Implementation of music in colorectal perioperative standard care—barriers and facilitators among patients and healthcare professionals

Abstract Aim Perioperative anxiety and pain are still prevalent among patients undergoing surgery. Inflammatory bowel disease and colorectal cancer patients are known to have higher anxiety rates than the general population. Perioperatively applied music intervention has been proven to be effective in reducing perioperative anxiety and pain, resulting in a decrease of intra‐operative sedative use, postoperative opioid requirement and neurohormonal stress response. IMPROVE evaluates the adherence to music intervention in colorectal perioperative standard care during systematic implementation. Method The Consolidated Framework for Implementation Research (CFIR) was used for implementation in three steps. This study addresses the first step in which barriers and facilitators for implementing perioperative music were identified by surveying patients who underwent colorectal surgery and healthcare professionals involved in perioperative care. Also, perioperative anxiety scores were assessed and data on perioperative pain was collected from the patients’ medical records. Results Fifty patients and 69 professionals (response rate 68.3%) were surveyed. For patients, all domains of the CFIR were facilitating implementation. The median reported preoperative and postoperative anxiety scores were 4.5 (1.0–7.0) and 3.0 (1.0–5.75) respectively. The median postoperative pain score on the first postoperative day was 2.8 (2.0–3.7). Also, for professionals most domains were facilitating, except for some factors related to work climate and culture among nurses. Conclusions In this study it was identified that facilitating factors for implementing music in standard perioperative care were more prominent in both patients and healthcare professionals and therefore successful implementation is probable. Also, this study provides a guideline for assessing facilitators and barriers in other settings.


INTRODUC TI ON
Preoperative anxiety is prevalent among surgical patients, described in 75% of the patients facing surgery [1,2]. Inflammatory bowel disease (IBD) and cancer patients are known to have higher anxiety rates than the general population [3][4][5]. Many of these patients undergo major surgical procedures, which cause additional psychological stressors [6,7]. Anxiety can cause resistance, which is defined as a range of negative emotions about the surgical procedures [8]. Resistance can affect the perioperative course, including more difficult induction and maintenance of anaesthesia, higher postoperative analgesic requirement and less satisfaction of the patient with the perioperative experience [8][9][10][11][12][13][14][15]. Preoperative anxiety and pain are also predictive factors for the intensity of postoperative pain [13]. Despite many efforts to optimize the treatment of acute postoperative pain, still 40%-65% of the patients experience moderate to severe postoperative pain [14,15]. Postoperative pain is related to increased morbidity, development of chronic postoperative pain, functional and quality of life impairment, poor postoperative recovery, prolonged opioid use and increased medical costs [16]. Pharmacological agents, such as opioids and benzodiazepines, are widely used interventions as treatment for anxiety and pain even though their side-effects and risks such as respiratory depression and dependency are well known [17][18][19]. Benzodiazepine usage often leads to sedation-related side-effects such as drowsiness, dizziness and light-headedness leading to falls and other accidents [20][21][22][23][24].
Therefore, non-pharmacological interventions should be considered in the perioperative standard of care of patients undergoing surgical procedures. The literature shows that perioperatively applied (recorded) music intervention can reduce preoperative anxiety and postoperative pain, resulting in a decrease of intraoperative sedative and postoperative opioid requirement [25][26][27].
Its effectiveness was also observed in the neuro-endocrine stress response, resulting in lower levels of postoperative cortisol levels [28]. As far as we know, music intervention does not pose any risks or side-effects, thus making it a safe intervention. However, even if clear evidence on effectiveness is obtained, implementation of new interventions in healthcare practice is known to be challenging [29][30][31]. Regarding perioperative music intervention, recently performed implementation studies achieved penetration rates of 36%-53% [32][33][34]. This was caused by several barriers for implementation which were not overcome before implementation.
Therefore, knowing and actively addressing these barriers in advance seems essential for a successful implementation.
The IMPROVE (IMPlementation of music inteRvention in periOperative standard carE) study was set up to achieve a sustainable implementation of music intervention in perioperative standard of care in a systematic manner [35]. This study aimed to identify the barriers and facilitators for implementing perioperative music intervention in colorectal surgery.

Study design
The IMPROVE study was a prospective monocentre study carried out in the IJsselland Hospital, Capelle aan den IJssel, the Netherlands.
The study protocol was approved by the Medical Ethics Committee of the Erasmus University Medical Centre (registration number MEC-2019-0563). Implementation was carried out according to the Consolidated Framework for Implementation Research (CFIR) [36] in three steps: (1) assessment of facilitators and barriers for implementing perioperative music intervention, (2) development of a tailored implementation strategy and (3) the implementation process. As mentioned before, this paper addresses the first step only. The complete study phase overview is depicted in Appendix S1 and was extensively described in the previously published protocol paper [35].

