Abstract
Among the vital signs collected during hospital triage, respiratory rate is an important parameter associated with physiological, pathophysiological, and emotional changes. In recent years, the importance of its verification in emergency centers due to the severe acute respiratory syndrome 2 (SARS2) pandemic has become very clear, although it is still one of the least evaluated and collected vital signs. In this context, infrared imaging has been shown to be a reliable estimator of respiratory rate, with the advantage of not requiring physical contact with patients. The objective of this study was to evaluate the potential of analyzing a sequence of thermal images as an estimator of respiratory rate in the clinical routine of an emergency room. We used an infrared thermal camera (T540, Flir Systems) to obtain the respiratory rate data of 136 patients, based on nostrils' temperature fluctuation, during the peak of the COVID-19 pandemic in Brazil and compared it with the chest incursion count method, commonly employed in the emergency screening procedures. We found a good agreement between both methods, with Bland-Altman limits of agreement ranging from -4 to 4 min(-1), no proportional bias (R(2) = 0.021, p = 0.095), and a strong correlation between them (r = 0.95, p < 0.001). Our results suggest that infrared thermography has potential to be a good estimator of respiratory rate in the routine of an emergency room.