This study used a quasi-experimental, single-arm design with pre- and post-intervention feasibility assessment conducted between June and September 2025 in the ICU of a Type B accredited hospital in Central Java, Indonesia. Eligible patients were identified daily and consecutively recruited during the study period.

A total of 180 patients were screened between June and August 2025, of whom 152 met the inclusion criteria and were consecutively enrolled in the study (Fig.). 28 patients were excluded (did not meet criteria or declined).

Inclusion criteria were:

Exclusion criteria were:

Lavender/essential oil allergy screening involved a baseline question regarding known allergies to fragrances or essential oils – patients with known allergy were excluded from the study. For patients without a known allergy a 5-minute bedside inhalation tolerance test was performed prior to the first night; if any adverse reactions occurred, aromatherapy was withheld.

The intervention protocol consisted of multiple non-pharmacological tools applied simultaneously: 1) Eye masks: soft cotton fabric eye masks (Lanaform, Belgium) with elastic bands, designed to block ambient light; 2) Earplugs: disposable polyurethane foam earplugs (3M™ 1100, USA) providing noise reduction rating of 29 dB; 3) Music therapy: instrumental relaxation music (tempo 60–80 beats per minute, volume 40–50 dB), delivered via bedside loudspeakers (Sony SRS-XB12, Japan). Music tracks were selected from a standardized relaxation playlist validated in prior ICU studies, consisting of slow-tempo instrumental music designed to promote relaxation and reduce environmental stress [23]; 4) Lavender aromatherapy: pure lavender essential oil (Lavandula angustifolia, Now Foods®, USA) diluted to 2% concentration in distilled water, was administered using an ultrasonic diffuser (InnoGear®, China) placed 1.5 m from the patient’s bed; 5) Lighting adjustment: overhead ICU fluorescent lamps were dimmed to <50 lux at patient bedside using adjustable dimmer switches installed in each cubicle; 6) Noise reduction: non-essential alarms and conversations near the patient were minimized. Staff were instructed to maintain voice levels <50 dB during nighttime hours. Noise reduction was implemented through staff education and the creation of an appropriate environment. ICU nurses received brief instructions to silence non-essential alarms when clinically appropriate, promptly responding to active alarms, and limiting non-urgent conversations at the patient's bedside during nighttime hours. Visual reminders were posted at the patient's bedside, and compliance with these rules was monitored by the nurse on duty during each shift.

The protocol was implemented over three consecutive nights from 21:00 to 06:00. Each night at 21:00, patients were assisted with the application of eye masks and earplugs, lavender aromatherapy diffuser was activated, lights were dimmed, and background music was played for 30 minutes prior to the sleep period. Earplugs and eye masks were maintained throughout the night unless removed by the patient. ICU staff adhered to a “quiet hours” policy to reduce avoidable disturbances. Intervention fidelity was monitored by bedside nurses and recorded in feasibility logs. The intervention was delivered individually at the bedside for each patient. All procedures were implemented by trained ICU nurses, under the supervision of the research team. To enhance protocol compliance, nurses assisted patients in applying earplugs and eye masks, monitored patient comfort throughout the night, and documented compliance using structured checklists. The individual ICU patient served as the unit of analysis, consistent with the unit of assignment.

Critical alarms remained active at all times. The protocol required that only non-critical alarms could be minimized (e.g., lowering non-urgent monitor volumes or relocating equipment alarms) while continuous physiological monitoring and rapid alarm escalation procedures were maintained. Nurses were instructed to immediately return devices to full volume if any clinical deterioration was observed. Adverse events (AE) were recorded using a structured AE log; trial staff reviewed logs daily. For aromatherapy intervention, patients underwent the pre-exposure tolerance check (5 minutes) under observation; any signs of respiratory distress, rash, or intolerance led to immediate discontinuation and reporting. No intervention-related AE occurred in this cohort.

The protocol, assessed by five experts, demonstrated high content validity across four criteria: relevance (1.00), clarity (0.92), simplicity (0.88), and applicability (0.96) with an overall average scale-level content validity index (S-CVI/Ave) of 0.94.

Data were collected using validated instruments to assess the feasibility of the intervention and its preliminary associations with sleep quality and pain intensity. Sleep quality was measured using the Richards-Campbell sleep questionnaire (RCSQ) [24], a widely used five-item visual analog scale ranging from 0 (very poor) to 100 (excellent): sleep depth, sleep latency, number of awakenings, efficiency of returning to sleep, and overall sleep quality. Pain intensity was evaluated daily using the numeric rating scale (NRS) [25], a simple and reliable tool in which patients rated their pain from 0 (no pain) to 10 (worst possible pain).

