Monitoring Respiratory Rate in Neonates Using the Rrate Mobile App

Catherine Njeru, Mark Ansermino, Dustin Dunsmuir, Jesse Coleman, Amy Ginsburg, William Macharia

Research output: Contribution to conferencePaper


Introduction: Monitoring the respiratory rate (RR) is an important part of the clinicalassessment of neonates.1 However, accurate RR measurement in clinical settings has beenelusive. RR measurement is especially challenging in neonates because of their irregular andperiodic breathing. There is no reference standard for RR measurement, and proposed methodslike visual counting and the Acute Respiratory Infection timer do not yield readily reproducibleresults.2 Capnography, though not the gold standard, attempts to give a reflection ofphysiological breathing by measuring expired carbon dioxide. There remains a need for a low-cost, simple and accurate tool to monitor RR in neonates. We undertook a study to evaluate theagreement between the RRate3 mobile app timer and Masimo Rad97 capnography for RRmeasurement in neonates.Methods: The study was conducted in the neonatal unit of Aga Khan University Hospital,Nairobi, where following informed consent, eligible neonates were enrolled. Data collectedincluded gestational and current age, sex, diagnosis, anthropometric measurements, and socio-demographic details of the mother. Paired observations were made by 3 trained observersusing the RRate mobile app, counting each neonate’s RR over a full minute. Each neonate wasalso simultaneously connected to a Masimo Rad97 monitor and the capnography waveformcontinuously recorded. The capnography wave forms were digitized and recorded with a customsoftware application. These were then printed out and the breaths manually counted. All datawere entered into a Microsoft Excel (Microsoft Excel, Washington, USA) spreadsheet. Bland-Altman analysis5 for replicated measurements was used to calculate bias and limits ofagreements between the average of the paired RRate observations and the manual counts fromthe capnography waveforms. The root mean square deviation was also calculated.Results: Between June and August 2019, 27 neonates were enrolled into the study. A total of130 paired observations were done but 7 were excluded from the final analysis: 5 were missinga paired RRate reading and 2 were identified as outliers by the interquartile range method.4 123paired observations were analysed and a Bland Altman plot generated (Figure 1). The biasbetween the RRate measurements and the capnography breath counts was 1.88 (95% CI -1.17,2.59) breaths per minute with limits of agreement of -9.75 (95% CI -8.53, -10.97) to 5.99 (95%CI 7.21, 4.77)breaths per minute. The root-mean-square deviation (RMSD) was ±4.4 (9.3%)breaths per minute.Discussion: There appears to be good agreement (< 10% RMSD) between the RRate mobileapp breath counts and Masimo Rad97 capnography. A few extreme outliers were observed onthe Bland Altman plot where the RRate counts were undercounted, especially at higher rates. Alarger study is needed to confirm these findings before the RRate Mobile App could be adaptedas a clinical tool to measure RR in neonates

Original languageUndefined/Unknown
Publication statusPublished - 1 Jan 2021

Cite this