Optimizing sleep during pediatric hospitalization is an important intervention to promote healing and patient and family well-being.1–4 Repeated overnight vital sign measurements are well-recognized sleep disruptors5–7 and have been targeted by recent quality improvement initiatives to improve sleep.7–12 Existing research demonstrates that reducing overnight vital sign monitoring is safe in specific populations, but it relies largely on proxy or subjective outcome measures, without directly measuring actual patient sleep quality or duration.7–12In this issue of Hospital Pediatrics, Bitterfeld et al13 report both subjective and objective improvements in sleep duration and quality after eliminating overnight vital sign monitoring for children hospitalized on medical-surgical units on hospital medicine teams. One key finding of this study is the overreporting of sleep duration and the underreporting of wake episodes on subjective measures compared with actigraphy, highlighting the importance of objective data. Another strength of the study is its novel design. The authors compared a group of patients receiving standard overnight vital sign monitoring with a group in which overnight vital sign monitoring was deferred, using a preintervention cohort as the control. Despite potential confounding related to seasonality and inherent interpatient variability, this design is novel in the pediatric sleep promotion literature and underused in pediatrics more broadly.14 Future studies could build on this work by using matched cohorts or by employing within-patient comparisons across sequential nights with and without overnight vital sign monitoring.Beyond its methodological contributions, this study’s results lay important groundwork for future quality improvement and deimplementation efforts. One notable finding is that deferring the 4 am vital sign check may be particularly beneficial. Patients in the intervention group were more likely to remain asleep at 7 am compared with controls, with no difference observed at midnight. Targeting the 4 am vital sign monitoring represents an especially attractive intervention for future work, as it may meaningfully improve sleep for patients and families without requiring prolonged intervals between assessments, an issue that often raises safety concerns and creates barriers to deimplementation.11,12Safety remains central to efforts to reduce overnight monitoring. In acutely ill or deteriorating patients, vital sign trends provide essential information that guides diagnosis and timely intervention. Hesitation around decreasing overnight vital sign monitoring largely reflects concern about missing early signs of clinical deterioration.15,16 The challenge then lies in determining how care teams can appropriately calibrate monitoring frequency with evolving patient acuity.The broad exclusion criteria used in this study (with approximately 80%-90% of patients excluded from eligibility) highlight the inherent difficulty of defining which patients are appropriate candidates for less frequent vital sign monitoring. Others have attempted to overcome this challenge by limiting work to specific patient populations, such as those admitted with failure to thrive or hyperbilirubinemia.10 Future studies could extend this approach to additional specific diagnoses. However, limiting improvement work to a relatively small population of patients will invariably have a limited scope of impact.The authors advocate for an individualized approach to decision-making, encouraging teams to participate in daily discussions of the necessity of overnight vital sign monitoring. This strategy supports collaborative and thoughtful decision-making but still misses systematic opportunities for deimplementation. Even the studies demonstrating success with rounding checklists still note difficulty in achieving consensus on which patients are safe for less frequent monitoring.11,12Emerging applications of artificial intelligence and machine learning represent a potential avenue for addressing this challenge.15,17 A high-performing prediction model could be integrated into a clinical decision support tool within the electronic medical record, prompting clinicians to modify vital sign monitoring frequency when appropriate.18 However, interpatient variability and low signal-to-noise ratios have the potential to hamper the effectiveness of predictive models, as evidenced by the mixed performance of early warning systems.19 Rather than focusing on predicting deterioration, future efforts might have more success by attempting to identify predictors of clinical stability.Although this study specifically targeted vital sign monitoring, the authors did measure the frequency of other overnight interventions, noting an average of 6 to 7 additional overnight care activities per patient. These care activities, which included toileting events, medication administration, intravenous flushes, lab draws, starting/stopping tube feedings, and other diagnostic procedures, should continue to serve as additional targets of future quality improvement work.Interestingly, despite improved sleep in the intervention group, there were no differences between the control and intervention groups in the number of overnight care activities. As the authors note, it is possible that improved sleep resulted in part from care clustering rather than deferral of vital sign monitoring alone. Efforts to decrease room entries and promote care clustering may therefore represent complementary strategies for improving sleep without compromising safety. If the average patient requires 6 care activities per night, do we have an opportunity for a less rigid monitoring strategy in which vitals are measured twice per night with other care activities as opposed to strictly every 4 hours? Alternatively, can similar information be obtained through visual assessments during other nursing cares without measuring vital signs?Improving sleep in the hospital setting continues to be an area ripe for improvement work. Adapting this study’s use of objective outcome measures and a control group will strengthen future quality improvement work, not just when evaluating the impact of reducing overnight vital sign monitoring, but also with projects aimed at adjusting timing of other overnight care activities. Vital sign monitoring frequency remains an important area of focus. The results of this study suggest that the 4 am vital sign check may be the most influential target. No study to date, including this one, has successfully and sustainably deimplemented vital sign monitoring across broad patient populations. We continue to apply a “one size fits all” model of care and, as a result, continue to overuse vital sign monitoring. Without a clearer consensus on which patients require monitoring—and at what frequency—widespread deimplementation will remain challenging. Opportunities for future scholarship in this area remain numerous.
Jones et al. (Mon,) studied this question.