Study population
All patients who underwent a surgical procedure for IBD or colorectal cancer (CRC) were informed regarding the study, surveyed and written informed consent was acquired, according to the Dutch Personal Data Protection Regulation, for anxiety assessment and data collection from the patients' medical record. Healthcare professionals involved in the perioperative process, including general surgeons, anaesthesiologists, gastrointestinal (GI) surgery ward and anaesthetist nurses, specialized nurses in GI oncology (nurse practitioner [NP]/physician assistant [PA]) and stoma care, and surgery and anaesthesiology residents, were also surveyed in this phase. A dedicated team of healthcare professionals was assembled at the implementation site. The dedicated team consisted of one person from each group of professionals, the dedicated team actively engaged in the study to reach effective implementation of the intervention.

Outcomes
Preoperative and postoperative anxiety scores were assessed on the day of inclusion and postoperative day (POD) 1, using the

What does this paper add to the literature?
Perioperative anxiety and pain, which are significantly improved by music, are still prevalent among inflammatory bowel disease and colorectal cancer patients. However, implementation of new interventions is challenging. This study describes essential barriers and facilitators for successful implementation of music and provides a framework to assess them in other settings. visual analogue scale for anxiety (VAS-A) on a scale of zero to 10 which is a reliable tool for assessing anxiety in the perioperative setting [37,38]. Postoperative pain scores (numerical rating scale, NRS) were extracted from the patient's medical health record until POD 14. Assessment of facilitators and barriers for implementing perioperative music intervention in the standard of care was evaluated by using qualitative research methods in the form of surveys.

Survey development and assessment
Two different surveys, for patients (Appendix S2) and healthcare professionals (Appendix S3), were developed. Both surveys were based on four of the five domains of the CFIR: (1) inner setting, (2) outer setting, (3) characteristics of the patients and healthcare professionals and (4) intervention. The fifth domain is the implementation process itself, which was beyond the scope of this study [36]. To ensure content validity of the surveys, the first drafts were carefully revised by members of the research team (EK/BS/ ). After consensus was reached, the survey was presented to the dedicated team for input. Any valuable comments were incorporated in the last version of the survey before it was distributed amongst healthcare professionals and patients.

Sample size
A minimum reduction of 12 mm on VAS for pain (corresponding to an NRS of 1.2) has been shown as clinically relevant [39]. To measure the impact of music intervention on postoperative pain we per-

Data analysis
The response rate was defined as the percentage of all returned surveys. Questionnaires without complete knowledge assessment were excluded from analysis. Descriptive statistics were used for outcomes of the current practices regarding anxiety and pain. Continuous data were presented using median and interquartile range (IQR).
The Kruskal-Wallis test was used to analyse differences between groups. Categorical data were analysed using the chi-squared test and Fisher's exact test if at least one group sample was less than five.
Ordinal data (e.g., age and working experience) were analysed using the Kruskal-Wallis test. The Bonferroni correction was applied in the case of multiple testing in the post hoc tests. Knowledge regarding music intervention was presented by calculating a median knowledge score (MKS, IQR). For every knowledge question, a knowledge score was assessed, which ranged from 0% to 100%. The MKS was assessed by calculating the median of the MKS of all respondents.
The definitions of facilitators and barriers were based on expert opinion (research team including an implementation expert (EI)), and were expressed in percentages. An MKS under 70.0% was defined as a barrier, 70.0% or higher as a facilitator. For other barriers and facilitators a cut-off point of 50.0% was used [42]. Data were analysed using R studio version 4.0.0.

Demographics
The median age of participants was 62.5 (IQR 21.8) years. The groups of patients being operated for IBD or CRC differed significantly regarding both age (P < 0.001) and educational level (P = 0.03).

Survey results
The complete survey results can be found in Appendix S2. In total, 50 patients were included, who all completed the survey. The majority of respondents had CRC (64.0%). IBD was the second most prevalent indication, with 20.0% of patients diagnosed with Crohn's disease and 8.0% with ulcerative colitis. The other 8.0% included colectomy based on constipation due to morbus Sjögren, connective tissue disease and diverticulitis.

Patients' characteristics
The MKS was 42.9% (26.4). Educational level and reason for surgery did not have significant influence on this outcome. Respondents scored a median of 8.0 (IQR 6.0-9.0) on the importance of music in their daily life on a scale of 0-10. Patients who listened to music more frequently rated the importance of music higher (P = 0.05) in the post hoc test. The median age of the patients choosing classical music was significantly higher (P < 0.001) than patients who did not choose classical music and for pop music significantly lower (P < 0.001) than patients who did not choose pop music. The respondents who would like to receive music also had a significantly higher MKS compared to the patients who did not (P < 0.001 for all perioperative moments). Table 1 gives an overview of the facilitators and barriers found based on the surveys.

