Methods: Focus groups with nurses and hospital-based physicians identified electronic health record (EHR) default settings that hardwired nighttime vitals and promoted q8h heparin as preferred venous thromboembolism (VTE) prophylaxis as major disruptions. Patient surveys also identified pain as a major disruption. The general medicine admission order set was modified via a required question asking whether vitals need to be continued through the night. Heparin q12h or enoxaparin was promoted for VTE prophylaxis. Resident education focused on minimizing sleep disruptions and addressing nocturnal pain. While new order sets went live Sept 2015 in two units, 1 “SIESTA unit” received more nursing education and added SIESTA into standard daily huddles to help identify sleep promotion. We compared pre-post use of sleep-promoting EHR orders using two sample test for proportions. Nocturnal staff entries into patient rooms obtained via hand hygiene heat sensors (GOJO SMARTLINKTM) were analyzed using interrupted time-series analysis (installed Aug 2015). Three times a week, patients reported perceived sleep disruptions using standard validated questions that correlate with objective sleep. Changes in HCAHPS (Hospital Consumer Assessment of Healthcare Providers and Systems) top-box scores were examined for 2 items: % quiet at night (always) & % pain well controlled (always).
Results: In the 6 months pre-post SIESTA, all 612 admission orders were reviewed. Admit orders without overnight vitals rose from 3% to 34% (p < .001), and admit orders with q12h heparin or qd enoxaparin (vs q8h heparin) rose from 15% to 42% (p < .001). From 244 nights of room entries, nocturnal staff entries into patient rooms initially dropped 44% post-SIESTA (-6.3 nightly disruptions per room; level change p<0.001). A significant slope change was noted after SIESTA was added to the nurses’ huddles in January (-1.31, p<0.001, see Figure). From July 2015-March 2016, 200 patients were surveyed. In multivariable logistic regression, a significant interaction was noted such that patients in the SIESTA unit post-implementation were 3 times (OR 3.35, p<0.05) more likely to report not being disrupted by vitals and 4 times more likely (OR 4.08, p<0.05) to report not being disrupted by medications. HCAHPS scores were also substantially higher in the SIESTA unit after the intervention with no change seen in the non-SIESTA unit: %quiet at night (+7%) & pain well-controlled (+9%).
Conclusions: SIESTA is the first hospital-based intervention to assess whether combining behavioral nudges and staff education can reduce objective nocturnal disruptions for patients and improve patient experience. SIESTA resulted in a significant reduction in disruptions caused by overnight vitals, medications, and staff entries into patient rooms. The decrease in patient-reported sleep disruptions suggests that these objective decreases in nocturnal disruptions translated to improvements in patient experience.