Primary feasibility outcomes were: recruitment rate, protocol completion (defined as participant completion of all 3 nights), per-component adherence (checklist), acceptability, and AE. Secondary exploratory outcomes: subjective sleep quality (RCSQ) and pain intensity (NRS). Objective environmental measures (bedside LAeq noise and bedside light level) were recorded in a subsample (n = 40) using calibrated instruments. The sample size was pragmatic for a feasibility study and included all eligible patients during the 3-month period (n = 152). No formal power calculation for hypothesis testing was performed; effect size estimates from this cohort will inform sample size calculations for future randomized controlled trials.

Blinding was not implemented in this study due to the nature of the intervention, which required ICU patients and nurses to be aware of the environmental modifications. Participants were informed of the interventions they received, and ICU nurses were responsible for administering and monitoring the protocol. Outcome assessment relied on validated self-report instruments (RCSQ and NRS), which minimized observer bias, but neither patients nor assessors were blinded to study condition.

Pre-post differences were assessed with paired t-tests (unless otherwise stated). Normality of difference scores was checked using Shapiro–Wilk tests; approximate normality supported use of parametric tests. We report means with standard deviations (SD), mean differences with 95% confidence intervals, paired t-statistics and p-values, and conservative Cohen’s d effect sizes calculated using the pooled SD: d = (Mean_post − Mean_pre) / SD_pooled, where SD_pooled = √[ (SD_pre² + SD_post²)/2]. For categorical pre-post changes (e.g., RCSQ category shifts), McNemar tests were used to formally assess whether observed category transitions were statistically significant beyond descriptive visualization. Missing data: none (all 152 completed). A significance level of α = 0.05 (two-tailed) was used. Data were analyzed using IBM SPSS Statistics 31.

A total of 152 adult patients were consecutively enrolled during the study period. Baseline characteristics are shown in Table 1. All 152 participants completed the full 3-night protocol (100% completion). Adherence was calculated as the proportion of scheduled components correctly delivered per patient across 3 nights (component checklist). Overall adherence was 94%. Component adherence (n = 152) included: eye mask 98%, earplug 96%, music therapy 92%, aromatherapy 94%, lighting adjustment 95%, staff quiet hours 88%. No intervention-related AE were recorded. Objective environmental monitoring in a subsample (n = 40) indicated reductions in bedside noise and light levels during intervention hours; however, quantitative logs were incomplete, so the results are presented descriptively.

Sleep quality, measured by the RCSQ, increased markedly after the intervention, while pain intensity, assessed with the NRS, decreased substantially (Table 2). Both outcomes demonstrated large effect size estimates, indicating potential benefit within this feasibility context. Confidence intervals for mean changes indicated the precision of these estimates. No null or negative results were observed, and no mediation or causal inference analyses were performed, as the study was designed to assess feasibility and preliminary effects only.

The clinical significance of these improvements is further illustrated in Table 3, which shows categorical shifts in outcome distributions. The proportion of patients reporting high sleep quality more than tripled, while the number experiencing mild pain nearly doubled. At the same time, the incidence of severe pain decreased significantly, demonstrating its clinical relevance within this feasibility context.

This single-center feasibility study demonstrates that a standardized multimodal environmental protocol can be implemented with high completion and adherence rates in an ICU setting and is associated with preliminary, clinically meaningful improvements in patient-reported sleep quality and reductions in pain intensity. Given the single-arm design, these results should be interpreted as preliminary associations rather than as evidence of effectiveness. These results are consistent with our initial expectation that the protocol would be feasible, acceptable, and associated with preliminary improvements in patient outcomes. The absence of AE and the 100% completion rate suggest that the intervention was well tolerated, indicating that the protocol may be suitable for integration into routine ICU care pending further evidence [23–25]. Importantly, the large effect sizes observed highlight the potential value of environmental interventions when structured into a combined protocol.

The novelty of this study lies in the systematic development and validation of a protocol that integrated earplugs, eye masks, music therapy, aromatherapy, noise reduction, and lighting adjustment into a single intervention. Previous research has generally examined individual non-pharmacological components in isolation, such as earplugs, eye masks, music, or noise reduction strategies, often reporting modest or inconsistent effects on sleep and pain-related outcomes [26–28]. While fragmented interventions address only single aspects of the ICU environment, this multimodal approach integrates multiple strategies. In this feasibility study, the combined protocol was associated with larger pre- and post-intervention changes, although synergistic effects cannot be confirmed within the context of a single-arm feasibility design [29–31].

