Associations of Ward Noise Exposure with Sleep Efficiency, Anxiety, and Length of Stay in Patients with COPD Complicated by Pneumonia

Abstract

Objective:

To examine the associations of ward noise exposure with sleep efficiency, anxiety status, and length of stay in patients with chronic obstructive pulmonary disease (COPD) complicated by pneumonia.

Methods:

This retrospective, single-center observational study included 90 hospitalized patients with COPD complicated by pneumonia. Clinical data were extracted from medical records, and ward noise and sleep data were obtained from routinely collected ward environmental and nursing monitoring records. Patients were categorized into high- and low-noise exposure groups according to 24-hour equivalent sound pressure level (Leq) using a 50-dB cutoff. Daytime and nighttime periods were defined as 06:00–22:00 and 22:00–06:00, respectively. State-Trait Anxiety Inventory scores at admission and discharge were compared between groups, and within-group changes were assessed using paired t-tests. Multivariable linear regression was performed to evaluate the associations among noise exposure, sleep efficiency, anxiety, and length of stay.

Results:

The high-noise exposure group had lower sleep efficiency and higher discharge state anxiety scores than the low-noise exposure group (both P < 0.05). State anxiety scores decreased from admission to discharge in both groups (high-noise group, P = 0.001; low-noise group, P < 0.001), whereas trait anxiety scores did not change significantly within either group (both P > 0.05). The high-noise exposure group also had a longer length of stay (P < 0.05). The 24-hour Leq was significantly correlated with nighttime Leq. Multivariable analysis showed that high-noise exposure was independently associated with decreased sleep efficiency, increased discharge state anxiety, and prolonged length of stay after adjustment for demographic factors, smoking history, disease severity, cardiovascular comorbidity, and baseline state anxiety where appropriate. Formal mediation analysis was not performed; therefore, the potential associations between noise exposure and sleep disruption, anxiety, and length of stay should be interpreted as exploratory.

Conclusion:

Ward noise exposure was associated with sleep efficiency, short-term anxiety status, and length of stay in patients with COPD complicated by pneumonia. Optimizing the inpatient acoustic environment may help improve sleep and psychological comfort, but prospective interventional studies are needed to determine whether noise reduction can shorten length of stay.

KEY MESSAGES

• (1)Ward noise exposure was associated with reduced sleep efficiency in hospitalized patients with chronic obstructive pulmonary disease complicated by pneumonia.
• (2)Patients with high-noise exposure may have higher state anxiety at discharge than those with low-noise exposure, whereas trait anxiety did not differ significantly between groups.
• (3)The association between ward noise and prolonged hospital stay may be indirectly mediated through impaired sleep quality rather than representing a direct causal relationship.
• (4)Optimizing the ward acoustic environment may support recovery as part of comprehensive inpatient management.

INTRODUCTION

Chronic obstructive pulmonary disease (COPD) complicated by pneumonia is a common cause of acute respiratory deterioration in older hospitalized patients and is associated with prolonged length of stay and increased clinical burden.[1,2] In this population, recovery depends not only on anti-infective therapy, bronchodilation, and respiratory support, but also on adequate sleep, emotional stability, and a supportive ward environment. Hospital noise is a modifiable environmental stressor, and recent evidence indicates that environmental noise exposure is associated with sleep disturbance and adverse health effects.[3,4,5] Patients with COPD and pneumonia may be particularly vulnerable because nocturnal arousal, fragmented sleep, and anxiety can aggravate dyspnea perception and reduce tolerance to treatment. However, evidence directly linking ward noise exposure to sleep efficiency, anxiety, and length of stay in this specific respiratory population remains limited. This study, therefore, retrospectively analyzed routinely collected ward noise, wearable sleep monitoring, psychological assessment, and clinical outcome data to explore the associations between ward noise exposure and sleep efficiency, anxiety status, and length of stay in patients with COPD complicated by pneumonia.

