Background: Ideally, rapid response systems identify at-risk inpatients and standardize ICU consultation. Scores and rapid response teams (RRT) are increasingly used to achieve these goals. Knowledge about differential and combined effects of RRT and scores on important outcomes is limited.
Methods: The EPOCH RCT (NCT01260831) evaluated the Bedside Paediatric Early Warning System (BedsidePEWS) versus no severity of illness score in patients <18 years and found similar hospital mortality and reduced late ICU admission. In this planned subgroup analysis we identified study outcomes where the effect of BedsidePEWS on the study outcome was different in hospitals with and without a RRT (interaction p<0.05). Next, the cluster-RCT regression analyses comparing BedsidePEWS versus control were repeated in each RRT sub-group. Results from RRT and no-RRT hospitals are reported as rate ratios (RR) with 95% confidence intervals.
Results: The 12 hospitals with no-RRT had 79985 discharges and 101 deaths. The 9 hospitals with RRT had 64754 discharges and 143 deaths. Interaction p-values suggested significantly differing effects of BedsidePEWS by RRT status for 4 outcomes. Amongst no-RRT hospitals, those with BedsidePEWS had more stat [RR 1.47 (1.32-1.64) p<0.01] and urgent calls to ICU [RR 1.27 (1.09-1.47) p<0.01]; and had less clinical deterioration [RR 0.84 (0.72-0.98) p=0.03] and urgent ICU admissions [RR 0.82 (0.70-0.95) p=0.01] than those without BedsidePEWS. Amongst RRT hospitals, those with BedsidePEWS had fewer stat calls [RR 0.88 (0.77-0.99) p=0.04] and similar urgent RRT calls [p=0.7], clinical deterioration [p=0.4] and urgent ICU admissions [p=0.4] than those without BedsidePEWS. Effects on mortality, late ICU admission, and cardiac arrests did not differ by RRT status.
Conclusions: Reduced clinical deterioration with BedsidePEWS in hospitals without RRT may reflect more efficient use of available expertise to manage sick inpatients. Differing effects of BedsidePEWS implementation in hospitals by RRT status may reflect cultural, resource and other differences that need further study.