The observed improvements could be related to reductions in environmental stressors such as noise and light, as well as sensory and relaxation effects of aromatherapy and music therapy. However, alternative explanations cannot be ruled out, including reassurance and increased staff engagement [26][27]. These mechanisms align with previous international studies reporting improved sleep quality with earplugs and eye masks [11][12] and reduced anxiety and pain with music therapy and lavender inhalation [27][28][32]. However, increased patient calming and staff attention may also have contributed.

Implementation fidelity was high: adherence exceeded 94% and no safety concerns reported. Minor barriers included occasional discomfort from earplugs or masks, which required reassurance from nurses. Overall, these results underscore the feasibility of integrating environmental modifications into ICU workflows as a non-invasive, low-cost adjunct to pharmacological management. Clinically, such interventions have the potential to influence broader patient outcomes, as suggested by previous literature [33][34], although these aspects were not assessed in this feasibility study.

The intervention protocol demonstrated high feasibility and acceptability in this setting, suggesting its potential applicability to broader ICU environments pending further evaluation. Because the protocol consisted of simple, low-cost, and non-invasive environmental modifications, it could be readily adopted in a variety of healthcare settings, including hospitals in low- and middle-income countries where pharmacological options may be limited or carry greater risks. The high completion rate achieved without financial incentives underscores the relevance of these findings to real-world practice. Although the study was conducted in Type B hospital with a short follow-up period, the intervention components are universally applicable and can be easily integrated into ICU care packages worldwide. Future multicenter studies will be valuable to confirm these results and to guide global policy development in this area.

Key limitations include the single-center, single-arm pre-post study design, short follow-up period (three nights), subjective outcome measures, and potential expectancy/observer bias as the intervention was administered by staff and patients were informed about it. Objective measures were included for a subsample but full objective sleep assessment (actigraphy/polysomnography) and multicenter randomized controlled trials are recommended.

Future studies should employ randomized controlled designs with appropriate control groups to strengthen causal inference regarding the intervention’s potential effectiveness. To complement self-reported outcomes, objective measurements, such as actigraphy, polysomnography, or physiological biomarkers should be included. Multicenter trials are recommended to enhance generalizability of results across diverse ICU environments and patient populations. In addition, longer-term follow-up periods are needed to evaluate the sustained impact of environmental modifications on clinical recovery, delirium incidence, ICU length of stay, and post-discharge quality of life. Exploring cost-effectiveness and integration into standard ICU protocols may also provide valuable insights for broader implementation.

The validated multimodal environmental modification protocol demonstrated both feasibility and acceptability in this ICU study and was associated with significant improvements in patient-reported sleep quality (by 17.2% in 3 nights) and reductions in pain intensity (by 1.9% in 3 nights). These preliminary results support progression to larger randomized controlled trials with objective sleep measures and longer follow-up periods to evaluate potential effectiveness and generalizability.

Barkah Waladani was responsible for the study concept and design, conducted the experiment, and drafted the manuscript. Endah Setianingsih contributed to data acquisition, supported the implementation of the intervention, and participated in the critical revision of the manuscript. Muhammad N.W. Hidayah performed the statistical analysis, assisted in the analysis and interpretation of data, and contributed to the revision of the final manuscript. All authors approved the final version of the article.

Ethics statements. The study was conducted in accordance with the ethical principles outlined in the World Medical Association Declaration of Helsinki for biomedical research involving human subjects. The protocol was reviewed and approved by the Health Research Ethics Committee of Universitas Muhammadiyah Gombong, Approval No. 110.6/II.3. AU/F/KEPK/V/2025. Written informed consent was obtained from all patients included in the study or their legally authorized representatives.

Data availability. The data that support the findings of this study are available from the corresponding authors on reasonable request. Data and statistical methods used in the article were examined by a professional biostatistician on the Sechenov Medical Journal editorial staff.

Conflict of interest. The authors declare that there is no conflict of interests.

Financing. The study was supported by a research grant from the Ministry of Education, Culture, Research, and Technology of the Republic of Indonesia (Kemendiktisaintek), Grant No. 0070/C3/AL.04/20251.

1. List of Research Program Funding Recipients 2025, Indonesia. Number: 0070/C3/AL.04/2025. Grant No. 127/C3/DT.04.00/PL/2025 (access date: 12.12.2025).