Protocolized Receptive Music Intervention to Reduce Care Refusal in Nursing Home Residents With Major Neurocognitive Disorders

1. Background and Rationale Behavioral and psychological symptoms of dementia (BPSD) are among the most challenging aspects of dementia care, affecting up to 90% of individuals at some point in their disease course. While apathy, anxiety, and depression are the most common BPSD, agitation, aggression, and opposition to care - particularly during direct nursing care activities (bathing, dressing, medication administration) - are frequently reported by both family and professional caregivers. Care refusal (also termed resistance to care) is defined as any verbal or non-verbal attitude of refusal to accept care, food, hygiene activities, or participation in daily activities. It interferes directly with caregivers' ability to perform activities of daily living assistance, such as bathing, dressing, meal support, and medication administration. It can be misinterpreted as aggression and may lead to inappropriate therapeutic escalation, including physical restraint or psychotropic medication prescription, both of which carry significant adverse effects. In all cases, care refusal should be interpreted as a genuine expression of refusal, even when cognitive autonomy is impaired. Non-pharmacological interventions (NPIs) are recommended as first-line approaches for BPSD management by French health authorities (HAS) and international expert bodies. Among NPIs, music therapy is one of the best-evidenced techniques, with demonstrated effects on apathy, anxiety, depression, cognitive and emotional competencies, and pain. However, its potential in reducing care refusal specifically has received limited study attention.
2. Rationale for the Intervention MUSIC CARE© is a validated clinical digital therapy tool (Class 1 Medical Device, CE-marked), developed in France, that uses algorithmically generated personalized music sequences based on the principles of hypnoanalgesia. It follows a standardized "U-sequence" protocol with three phases: an induction phase (stimulating, gradually slowing), a relaxation phase (minimal instrumentation, slow tempo), and a wake phase (gradual return toward initial parameters). Sequences are exclusively composed for MUSIC CARE©, are instrumental, and are free of lyrics to minimize the risk of emotional re-activation through familiar melodies. The tool has a catalogue of 58 sequences across a wide variety of musical styles and can be adapted to individual musical preferences. MUSIC CARE© has been deployed across LNA Santé establishments since 2021. Clinical observation suggested that while the tool is available, its use in care refusal situations remains inconsistent and practitioner-dependent. Based on scientific literature indicating that psychosocial interventions require structured, protocolized implementation (at least twice weekly, minimum 12 sessions) to demonstrate measurable effects on BPSD, this study was designed to evaluate the impact of formalizing MUSIC CARE© use as a care prescription during hygiene situations.
3. Study Design OPPOZIC' is a prospective, multicenter, cluster-randomized, open-label study conducted across 15 EHPAD (nursing homes) within the LNA Santé group in France. Cluster randomization is performed at the EHPAD level: all residents from a given EHPAD are assigned to the same arm (Control, Interventional 1, or Interventional 2). This approach was chosen to ensure feasibility of uniform staff practice within each facility and to enable assessment of protocolization effects at the facility level. Randomization is centralized by LNA Santé before the first resident is enrolled in each facility. The study involves two populations:

• Residents: elderly residents with major neurocognitive disorders presenting care refusal during hygiene care (primary population; n=60)
• Caregivers (soignants): nursing and care staff performing hygiene care for enrolled residents (secondary population; approximately 10-30 per EHPAD, for a total estimated n=150-450) The study is open-label: blinding is not feasible for a behavioral/digital therapeutic intervention.

1. Intervention Description by Arm Arm A - Control: Standard Care (Current Practice) No change to current facility care practices is required. MUSIC CARE© is available to care staff and may be used at their own clinical judgment and discretion, as is currently practiced in LNA Santé facilities. When a care refusal situation is identified, the resident's medical file is reviewed for potential inclusion. This arm reflects real-world current standard of care and serves as the comparator for the protocolized intervention arms. Arm B - Interventional 1: Protocolized MUSIC CARE© During Hygiene Care (20 minutes) MUSIC CARE© is protocolized and formally integrated into the resident's personalized care plan (equivalent to a medication prescription). When a care refusal situation is identified, the care team implements a structured MUSIC CARE© intervention during hygiene care. A caregiver enters the resident's room with the MUSIC CARE© tablet and speaker/headphones at the time of care. A 20-minute U-sequence, personalized to the resident's musical preferences, is programmed and played throughout the care session (preparation, hygiene care, installation). This protocol is applied daily for 4 weeks. Arm C - Interventional 2: Protocolized MUSIC CARE© Before and During Hygiene Care (40 minutes total) The approach is identical to Arm B, with the addition of a 20-minute music session delivered before the hygiene care. The caregiver visits the resident 20 minutes before the care session to start the first U-sequence, then programs a second U-sequence during the care itself. Total daily music intervention: 40 minutes. This protocol is applied daily for 4 weeks.
2. Standard Care Program All arms continue to receive standard geriatric nursing care within the EHPAD (medical monitoring, nutritional support, social activities, other non-pharmacological approaches as clinically indicated). MUSIC CARE© use in control arm facilities is not prohibited; it is simply not protocolized.
3. Visit Schedule and Procedures Resident pathway:

• Pre-selection (Day -7) Eligibility check; information letters provided to resident and/or representative; 7-day reflection period
• Inclusion / Baseline (Day 0) Written informed consent; demographics; MMSE; NPI-ES (multidisciplinary team); psychotropic medications (last 4 weeks); start of care refusal and duration tracking
• Weeks 1-4 (S1-S4) - daily Degree of care acceptance/refusal (4-level scale); duration of hygiene care session; adverse event monitoring
• Follow-up assessment (Week 4 ± 3 days) NPI-ES; psychotropic medications (J0→S4); protocolization data (sessions completed, staff involved)
• End-of-study assessment (Week 8) NPI-ES; psychotropic medications (S4→S8); adverse events Caregiver pathway:
• Pre-selection (Day -7) Information letter provided; 7-day reflection period Inclusion (Day 0) Written informed consent; Maslach Burnout Inventory (MBI) - anonymous
• Week 4 (S4) Maslach Burnout Inventory (anonymous)
• Week 8 (S8) Maslach Burnout Inventory (anonymous)

1. Assessment Tools

4-Level Ordinal Care Acceptance/Refusal Scale - Primary Outcome Measure

A structured 4-category observational scale recorded daily by caregivers during hygiene care:

• Category 1: No refusal - complete care performed without negotiation
• Category 2: Complete care performed with negotiation
• Category 3: Partial care performed with negotiation
• Category 4: Complete refusal or failure (including paradoxical anxiety activation or emotional re-activation triggered by music) Score range: 1 (best) to 4 (worst). Category 4 is treated as complete refusal/failure in all statistical analyses (per CPP recommendation). Lower score = better care acceptance. Assessed daily, Days 1-28.

NPI-ES (Neuropsychiatric Inventory - Nursing Staff version) The NPI-ES is a validated caregiver-report instrument assessing 12 behavioral and neurovegetative domains in patients with dementia: delusions, hallucinations, agitation/aggression, depression/dysphoria, anxiety, euphoria, apathy/indifference, disinhibition, irritability/emotional lability, aberrant motor behavior, sleep, and appetite/eating disorders. Each domain score = Frequency (1-4, or 0 if absent) × Severity (1-3, or 0 if absent). Total behavioral score (10 domains): 0 to 120; higher score = more severe neuropsychiatric symptoms. A caregiver professional burden subscale (0-5 per domain) is assessed separately and is not included in the total score.

Maslach Burnout Inventory (MBI) - Caregiver Outcome

The MBI is a 22-item self-report questionnaire assessing professional burnout, completed anonymously by caregivers. Each item rated on a frequency scale (0 = never to 6 = daily). Three subscales:

• Emotional Exhaustion (9 items, total 0-54): score <17 = low burnout; 18-29 = -moderate; ≥30 = high. Higher = greater burnout.
• Depersonalization (5 items, total 0-30): score <5 = low burnout; 6-11 = moderate; ≥12 = high. Higher = greater burnout.
• Personal Accomplishment (8 items, total 0-48): score ≥40 = low burnout; 34-39 = moderate; ≤33 = high. Higher = LESS burnout (reversed direction for this subscale). (8) Statistical Analysis Plan Statistical analyses will be performed using R software (v4.3.2 or later) by the LNA Santé Research Support Unit. All tests are two-sided at a global alpha of 5%. p < 0.05 is considered statistically significant. 95% confidence intervals are reported throughout.
• Descriptive statistics: Continuous variables described as mean ± SD (with normality assessment) and median [Q1-Q3, min-max]. Categorical/discrete variables as n (%).
• Primary outcome - care acceptance/refusal scale (ordinal, 4 categories):

The primary outcome (score 1-4) may be analyzed as a discrete variable using Pearson's Chi-square or Fisher's exact test. A Poisson regression (or logistic regression focusing on score 4 = refusal/failure) will estimate odds ratios between the three randomization groups. Category 4 (paradoxical reaction/anxiety) is treated as complete refusal/failure in all analyses (per CPP requirement).

• Between-group comparisons (3 arms): ANOVA or Kruskal-Wallis for continuous variables; Chi-square or Fisher's exact test for categorical variables. Pairwise comparisons: t-test or Mann-Whitney U, Chi-square; no p-value correction given exploratory nature.
• Temporal evolution (repeated measures at J0, S4, S8): Mixed-effects models with patient/resident as random effect, using the most appropriate regression (linear, logistic, or Poisson). NPI-ES total and subscale scores, Maslach subscale scores, and psychotropic medication use will be analyzed at each time point.
• Exploratory analyses: Subgroup analyses if sample sizes allow. Multivariate models if confounders are identified.
• Sample size: 60 residents (4/center × 15 centers) over 18 months. Defined based on inclusion potential, given the exploratory nature of the study. Caregiver participation: estimated 10-30 per EHPAD (total ~150-450). (9) Benefit/Risk Analysis

Expected benefits for residents: Reduction in care refusal and distress during hygiene care; possible reduction in psychotropic medication use; improved quality of care interactions.

Expected benefits for caregivers: Facilitated hygiene care; improved caregiver-resident relationship; reduced professional burnout; enhanced sense of professional accomplishment.

Expected benefits for LNA Santé: Evidence-based protocol for MUSIC CARE© as a non-pharmacological care prescription; standardized clinical practice.

Identified risks (minimal):

• Paradoxical anxiety activation or emotional re-activation (triggered by music associated with a past trauma): managed by study exit as per principal investigator decision; treated as complete refusal (score 4) in statistical analyses per CPP requirement.

This study is classified under Loi Jardé, Category 2 (minimal risk and constraints). No ANSM authorization is required.