Background: Although 1 in 10 people suffer from chronic insomnia, the prevalence among hospitalized patients is likely higher, due to common comorbidities such as depression and chronic pain in this population. We aimed to assess the prevalence of undiagnosed insomnia among hospitalized adults and examine the associations between severity of insomnia and in-hospital sleep duration and efficiency.
Methods: We conducted a prospective cohort study of general medicine inpatients age 50 and older. We excluded patients with prior diagnosed sleep disorders, were unable to walk, or had an ICU stay. Sleep duration and efficiency were measured by wrist actigraphy. Patients answered the 7-item Insomnia Severity Index (score < 7 no insomnia, 8-14 subthreshold insomnia, 15-21 moderate insomnia, 22-28 severe insomnia). Random effects linear regression models clustered by subject and controlling for patient demographics were used to test the association between insomnia severity and objective sleep metrics.
Results: From June 2010 to August 2015, 446 patients were enrolled. The majority of participants were African American (75%) and female (55%), with a mean age 65. Average inpatient sleep duration and efficiency were 5.25 hours (315 min (SD=139) and 70% (95% CI: 69, 71), respectively. Roughly 1 in 4 patients (23%) screened positive for insomnia; 58 (13%) had moderate insomnia, 46 (10%) had severe insomnia. And additional 112 (25%) had subthreshold insomnia, bringing the total with insomnia to 216 (45%). Compared to patients without insomnia, patients with insomnia obtained less sleep: those with moderate insomnia obtained ~1 hour less sleep (57.4 min [-90.8, -24.0], p=0.001) and those with severe insomnia obtained 70 minutes less sleep (70.2 min [-114.6, -25.8], p=0.002). Patients with insomnia were younger (62 vs. 67 years, p<0.001), more likely to have COPD or asthma (30% vs. 19%, p=0.005), and had higher Charlson comorbidity scores (17% vs. 7%, p=0.02) than non-insomnia patients. Patients with worse insomnia were more likely to receive pharmacologic sleeps aids in the hospital (14% vs. 5%, p<0.001) and more likely to report noise disruptions (65% vs. 48%, p=0.03). Patients with insomnia were significantly more likely to have an ER visit 30-days after discharge (27% vs. 17%, p=0.03). Although not statistically significant, patients with severe insomnia reported more 30-day readmissions after discharge (38% vs. 22%, p=0.07).
Conclusions: One in four hospitalized adults without previously diagnosed sleep disorders have insomnia symptoms. Patients with insomnia had shorter in-hospital sleep duration, lower sleep efficiency, reported more disruptions from noise, and were more likely to have an ER visit after discharge. Given this high prevalence of insomnia, it is important to train hospital staff to recognize, screen, and treat patients for insomnia, particularly younger sicker patients and those with COPD or asthma.