UCSF

Continuous monitoring of respiratory rate (RR), especially in the ICU, is important to identify patients with respiratory compromise. Currently, there are two methods used to assess RR, visual assessment (VA) and impedance pneumography (IP). While VA is easy to use, RR assessments are intermittent, time-consuming, and often inaccurate. The IP method, while continuous, is plagued with alarms, which can lead to alarm fatigue in clinicians. Electrocardiographic (ECG) derived RR methods, or EDR, have been examined but have not been validated for use in the hospital setting.This dissertation was designed to examine RR assessment using the three methods identified above. Three overall aims were studied, including: (1) a systematic review of the literature describing the strengths and limitations of the three methods (VA, IP, and EDR); (2) an evaluation of the number and types of RR alarms (i.e., parameter and apnea) in 461 ICU patients and the association of RR alarms to demographic, clinical characteristics, and supportive therapies; and (3) an evaluation of RR agreement between the IP method and a novel combined-ECG derive (combined-EDR) method in 100 ICU patients. Chapter #2: Of the 78 studies identified in the systematic review, full manuscripts for 23 studies were reviewed, and four studies were included in this review. Given the paucity of research and the fact that no studies have compared all three methods in the same patients, no definitive conclusions can be drawn about the accuracy of these three methods. Chapter #3: RR parameter alarms (high >30 breaths per minute [bpm] or low <5 bpm) and apnea (>20 seconds of no breathing) were examined in 461 ICU patients. These parameters were selected because it was the hospital’s standard default parameter alarm limits in all the ICUs at our hospital. A total of 159,771 RR type alarms over 48,000 hours of monitoring occurred (67 RR alarms/bed/day). The majority of the alarms (82.5%; n=131,827) were high parameter alarms. RR alarm occurrence rates were associated with: the type of ICU (p<0.01), mechanical ventilation (p<0.01); and the lack of a ventricular assist device, or pacemaker (p<0.01). Male gender was associated with low (p<0.01) and apnea (p<0.05) alarms. Chapter #4: This study was designed to examine the agreement between the IP and combined-EDR method for normal RR; low RR (<5 breaths per minute (bpm); and high RR (>30 bpm) in 100 ICU patients. For normal RR, a significant bias difference -1.00 + 2.11 (95% CI -1.60 to -0.40) and LOA of -5.13 to 3.13 was found between the two methods. For low RR, a significant bias difference of -16.54 + 6.02 (95% CI: -18.25 to -14.83) and a 95% LOA of -28.33 to - 4.75 were found. For high RR, a significant bias difference of 17.94 + 12.01 (95% CI: 14.53 to 21.35) and 95% LOA of -5.60 to 41.48 were found. The combined-EDR method had good agreement with the IP method for measuring normal breathing. Whereas the combined-EDR method was consistently higher than low IP RR and almost always lower than high IP RR. This study should be replicated in a larger sample and include confirmation with VA. The overall findings of this dissertation research show that there are very few studies that have examined the three RR methods. For the IP method, high parameter RR alarms are the most common type of alarm. Occurrence rates were associated with the type of ICU, mechanical ventilation, and the lack of a ventricular assist device, or pacemaker. Male gender was associated with low parameter and apnea (p<0.05) alarms. These data suggest that the combined-EDR method is comparable to the IP method with regards to normal RR, was consistently higher than low IP RR, and almost always lower when comparing high RR. This dissertation adds to scientific knowledge regarding RR alarms using the IP method. However, further research is needed to test the combined-EDR method to the gold standard VA method to determine the accuracy of this method.