METHODS

Study Design and Ethics

This retrospective, single-center observational cohort study included hospitalized patients treated in the Department of Respiratory Medicine of Hangzhou Third People’s Hospital between January 2024 and December 2025. Bedside sound-level monitoring and wearable sleep monitoring were performed as part of routine ward environmental and nursing quality management rather than as a research intervention, and the present study analyzed these pre-existing monitoring records together with electronic medical records. The study protocol was approved by the ethics committee of Hangzhou Third People’s Hospital (Approval No. 2026KA111), and the requirement for informed consent was waived due to the retrospective use of existing clinical and monitoring data. The study did not alter clinical diagnosis, treatment decisions, or nursing arrangements.

Participants

A total of 126 hospitalized patients with COPD complicated by pneumonia were assessed for eligibility between January 2024 and December 2025. Among them, 36 patients were excluded, including 8 who did not meet the age or length-of-stay criteria, 6 with previously diagnosed severe anxiety disorder or sleep disorder, 7 with acute respiratory failure requiring mechanical ventilation, 5 with malignancy or end-stage organ failure, 4 with severe hearing impairment or cognitive dysfunction, and 6 transferred to another department, discharged against medical advice, or with incomplete monitoring records. Ultimately, 90 patients were enrolled in the final analysis.

Baseline data were extracted from medical records, including age, sex, body mass index (BMI), smoking history, COPD severity, Pneumonia Severity Index (PSI) score, and major comorbidities. COPD severity was assessed according to the Global Initiative for Chronic Obstructive Lung Disease (GOLD) spirometric criteria, based on the most recent post-bronchodilator forced expiratory volume in one second (FEV1%) predicted. Pneumonia severity was evaluated using the PSI score, which stratifies short-term mortality risk using demographic characteristics, comorbidities, physical examination, laboratory results, and radiographic variables.[6]

Inclusion criteria were as follows: age 45–85 years; diagnosis of COPD according to the GOLD criteria and the Chinese expert consensus on acute exacerbation COPD;[2,7] pneumonia diagnosis diagnosed based on symptoms, imaging, and laboratory evidence; clear consciousness after admission and ability to complete questionnaire assessment; and a hospital stay of at least 5 days.

Exclusion criteria were as follows: previously diagnosed severe anxiety disorder or sleep disorder; acute respiratory failure requiring mechanical ventilation; malignancy or end-stage organ failure; severe hearing impairment or cognitive dysfunction; transfer to another department or discharge against medical advice during hospitalization.

Noise Exposure Measurement and Grouping

Ward noise data were obtained from continuously recorded bedside monitoring using a metrologically calibrated Class 1 digital sound level meter (AWA6228+, Hangzhou Aihua Instruments Co., Ltd., Hangzhou, China), with A-weighting and fast time weighting. The instrument was calibrated before monitoring using an acoustic calibrator (AWA6221A, Hangzhou Aihua Instruments Co., Ltd.; 94 dB at 1 kHz). Monitoring was performed near the patient’s bedside, approximately 1.2 m above the floor and away from walls or major reflective surfaces to reduce potential effects of sound reflection. Equivalent continuous sound levels (Leq) and maximum sound levels (Lmax) were recorded separately for daytime (06:00–22:00) and nighttime (22:00–06:00), consistent with the WHO convention for daytime and nighttime intervals.[3]

Patients were stratified into a high-noise exposure group and a low-noise exposure group according to the 24-hour Leq, using 50 dB as the cutoff. The 24-hour Leq was selected as the grouping threshold because it reflected the overall acoustic burden during hospitalization, including both daytime care-related noise and nighttime environmental noise. Nighttime Leq (22:00–06:00) was calculated as the exposure metric corresponding to the sleep-monitoring period. The correlation between 24-hour Leq and nighttime Leq was examined to assess consistency between the grouping indicator and nighttime exposure. Lmax was used to characterize impulsive noise events, such as alarms or sudden equipment sounds, but was not included in the main regression model because peak events were irregular and the sample size was limited. At least three full days of monitoring records were available for each patient. This threshold was used for clinical interpretability and was not intended to define a universal biological cutoff. Of the 90 patients, 45 were classified into the high-noise group and 45 into the low-noise group. Grouping was based solely on objective monitoring and did not affect clinical care.

