Document resource
Objectives To explore the relationships between the pulse/breathing rate ratio (PBR) and the risk of death and critical illness.Design This was a retrospective observational study.Setting This was a single-centre study from a tertiary hospital in Tianjin, China.Participants The study population consisted of patients aged ≥16 years who underwent consultation and were monitored for medical purposes at the clinic.Interventions Between April 2021 and December 2021. Before the patients received any medical intervention, vital signs were measured, and the PBR and the National Early Warning Score (NEWS) were calculated on the basis of the above measured indicators.Primary and secondary outcome measures The associations of PBRs with death and critical illness were evaluated.Results A total of 1048 outpatients with fever were included. Restricted cubic spline (RCS) bars were used to explore potential nonlinear associations between PBR and mortality and critical illness. The abilities of the PBR and NEWS to predict the risk of death and critical illness were compared through decision curve analysis (DCA). The RCS showed a U-shaped nonlinear distribution of associations between PBRs and death and critical illness (nonlinear p values of p=0.036 and p=0.005, respectively). The risk of mortality was lowest between 4.6 and 6.2 for PBR, with the risk of mortality increasing progressively with decreasing PBR for PBR<4.6, and the risk of mortality increasing progressively with increasing PBR for PBR>6.2. The risk of critical illness was lowest when the PBR was between 4.6 and 5.5, and the risk increased gradually with decreasing PBR for PBR<4.6 and with increasing PBR for PBR>5.5. The DCA results revealed that the value of the PBR in predicting death was similar to that of the NEWS. In the DCA, the net benefit achieved by the NEWS ranged from 3% to 10% threshold probability, and the net benefit achieved by the PBR ranged from 4% to 10% threshold probability, with similar net benefits for the NEWS and PBR and 1% for both the PBR and the NEWS when the threshold probability was 7% (equivalent to correctly identifying 10 mortalities per 100 patients).Conclusions The PBR helps to predict the risk of mortality and critical illness in acutely ill patients, and its value in predicting mortality is similar to that of the NEWS.