Evaluation of Nonpharmacologic Interventions and Sleep Outcomes in Hospitalized Medical and Surgical Patients: A Nonrandomized Controlled Trial

IMPORTANCE: Inadequate sleep negatively affects patients' physical health, mental well-being, and recovery. Nonpharmacologic interventions are recommended as first-choice treatment. However, studies evaluating the interventions are often of poor quality and show equivocal results. OBJECTIVE: To assess whether the implementation of nonpharmacologic interventions is associated with improved inpatient night sleep. DESIGN, SETTING, AND PARTICIPANTS: In a nonrandomized controlled trial, patients were recruited on the acute medical unit and medical and surgical wards of a Dutch academic hospital. All adults who spent exactly 1 full night in the hospital were recruited between September 1, 2019, and May 31, 2020 (control group), received usual care. Patients recruited between September 1, 2020, and May 31, 2021, served as the intervention group. The intervention group received earplugs, an eye mask, and aromatherapy. Nurses received sleep-hygiene training, and in the acute medical unit, the morning medication and vital sign measurement rounds were postponed from the night shift to the day shift. All interventions were developed in collaboration with patients, nurses, and physicians. MAIN OUTCOMES AND MEASURES: Sleep was measured using actigraphy and the Dutch-Flemish Patient-Reported Outcomes Measurement Information System sleep disturbance item bank. Other outcomes included patient-reported sleep disturbing factors and the use of sleep-enhancing tools. RESULTS: A total of 374 patients were included (222 control, 152 intervention; median age, 65 [IQR, 52-74] years). Of these, 331 were included in the analysis (195 [59%] men). Most patients (138 [77%] control, 127 [84%] intervention) were in the acute medical unit. The total sleep time was 40 minutes longer in the intervention group (control: median, 6 hours and 5 minutes [IQR, 4 hours and 55 minutes to 7 hours and 4 minutes]; intervention: 6 hours and 45 minutes [IQR, 5 hours and 47 minutes to 7 hours and 39 minutes]; P < .001). This was mainly due to a 30-minute delay in final wake time (median clock-time: control, 6:30 am [IQR, 6:00 am to 7:22 am]; intervention, 7:00 am [IQR, 6:30-7:30 am]; P < .001). Sleep quality did not differ significantly between groups. For both groups, the main sleep-disturbing factors were noises, pain, toilet visits, and being awakened by hospital staff. Sleep masks (23 of 147 [16%]) and earplugs (17 of 147 [12%]) were used most. Nightly vital sign checks decreased significantly (control: 54%; intervention: 11%; P < .001). CONCLUSIONS AND RELEVANCE: The findings of this study suggest that sleep of hospitalized patients may be significantly improved with nonpharmacologic interventions. Postponement of morning vital sign checks and medication administration rounds from the night to the day shift may be a useful way to achieve this. TRIAL REGISTRATION: Netherlands Trial Registry Identifier: NL7995