Automated Respiration Rate to Improve Accuracy of the Electronic Cardiac Arrest Risk Triage Score (eCART) Algorithm

Overview

Currently, breathing rate and heart rate are checked by nursing staff manually every few hours and entered into a patient's medical record. The investigators are doing this study to see if a device that will automatically record breathing rate and heart rate every 15 minutes is as accurate as the manual measurement. The investigators will also see if these measurements, taken every 15 minutes, will help us predict adverse events more quickly and accurately than the measurements taken every few hours.

Both cardiac arrest and sepsis are primarily identified by vital sign abnormalities. However, the practice of nurses and their designees routinely checking hospitalized patients' vital signs every four to eight hours throughout the day and night has remained essentially unchanged for over one hundred years. While respiratory rate has been shown to be the most predictive vital sign for adverse events on the wards, it is often inaccurately measured and poorly documented. For example, a disproportionate amount of respiratory rates are recorded as either 18 or 20 breaths/min, which is often higher than actual rates. We have previously statistically derived a physiology-based early warning score, called the electronic cardiac arrest risk triage score (eCART), using vital signs and lab values. The eCART was more accurate than scores commonly used in hospitals today. However, the vital sign values utilized for our score were manually collected by nursing staff every four hours. Recent technological advances have allowed for high-frequency measurement of pulse and respiratory rate using a cableless respiration monitor. These devices allow for more frequent and potentially more accurate measures of respiration, which may enhance the prediction ability for detecting adverse events on the wards. In addition, the increase in monitoring frequency may result in earlier detection of adverse events, which could translate into further improvements in patient outcomes. A subset of patients may be continuously measured using a telemetry system. The alarms and ECGs from these patients are monitored by staff in a centralized station in the hospital. When a clinical event requiring action is observed, a call is made to the unit alerting the clinical staff that action is necessary. We will collect the continuous measurements collected from this system and compare these continuous measurements to the high frequency and manual measurements.

Conditions

Interventions

Eligibility

Locations (1)

Outcomes