COPD Severity Assessment

COPD severity was classified according to the GOLD spirometric grading system, based on the most recent post-bronchodilator FEV1% predicted. COPD was classified into four grades: Grade 1 (mild), FEV1 ≥ 80% predicted; Grade 2 (moderate), 50% ≤ FEV1 < 80% predicted; Grade 3 (severe), 30% ≤ FEV1 < 50% predicted; and Grade 4 (very severe), FEV1 < 30% predicted. The COPD severity was assessed by trained respiratory physicians.

Outcome Measures

(1) Sleep Efficiency (Primary Outcome)

Sleep efficiency was calculated based on sleep monitoring data from wrist-worn actigraphy devices (ActiGraph wGT3X-BT, ActiGraph LLC, Pensacola, FL, USA) collected during 22:00–06:00. Patients wore the device on the non-dominant wrist. Raw triaxial acceleration data were sampled at 30 Hz and aggregated into 60-second epochs using ActiLife software (version 6.13.4; ActiGraph LLC). Sleep–wake status was estimated using the validated actigraphy algorithm embedded in the software. Recorded variables included time in bed, estimated total sleep time, wake after sleep onset, and wake periods. Sleep efficiency was calculated as total sleep time divided by time in bed and expressed as a percentage. Higher values indicated better sleep continuity. Wrist actigraphy has been widely used for objective sleep-wake estimation and has shown acceptable agreement with polysomnography for sleep duration and sleep efficiency in validation studies, although it cannot fully replace polysomnography.[8] Morning sleep logs were used to check monitoring plausibility, and the average sleep efficiency across all monitored nights was used for analysis.

(2) Anxiety (Primary Outcome)

Anxiety status was assessed using the Chinese version of the State-Trait Anxiety Inventory (STAI). The STAI includes two 20-item subscales: state anxiety, reflecting transient anxiety at the time of assessment, and trait anxiety, reflecting a relatively stable anxiety proneness. Each item is rated on a four-point scale, and each subscale ranges from 20 to 80 points, with higher scores indicating greater anxiety.[9] The Chinese version has demonstrated good reliability and factorial validity in previous validation studies, with reported Cronbach’s alpha coefficients of 0.91 for state anxiety and 0.92 for trait anxiety.[10] In the present study, the STAI demonstrated good internal consistency, with Cronbach’s alpha coefficients of 0.89 and 0.90 for the state and trait subscales, respectively. STAI data recorded within 24 hours after admission and within 24 hours before discharge were extracted for analysis.

(3) Length of Stay (Secondary Outcome)

Length of stay was defined as the number of days from admission to the medically approved discharge date, according to electronic medical records. If discharge was delayed for non-medical reasons, only days up to the medically approved discharge date were counted.

Quality Control

Data extraction was conducted by two investigators independently using a standardized form. Discrepancies were resolved by discussion with a third reviewer. Completeness and logical consistency checks were performed before analysis.

Statistical Analysis

Statistical analyses were performed using R software version 4.5.3 (R Foundation for Statistical Computing, Vienna, Austria). Continuous variables were assessed for normality and expressed as mean ± standard deviation or median (interquartile range), as appropriate. Between-group comparisons were performed using the independent-samples t-test or Mann–Whitney U test for continuous variables and the chi-square test or Fisher’s exact test for categorical variables. Within-group changes in state and trait anxiety scores from admission to discharge were assessed using paired t-tests; between-group comparisons at each time point were performed using independent-samples t-tests. Pearson correlation analysis was used to examine exploratory relationships among 24-hour Leq, nighttime Leq, sleep efficiency, discharge state anxiety score, and length of stay. Multivariable linear regression was used to examine the associations of ward noise exposure with sleep efficiency, discharge state anxiety score, and length of stay. High-noise exposure was entered as the main independent variable. Covariates were selected a priori according to clinical relevance and included age, sex, BMI, smoking history, COPD severity, PSI score, and cardiovascular comorbidity. For the model with discharge state anxiety score as the dependent variable, admission state anxiety score was additionally included to adjust for baseline anxiety. Multicollinearity was assessed using variance inflation factors (VIFs), and model assumptions were evaluated using standardized residuals, normal probability plots, and residual-versus-predicted plots. Formal mediation analysis was not conducted; therefore, any explanation involving sleep disruption as a pathway between noise exposure and length of stay was treated as exploratory. All tests were two-sided, and P < 0.05 was considered statistically significant.

