A Prospective Study of Nighttime Vital Sign Monitoring Frequency and Risk of Clinical Deterioration

The routine practice of collecting vital signs every 4 hours in hospitalized ward patients has been perpetuated since as early as 1893, but there is little evidence to support this tradition¹. Although vital signs can be indicative of impending clinical deterioration, routine nighttime vital sign monitoring adds to the already fragmented sleep of inpatients. Sleep disruptions are prevalent among ward patients and are associated several negative health outcomes, including elevated blood pressures and delirium^2,3. Overnight vital sign checks not only are an especially bothersome disruptor but also can deplete crucial health care resources. Track and trigger systems, such as the Modified Early Warning Score (MEWS), have been used to identify high-risk patients for critical interventions⁴. The present study investigated whether the MEWS could identify low-risk-patients who might forgo overnight vital sign monitoring.

Methods

We conducted a prospective cohort study of consecutive adult inpatients at a 550-bed academic institution between November 4, 2008, and August, 31 2011. All vital signs were extracted from the electronic medical record (EPIC; EPIC Systems Corporation), and a MEWS was calculated for each data set obtained on the general floors, with forward imputation used for incomplete sets. The MEWS most closely preceding 11pm each evening was used to stratify patients. The number of nighttime (11pm to 6am) disruptions for vital sign monitoring and the occurrence of adverse events, defined as intensive care unit transfers or cardiac arrests in the next 24 hours (11pm to 11pm), were compared across all MEWS categories.

Results

A total, 54,096 patients were included in the study, accounting for 182,828 patient-days on the wards and 1,699 adverse events. The patient sample was 43.0% male and had a median age of 56 years (interquartile range, 40-68 years). The median evening MEWS was 2 (interquartile range, 1-2). The adverse event rate increased with higher evening MEWS, from a rate of 5.0 per 1000 patient-days (when the MEWS was less than or equal to 1) to 157.3 per 1000 patient-days (when the MEWS was greater than or equal to 7) (P=0.003 for trend) (Figure). However, the frequency of vital sign disruptions was unchanged, with a median of 2 vital sign checks per patient per night and at least 1 disruption from vital sign collection 99.3% of the nights regardless of MEWS category. Almost half of all nighttime vital sign disruptions (45.0%) occurred in patients with a MEWS of 1 or less.

Figure.

Bars represent the adverse event rate (/1000 patient-days). Line represent nighttime vital signs (per patient median).

Discussion

To our knowledge, this is the first study of its kind to critically examine the practice of vital sign collection on medical wards. Our study found that overnight vital signs are collected frequently among ward patients regardless of their risk of clinical deterioration. The evening MEWS identified a low-risk subset of patients who had significantly fewer adverse events but had overnight vital signs taken at a similar rate as high-risk patients. This suggests that the nighttime frequency of vital sign monitoring for low-risk medical inpatients might be reduced. Such a reduction could have dramatic benefits to patient sleep, considering that vital sign checks have been shown to be the environmental factor most disruptive to patient sleep⁵. In addition to being linked to negative health outcomes and patient distress during inpatient care, sleep deprivation may be an important factor in the post-hospitalization syndrome that has recently been implicated as a cause for 30-day readmissions⁶.

Given that half of all nighttime vital sign checks are for low-risk patients, this reduction could also have significant health care resource implications. With fewer vital sign collections required on most patients, overnight nurse staffing could be moderated. This would represent a significant cost savings, considering that nursing wages account for one-quarter to one-third of a hospital’s operating budget⁷. A tailored approach would also enable a reallocation of resources away from the sleep-deprived, low-risk patients to more careful monitoring of high-risk patients. Evidence suggests that this reallocation could improve patient safety. For example, investigations have shown that adverse events on the medical wards are often closely preceded by vital sign changes, indicating that some of them may be predicted and prevented with individualized increases in monitoring⁸.

Although the cohort is large, this study is limited to a single institution. In addition, the MEWS only represents vital sign data and does not incorporate more nuanced markers of clinical status. As such, further prospective study of the safety and benefit of personalized vital sign collection is warranted. There is also promise in using new wireless sensor technology and machine-learning algorithms to individualize the frequency, reporting, and interpretation of vital signs and their trends. Routine current practice is the reflexive and inflexible collection of vital signs every 4 hours both day and night, without consideration of the patient’s risk for deterioration. Given these findings, further study of approaches to tailor vital sign collection based on risk of clinical deterioration is warranted and may help improve the patient experience and safety in hospitals.