of how effective they think the music is compared to the CTL. To minimize possible interruptions and noise during intervention periods, the nurse responsible for the patient will be met to discuss and plan for the best time to start the intervention period, prior to a scheduled bed turn.

Due to the current context of the pandemic and this study being conducted in an ICU setting, it is possible that limited family visits and nursing staff shortages will impact recruitment from these populations. To mitigate these possible limitations, meeting families at the bedside will be coordinated with the nursing staff.