RESULTS

Baseline Characteristics

No significant differences were observed between the two groups in age, sex distribution, BMI, smoking history, COPD severity, PSI score, or major comorbidities (P > 0.05), indicating good baseline comparability (Table 1).

Table 1.

Baseline Characteristics of the Two Groups [n (%)]

Characteristic	High-noise group (n = 45)	Low-noise group (n = 45)	t/χ2 value	P-value
Age (years)	68.42 ± 7.36	67.91 ± 7.12	0.334	0.739
BMI (kg/m2)	24.12 ± 2.83	23.86 ± 2.71	0.445	0.658
Male [n (%)]	29 (64.44)	27 (60.00)	0.186	0.666
Smoking history [n (%)]	26 (57.78)	24 (53.33)	0.178	0.673
Moderate-to-severe COPD [n (%)]	31 (68.89)	30 (66.67)	0.051	0.821
PSI Score (mean ± SD)	92.13 ± 12.47	90.86 ± 11.98	0.486	0.628
Cardiovascular comorbidity [n (%)]	18 (40.00)	17 (37.78)	0.046	0.830

Note: BMI, body mass index; COPD, chronic obstructive pulmonary disease; PSI, Pneumonia Severity Index.

STAI Anxiety Scores

At admission, state and trait anxiety scores did not differ significantly between groups (P > 0.05). At discharge, the high-noise exposure group had significantly higher state anxiety scores than the low-noise group (P < 0.05). In paired comparisons, state anxiety scores decreased significantly from admission to discharge in both groups (high-noise group, P = 0.001; low-noise group, P < 0.001), whereas trait anxiety scores did not change significantly within either group (both P > 0.05) (Table 2).

Table 2.

Comparison of STAI Scores between Groups (Points, Mean ± SD)

Group	State anxiety			Trait anxiety
	Admission	Discharge	Within-groupP-value	Admission	Discharge	Within-groupP-value
High-noise group (n = 45)	48.36 ± 6.12	44.82 ± 5.74	0.001	46.17 ± 5.88	45.63 ± 5.61	0.182
Low-noise group (n = 45)	47.91 ± 6.05	39.96 ± 5.28	<0.001	45.92 ± 5.71	44.98 ± 5.43	0.164
t-value	0.347	4.148	—	0.206	0.562	—
P-value	0.729	<0.001	—	0.837	0.576	—

Noise Exposure, Sleep Efficiency, and Length of Stay

Noise exposure metrics, sleep efficiency, and length of stay are summarized in Table 3. The high-noise exposure group had significantly higher 24-hour Leq, nighttime Leq, and nighttime Lmax, together with lower sleep efficiency and a longer length of stay, than the low-noise exposure group (all P < 0.001). Correlation analysis showed a strong positive correlation between 24-hour Leq and nighttime Leq (r = 0.82, P < 0.001). Nighttime Leq was negatively correlated with sleep efficiency (r = −0.49, P < 0.001).

Table 3.

Comparison of Noise Exposure Metrics, Sleep Efficiency, and Length of Stay between Groups

Group	24-hour Leq, dB	Nighttime Leq, dB	Nighttime Lmax, dB	Sleep efficiency, %	Length of stay, days
High-noise group (n = 45)	52.84 ± 2.76	51.36 ± 3.12	65.42 ± 5.18	76.84 ± 8.12	11.62 ± 2.14
Low-noise group (n = 45)	45.91 ± 2.83	43.28 ± 3.06	56.73 ± 4.96	84.21 ± 7.35	9.78 ± 1.87
t-value	11.76	12.403	8.128	4.587	4.231
P-value	<0.001	<0.001	<0.001	<0.001	<0.001

Note: Leq, equivalent continuous sound level; Lmax, maximum sound level. nighttime Leq was calculated from 22:00 to 06:00 and averaged across the same monitored nights as sleep efficiency.

