Impact of Rapid Response Teams on Pediatric Care: An Interrupted Time Series Analysis of Unplanned PICU Admissions and Cardiac Arrests

Abstract

Pediatric rapid response teams (RRTs) are expected to significantly lower pediatric mortality in healthcare settings. This study evaluates RRTs' effectiveness in decreasing cardiac arrests and unexpected Pediatric Intensive Care Unit (PICU) admissions. A quasi-experimental study (2014-2017) at King Abdulaziz University Hospital, Jeddah, Saudi Arabia, involved 3261 pediatric inpatients, split into pre-intervention (1604) and post-intervention (1657) groups. Baseline pediatric warning scores and monthly data on admissions, transfers, arrests, and mortality were analyzed pre- and post-intervention. Statistical methods including bootstrapping, segmented regression, and a Zero-Inflation Poisson model were employed to ensure a comprehensive evaluation of the intervention's impact. RRT was activated 471 times, primarily for respiratory distress (29.30%), sepsis (22.30%), clinical anxiety (13.80%), and hematological abnormalities (6.7%). Family concerns triggered 0.1% of activations. Post-RRT implementation, unplanned PICU admissions significantly reduced (RR = 0.552, 95% CI 0.485-0.628, p < 0.0001), and non-ICU cardiac arrests were eliminated (RR = 0). Patient care improvement was notable, with a -9.61 coefficient for PICU admissions (95% CI: -12.65 to -6.57, p < 0.001) and a -1.641 coefficient for non-ICU cardiac arrests (95% CI: -2.22 to -1.06, p < 0.001). Sensitivity analysis showed mixed results for PICU admissions, while zero-inflation Poisson analysis confirmed a reduction in non-ICU arrests. The deployment of pediatric RRTs is associated with fewer unexpected PICU admissions and non-ICU cardiopulmonary arrests, indicating improved PICU management. Further research using robust scientific methods is necessary to conclusively determine RRTs' clinical benefits.

Keywords: before-and-after study; cardiac arrest; intensive care units; patient care team; pediatrics; time series studies.

Figure 1

This figure illustrates the results of the Interrupted Time Series (ITS) analysis focusing on the impact of the pediatric rapid response team intervention on unplanned Pediatric Intensive Care Unit (PICU) admissions. The plot displays the observed un-planned PICU admission rates over time alongside the fitted values predicted by the ITS model. A vertical grey dashed line indicates the intervention’s implementation in September 2015, demarcating the pre- and post-intervention phases. The trend lines provide a visual representation of unplanned PICU admissions, demonstrating a notable decrease following the introduction of the RRTs intervention.

Figure 2

An Interrupted Time Series (ITS) analysis displays the monthly counts of non-ICU cardiorespiratory arrests in pediatric patients from January 2014 to July 2017, with observed data and fitted model values illustrated in distinct line styles. The implementation of the pediatric rapid response team in September 2015 is denoted by a grey dashed line, serving as a demarcation for comparing trends and assessing the intervention’s impact in the pre- and post-intervention periods.

Figure 3

Sensitivity test by bootstrapped coefficient distribution for unplanned PICU admissions. This histogram depicts the distribution from 1000 bootstrapped samples analyzing the impact of a healthcare intervention on unplanned PICU admissions. The original mean difference of −9.61 (red dashed line) suggests a reduction in admissions post-intervention. The 95% confidence interval, ranging from −3.65 to 4.04 (green dashed lines), indicates substantial variability in the intervention’s effect.