Perioperative anxiety and pain
No significant differences were found in anxiety scores preoperatively (P = 0.7) or postoperatively (P = 0.2) between CRC and IBD patients. For 29 patients preoperative pain scores of the morning before surgery were available; the median was zero (IQR 0-0) and did not differ significantly between CRC and IBD patients. The postoperative pain score (NRS 0-10) on POD 1 was 2.8 (IQR 2.0-3.7).
When comparing groups based on diagnosis, significant differences

Healthcare professionals Patients
Facilitators Current

Demographics
The health professional groups differed significantly from each other regarding age (P < 0.001), work experience (P = 0.01) and workload (P < 0.001). The Dunn post hoc test with Bonferroni correction showed that nurses and physicians were older, P < 0.001 and P < 0.001 respectively, and had more working experience, P = 0.008 and P = 0.05 respectively, than residents. Physicians (P < 0.001) and residents (P < 0.001) worked significantly more hours than nurses.

Survey results
The complete survey results can be found in Appendix S3. The  interventions reducing perioperative anxiety and pain, believed in the effectiveness of perioperative music intervention and believed that music should be part of standard care. Most of the respondents were already informed on the effects of music (92.3%, 60/65). This existing awareness of the effects of music did not affect the MKS (P = 0.40).
Respondents scored their confidence/ability in successfully applying music intervention with a median of 8.0 (IQR 7.0-8.0); no significant differences were found between healthcare professionals (nurse vs. resident vs. physician; P = 0.184). Table 2 shows what healthcare professionals require, mainly equipment and instructions, in order to successfully ensure implementation of perioperative music.

DISCUSS ION
This is the first study assessing barriers and facilitators for implementing preoperative, intra-operative and postoperative music intervention during hospital admission through surveys of postoperative patients and healthcare professionals (Appendix S4).
Furthermore, anxiety and pain in the perioperative phase before implementation were also evaluated.
A reasonably high response rate in healthcare professionals this may lead to reduction in preoperative anxiety experienced by the patients [46]. Furthermore, the use of standardized postoperative opioid regimens leads to more controlled pain experiences in patients. In the final step of the IMPROVE study assessment of the effects of the music intervention on medication use for anxiety and pain will be carried out.
This first phase of the IMPROVE study was to identify determinants which may influence the process of implementing music intervention as an alternative and/or supplement to reducing perioperative anxiety and pain. By identifying the facilitators and barriers, a tailored implementation strategy to achieve swift implementation and proper adherence to the intervention in clinical practice can be conducted. The implementation strategy will be based on the Expert Recommendations for Implementing Change (ERIC) [47].
Based on the ERIC, the proper mechanism of action will be selected separately for each strategy, since it is known that tailored multilevel strategies are more likely to be effective than single strategies [45]. For example, in this phase it was found that the knowledge regarding music intervention in healthcare is low; this is considered a barrier for implementation. The implementation strategy for this would be developing and distributing educational materials, for example pocket cards for nurses with information about how to apply the intervention, in order to handle this barrier [35]. Finally, the implementation strategies will be operationalized by using the seven dimensions proposed by Proctor et al. [48]. A guideline to improve the reporting of implementation strategies in research studies and to stimulate further identification of elements pertinent to implementation strategies is proposed by the authors.

Strengths and limitations
The most important strength of this study is that it assessed barriers and facilitators according to one of the most widely used frameworks for implementation. Second, apart from the healthcare professionals' perspectives, patients' perspectives were also assessed. Lastly, the surveys developed for this study make it possible to identify facilitators and barriers in other settings. Nevertheless, several limitations must be considered. This study is conducted in one specific surgical population and in one regional teaching hospital in the Netherlands. The identified facilitators and barriers may therefore not be completely applicable to other settings (e.g., psychiatrics population, tertiary referral hospital) and interventions.
However, the surveys developed for this study can be used to detect facilitators and barriers in other settings. Furthermore, the number of patients included is low, questioning the validity of the survey results. As for all surveys, almost one-third of the healthcare professionals did not respond which suggests that selection bias cannot be ruled out.

CON CLUS ION
In this first step of the IMPROVE study several important facilitators and barriers for implementing music intervention in the perioperative standard of care were identified, in which the facilitating factors were more prominent for a specific setting. The most important barriers for implementation were lack of knowledge regarding the intervention in both patients and healthcare professionals and the professional climate and culture factors among the nurses. Overall, the attitudes, perceptions, beliefs of patients and healthcare professionals regarding music intervention were facilitatory and therefore successful implementation is expected. Lastly, the surveys developed for this study can be used as a tool to identify facilitators and barriers in other settings.

CO N FLI C T O F I NTE R E S T
None declared. can take responsibility for the integrity of the data and the accuracy of the data analysis.

E TH I C A L A PPROVA L
The study protocol was approved by the Medical Ethics Committee of the Erasmus University Medical Centre (registration number MEC-2019-0563).

PR ER EG I S TR ATI O N O F S TU DY O R A N A LYS I S PL A N D I SCLOS U R E
No preregistration exists for the studies reported in this paper.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.