Exploratory Correlation Analysis

To further explore the potential correlations between nighttime noise exposure and sleep efficiency, anxiety, and length of stay, Pearson correlation analyses were performed. Nighttime Leq was positively correlated with discharge state anxiety score (r = 0.43, P < 0.001) and length of stay (r = 0.38, P < 0.001). Sleep efficiency was negatively correlated with discharge state anxiety score (r = −0.46, P < 0.001) and length of stay (r = −0.40, P < 0.001). These findings provide preliminary evidence for interrelationships among ward noise exposure, sleep disruption, anxiety status, and length of stay, but should not be interpreted as evidence of a confirmed causal or mediation pathway.

Multivariable Analysis

Multivariable linear regression models were constructed with sleep efficiency, discharge state anxiety score, and length of stay as dependent variables. After adjustment for age, sex, BMI, smoking history, COPD severity, PSI score, and cardiovascular comorbidity, high-noise exposure remained independently associated with lower sleep efficiency and a longer length of stay. In the model for discharge state anxiety, admission state anxiety score was additionally included as a baseline psychological covariate, and high-noise exposure remained significantly associated with higher discharge state anxiety. Multicollinearity diagnostics showed no obvious collinearity among covariates, with VIF values ranging from 1.06 to 1.84 across the three models. Residual diagnostics showed approximate normality, linearity, and homoscedasticity (Table 4).

Table 4.

Results of Adjusted Multivariable Linear Regression Analyses

Dependent variable	Independent variable	β	SE	t-value	P-value	Model R2	95% CI
Sleep efficiency	High-noise exposure	–5.92	1.29	–4.589	<0.001	0.34	–8.45 to −3.39
Discharge state anxiety	High-noise exposure	2.86	0.91	3.143	0.002	0.52	1.07–4.65
Length of stay	High-noise exposure	1.53	0.39	3.923	<0.001	0.37	0.76–2.30

Note: The sleep efficiency and length-of-stay models were adjusted for age, sex, BMI, smoking history, COPD severity, PSI score, and cardiovascular comorbidity. The discharge state anxiety model was additionally adjusted for admission state anxiety score. VIF values ranged from 1.06 to 1.84 across the three models. β, regression coefficient; SE, standard error; CI, confidence interval.

DISCUSSION

Patients with COPD complicated by pneumonia often have reduced respiratory reserve, systemic inflammation, dyspnea, and frequent nighttime symptoms, making restorative sleep particularly important during hospitalization.[11] Nighttime noise may disturb sleep through repeated arousals, sleep fragmentation, sympathetic activation, and reduced sleep continuity.[4,12,13] In patients with respiratory disease, these effects may further increase dyspnea perception, exacerbate emotional stress, and impair treatment intolerance. Therefore, ward noise should be considered not only a comfort issue, but also a potentially modifiable component of supportive inpatient care.

Hospital noise sources include sustained background noise and impulsive events such as monitor alarms, staff conversations, nursing procedures, door opening, and movement of medical equipment. The present study showed that higher ward noise exposure was associated with lower sleep efficiency. In addition, 24-hour Leq was strongly correlated with nighttime Leq, supporting its use for group stratification, whereas nighttime Leq provided a more direct metric of sleep-related noise exposure. These findings are consistent with reviews showing that hospital and intensive-care unit noise can disrupt sleep and impair patient outcomes.[5,12,14]

State anxiety scores at discharge were higher in the high-noise exposure group, whereas trait anxiety scores did not differ significantly between groups. In within-group paired comparisons, state anxiety scores decreased from admission to discharge in both groups. This pattern is consistent with the construct of the STAI: trait anxiety reflects a relatively stable psychological tendency, whereas state anxiety is more sensitive to acute illness, hospitalization, environmental stimulation, and perceived discomfort. Sleep disturbance and anxiety may interact bidirectionally; impaired sleep can heighten emotional reactivity, whereas anxiety can further impair sleep initiation and maintenance.[15,16] In patients with COPD, anxiety may also amplify subjective dyspnea and reduce adherence to treatment or rehabilitation. These mechanisms may explain why noise exposure was more closely associated with state anxiety than trait anxiety.

