A comparison of pulse oximetry and respiratory rate in patient screening

Objective: To examine how well respiratory rate correlates with arterial oxygen saturation status as measured by pulse oximetry, and determine whether respiratory rate measurements detect oxygen desaturation reliably. Methods: Respiratory rate (RR) and oxygen saturation (SaO(2)) were measured prospectively on 12 096 consecutive adult emergency department triage patients at a university medical center. Respiratory rate was measured by counting ausculated breath sounds for 1 min. Pulse oximetry was used to measure SaO(2). Measurements were analysed by age (with one group for 18-19 year olds, groups for every 10 yr from age 20 to age 60, and groups for every 5 yr for subsequent ages). Pearson correlation coefficients were calculated for each age group as well as the weighted average coefficient. Cases having oxygen saturation below 90% were examined to determine how frequently they exhibited increased RR (increased RRs were defined as any rate in the upper five percentile by age. Results: Correlation coefficients ranged from 0 . 379 to -0 . 465 with a weighted mean of -0 . 160. Coefficients for ages IS through 70 years (representing 10 740 patients) all had magnitude <0 . 252. Overall, only 33% of subjects with oxygen saturation below 90% exhibited increased RR. Conclusions: Respiratory rate measurements correlate poorly with oxygen saturation measurements and do not screen reliably for desaturation. Patients with low SaO(2) do not usually exhibit increased RR. Similarly, increased RR is unlikely to reflect desaturation.