Sleep disturbances are prevalent among hospitalized seniors and often neglected by clinicians, despite their potential for negative health outcomes. One of the primary reported causes of sleep loss in hospitals is environmental noise, a new publicly reported hospital quality measure for Medicare patients. Using objective methods, we characterized noise levels in patient rooms and its association with sleep among older inpatients.


From April to August of 2010, we conducted a prospective cohort study of community‐dwelling inpatients 50 years and older hospitalized on the University of Chicago general medicine services. Noise levels in patient rooms were measured nightly by continuous sound level monitoring to determine average and percentile noise measures. Objective sleep data were obtained using wristwatch actigraphy and analyzed to determine sleep duration and sleep efficiency. Patients reported in‐hospital sleep quality and disruptions as well as baseline sleep hygiene using standard questionnaires. Descriptive statistics were used to characterize noise levels and sleep measures. Random‐effects multivariate linear regression models, clustered by subject, were used to test the association between noise levels and inpatient sleep characteristics, controlling for patient demographics and baseline sleep hygiene.


Of 37 consented patients, 23 (62%) completed at least 1 night of actigraphy, noise monitoring, and surveys for a total of 36 nights. Patient room noise levels universally exceeded the World Health Organization (WHO) recommendations for hospitals (mean, 56.3 dB; 95% CI, 54.4–58.3 dB; vs. WHO, 30 dB; P < 0.001). Maximum noise levels were louder than a chain saw at 15 m (mean, 87.9 dB; 95% CI, 83.5–92.2 dB), and the top fifth percentile noise level (L5) was similar to dogs barking (mean, 60.6 dB; 95% CI, 59.1–62.0 dB). Inpatients slept an average of 283 minutes (95% CI, 239–326 minutes), which was significantly less than their self‐reported baseline sleep (mean, 379 minutes; 95% CI, 329–428 minutes; P = 0.008). In multivariate analyses, patients with higher recorded sound levels (loudest tercile of L5) had significantly lower sleep duration (Δ = −112 minutes, 95% CI, −218 to −6 minutes; P = 0.039) and sleep efficiency (Δ = −18.5%; 95% CI, −29.5% to −7.6%; P = 0.001). Roughly half of patients (53%) complained of noise on surveys, and these complaints were associated with significantly higher maximum noise levels (92.9 vs. 82.0 dB, P = 0.006). Surveys also revealed that the most commonly reported source of noise was staff conversation (41% of patients reported).


Hospital noise levels are far from acceptable, with maximums exceeding the noise level of a chainsaw. For patients exposed to the loudest tercile of sound levels, noise could account for almost 2 hours of sleep loss every night hospitalized. Given the magnitude of this association and the known adverse health effects of sleep loss, interventions to reduce hospital noise levels are needed.


J. Yoder ‐ University of Chicago, Pritzker School of Medicine, student; A. Fazal ‐ University of Chicago, Biological Sciences Division, student; P. Staisiunas ‐University of Chicago, Department of Medicine, employment; D. Meltzer ‐University of Chicago, Department of Medicine, employment; Agency for Healthcare Research and Quality, research funding (5U18HS016967‐04); National Institute on Aging, research funding (5K24AG031326‐03); K. Knutson ‐ University of Chicago, Department of Medicine, employment; E. Van Cauter ‐ University of Chicago, Department of Medicine, employment; V. Arora ‐ University of Chicago, Department of Medicine, employment; National Institute on Aging, research funding (1K23AG033763‐01A1)