The high-noise group had a longer average length of hospital stay. Exploratory correlation analysis further showed that higher nighttime Leq was associated with lower sleep efficiency, higher discharge state anxiety, and a longer length of stay, whereas lower sleep efficiency was associated with higher anxiety and a longer length of stay. These findings are consistent with a possible pathway linking ward noise, sleep disturbance, psychological burden, and prolonged hospitalization. However, because correlation analysis and multivariable regression cannot establish temporal mediation or causality, this pathway remains hypothetical and should be verified in prospective studies with formal mediation analysis. Length of stay is influenced by multiple factors, including baseline respiratory function, pneumonia severity, comorbidities, treatment response, discharge planning, and institutional workflow. Therefore, the observed association should be interpreted as exploratory rather than causal.

From a clinical perspective, ward noise is a modifiable environmental factor. Common nighttime noise sources, including device alarms, staff communication, corridor activity, bedside procedures, and visiting-related sound, can be addressed through workflow optimization and nursing management. Measures such as quiet-hour protocols, rational scheduling of nighttime nursing activities, alarm threshold optimization, staff education, soft-closing equipment, and environmental acoustic improvements may help reduce nighttime noise stimulation.[17,18,19] For patients with COPD complicated by pneumonia, who often require close monitoring and repeated interventions, noise control should be balanced with clinical safety. The practical goal is not absolute silence, but reduction of avoidable and impulsive noise while maintaining timely monitoring and care.

This study has several limitations. First, it was a single-center retrospective observational study with a relatively small sample size, which limits causal inference and external validity. Differences in hospital architecture, ward layout, bed occupancy, nursing workflow, alarm systems, and patient severity across institutions may influence both noise exposure and clinical outcomes. Second, although objective noise monitoring was used, the study did not fully distinguish specific noise sources, frequency characteristics, or individual sensitivity to noise. Third, although wearable sleep monitoring provides feasible continuous assessment in real-world wards, it cannot completely replace polysomnography for detailed sleep architecture evaluation. Fourth, although several clinical confounders were adjusted in the multivariable models, residual confounding from unmeasured factors, such as medication use, nocturnal oxygen therapy, cough frequency, sputum burden, and nighttime nursing interventions, cannot be excluded. Finally, no formal mediation analysis was conducted; therefore, the potential pathway linking ward noise, sleep impairment, anxiety, and prolonged length of stay remains exploratory.

CONCLUSIONS

Ward noise was associated with lower sleep efficiency, higher discharge state anxiety, and a longer length of stay in patients with COPD complicated by pneumonia. These findings support attention to the inpatient acoustic environment. However, prospective interventional studies are needed to establish causality and clarify the underlying mechanisms linking ward noise exposure with sleep and anxiety.

Availability of Data and Materials

All data generated or analyzed during this study are included in this published article. Further details are available from the corresponding author on reasonable request.

Author Contributions

YueFang Xu and Xia Feng contributed to the conception and design of the study. YinFang Wu and HongXia Liang performed data collection and clinical data extraction. YinFang Wu and HongXia Liang conducted data analysis and statistical analysis. YinFang Wu drafted the manuscript. Xia Feng critically revised and edited the manuscript. All authors reviewed and approved the final manuscript and agree to be accountable for all aspects of the work. YueFang Xu is the guarantor of this work.

Ethics Approval and Consent to Participate

This study was approved by the institutional ethics committee of Hangzhou Third People’s Hospital (Approval No. 2026KA111). Because of the retrospective study design, the requirement for informed consent was waived by the ethics committee.

Conflicts of Interest

The authors declare no conflict of interest.

Funding Statement

This research received no external funding.