Bedside Pediatric Early Warning System Does Not Reduce All-Cause Mortality

March 1, 2018

Article

Hospitals utilizing BedsidePEWS did not see a reduction in mortality, but the rate of late admissions to the ICU was decreased.

Christopher Parshuram, MBChB, DPhil, FRACP

Implementing the Bedside Pediatric Early Warning System (BesidePEWS) is inconsequential in reducing all-cause mortality among hospitalized pediatric patients, but does result in a lowered incidence of late admission to the intensive care unit (ICU), according to research.

The study explored the use of BesidePEWS in 21 hospitals in 7 countries—Belgium, Canada, England, Ireland, Italy, New Zealand, and the Netherlands—from February 28, 2011, to July 19, 2015. It utilized data from 144,539 patients, and 559,443 patient-days.

Investigators found the rate of all-cause mortality to be 1.93 per 1000 discharges at hospitals with BesidePEWS and 1.56 per 1000 discharges with usual care (adjusted group rate difference, 0.01; 95% CI, −0.80 to 0.81 per 1000 patient discharges; adjusted odds ratio [adjOR], 1.01; 95% CI, 0.61 to 1.69; P = .96).

Led by Christopher Parshuram, MBChB, DPhil, FRACP, a staff physician at the Hospital for Sick Children and an associate professor of critical care medicine and pediatrics at the University of Toronto, the team noted that these findings do not support the use of BedsidePEWS to intervene in regard to mortality reduction.

“This failure of the BedsidePEWS intervention to positively reduce mortality rates may be difficult to reconcile with the expectation that outcomes can be improved if early clinical changes are identified,” Neil A. Halpern, MD, MCCM, wrote in an accompanying editorial.

Although BedsidePEWS lacked a decrease in all-cause mortality, the system is still a highly validated and practically oriented system of care that is beneficial for clinical routines, Parshuram told MD Magazine.

Neil A. Halpern, MD, MCCM

“It was chosen as the best in class to compare vs. usual care in the EPOCH trial, the largest randomized trial of hospitalized children to date,” he said. “It studied the implementation of a complex healthcare intervention, that involved changing the core business of hospitals. That hospital leaders—CEOs and other leaders—agreed for their hospitals to participate, is [a] testimony to their desire to help generate this type of evidence to inform [the] delivery of care and may be seen as a template for future studies.”

The composite outcome of late admission to the ICU—defined by the authors as clinical deterioration events—did show a significant reduction for hospitals using BedsidePEWS. In total, there were 386 significant deterioration events (BedsidePEWS hospitals, n = 127; usual care hospitals, n = 259). These events consisted of 15.3% of urgent ICU admissions at BedsidePEWS hospitals and 22.0% at usual care hospitals.

Late admissions to the ICU occurred at 0.5 per 1000 patient-days at hospitals with BedsidePEWS and 0.84 per 1000 patient-days at hospitals with usual care (adjusted between-group rate difference, −0.34; 95% CI, −0.73 to 0.05 per 1000 patient-days; adjRR, 0.77; 95% CI, 0.61 to 0.97; P = .03).

“This is the key objective of Early Warning Systems, and for clinicians and individual patients alike an important objective—avoiding late ICU admission,” Parshuram said.

The reduction in late ICU admission did not translate into a measurable difference in mortality, due to the mortality rate being too low for a significant effect to be found in the mixture of hospitals studied, Parshuram said. In turn, he said, this poses a question about the relevance of seeking lower mortality when rates are close to zero, and not all patients are curable.

Ultimately, Halpern concurred with Parshuram, noting that “it is possible that there were not enough deaths overall to show a difference. Regardless, some have suggested that mortality as a primary outcome is limited,and perhaps mortality would be a better outcome measure if combined with other parameters that may be improved by various interventions.”

The study, "Effect of a Pediatric Early Warning System on All-Cause Mortality in Hospitalized Pediatric PatientsThe EPOCH Randomized Clinical Trial," was published in JAMA.