Background:

Although sleep is important for recovery from acute illness, it is often hampered by noise in hospitals. Interestingly, some patients are more vulnerable to noise disruptions, which is now a publicly reported quality measure for hospitals. Perceived control over sleep, or the belief of personal ability to bring about a health outcome, could be responsible for this variation. The aim of this study was to assess the association between perceived control over sleep and inpatient sleep time in after controlling for noise levels.

Methods:

All non–institutionalized patients over age 50, who were ambulatory on admission and admitted to a general medicine ward were eligible. Patients with prior sleep disorders, in the hospital greater than 72 h, prior ICU stay or in respiratory isolation were excluded. Perceived control over sleep was measured using the Sleep Self–Efficacy Scale (SSE) which is a 9 item scale which ranges from 9 to 45 and asks patients to report their confidence ranging from 1 (Not Confident) to 5 (Very Confident) of one’s ability to carry out activities related to sleep (i.e., lie in bed, feeling mentally relaxed). Patients were also asked daily about whether their sleep was disrupted by noise the night before. Sleep in hospital was measured nightly using wrist actigraphy (Actiwatch). Noise level was measured nightly in dB using bedside Larson Davis sound meters. Descriptive statistics and multivariate linear regression were used to discern the association between perceived control and sleep duration controlling for noise and clustered by subject.

Results:

From April 2010 to August 2011, 76 patients (61%) were enrolled whose mean age was 67 [pm] 12 years. Two thirds were African American, and 57% were female. Mean sleep time was 329 [pm] 123 min. The Lmin in the loudest tertile was 37 dB and the Lmax in the loudest tertile was 94 dB. Roughly half (46%) of patients complained of noise. Median SSE was 35 (IQR 26–41). In unadjusted analyses, patients with high SSE had 58 min more sleep time (95% CI [15–100], P = 0.009). This effect remained significant after controlling for noise level, sex, age, and race and was equivalent to the magnitude of sleep loss for patients in the loudest rooms (-52 min). Patients with high SSE had 70% lower odds of reporting noise–disrupted sleep [OR 0.31 (0.13, 0.73), P < 0.05] in both adjusted and unadjusted analyses. Sensitivity analysis using raw SSE scores showed similar associations with sleep duration and patient report of noise–disrupted sleep.

Conclusions:

Controlling for noise levels, high Sleep Self–Efficacy among hospitalized patients is associated with longer sleep duration and fewer complaints of noise. In addition to noise control, we should consider interventions (coaching, empowerment) to boost Sleep Self–Efficacy (SSE) in order to improve in–hospital sleep.