SCCM Resource Library-Children's Hospitals Are Associated With Improved Outcomes in PICU Patients With Sepsis

Video Transcription

Good morning. My name is Ashish Nagpal. I'm a pediatric intensivist at Oklahoma Children's Hospital, and I'm going to be discussing our  study in which we evaluated the impact of initial location of resuscitation on clinical outcomes in PICU patients with sepsis.  Based on the Point Prevalence Sprout Study, sepsis is a major cause of morbidity and mortality worldwide, with mortality approaching 25%.  A recent study using a nationwide database in the United States evaluated the clinical outcomes of children with sepsis comparing patients who were transferred to those who  remained at an outlying facility. And what they found was that children who were transferred were younger, sicker, had higher organ dysfunction, and increased mortality.  However, in adjusted analysis, no mortality difference remained. And so the impact of initial location of resuscitation on clinical outcomes in children with sepsis remains unclear.  And so in our study, we compared in children along the spectrum of sepsis who were admitted  to PICUs and received their initial resuscitation at a community-based hospital were compared  to patients who received their initial resuscitation at a virtual pediatric systems or VPS participating hospital.  And our hypothesis was that among critically ill children along the spectrum of sepsis who received their initial resuscitation at a VPS participating hospital will have improved  clinical outcomes compared to those who received their initial resuscitation and management at a non-VPS participating hospital.  VPS is a prospectively constructed, highly quality controlled database that includes PICU patients from over 200 PICUs.  And we used it to conduct a retrospective cohort study in patients with a primary diagnosis of sepsis, severe sepsis, or septic shock over a 10-year period.  We included patients who had a primary diagnosis along the spectrum of sepsis who had their  diagnosis within 24 hours of admission, were 30 days to 18 years of age, only from US-based PICUs, and had to come from an emergency department.  No restrictions were made on the type of PICU and unit and staff characteristics, for example. Our primary outcome was mortality, and our secondary outcomes were intensive care unit  length of stay, ventilator free days, and hospital length of stay. Our statistics were descriptive data were presented as medians with interquartile ranges  for continuous variables and frequencies with percentages for categorical variables.  Multiple comparisons of descriptive and outcome data were made in the usual fashion for continuous and categorical variables.  We used logistic regression to evaluate mortality, and linear regression to evaluate our continuous  outcome variables in ICU length of stay, hospital length of stay, and ventilator free days, and statistical significance was defined as a P-value less than 0.05.  Looking at our results, the majority of patients in our cohort had a diagnosis of sepsis with or without shock, with septic shock being the second most common diagnosis.  We started with just over 64,000 patients in our data set, and after applying our inclusion criteria with sepsis star codes, sepsis onset less than one day, presenting to emergency  department within our age group, and removing two patients who had a length of stay less than 0 days, we had just over 9,800 patients for evaluation of mortality, ICU length of  stay, and hospital length of stay. And after evaluating patients with mechanical ventilation records, whose mechanical ventilation  started within 48 hours of their sepsis diagnosis, we had just over 5,500 patients for the evaluation of mechanical ventilation.  Looking at demographic and outcome data in our cohort, using simple comparisons and unadjusted  analysis, our cohort included just over 9,800 patients, 2,500 of whom came from a non-VPS emergency department. There are no differences between the two groups in gender or weight.  Our cohort had a higher percentage of patients at the extremes of age, and were more likely to be Caucasian, presenting from a non-VPS emergency department.  Among patients coming from a non-VPSED, the majority of patients were transported via ground.  However, transportation had a high missingness rate, with no transportation information reported in over 4,000 patients, 914 of whom came from a non-VPSED.  And this represented a 36% missingness rate. In simple comparisons looking at our outcome data, severity of illness and probability  of death based on PIM-3 scoring was higher, presenting to a non-VPSED. Presenting to a non-VPSED, you are more likely to require mechanical ventilation.  However, there were no differences between the two groups in the number of ventilator-free days.  Hospital length of stay and IC length of stay were one day shorter, presenting to a VPS emergency department. And overall mortality in our cohort was 6.3%.  9.2% presenting to a non-VPSED versus 5.2% presenting to a VPSED. And this was statistically significant.  Presented in this table are the univariate analysis of cofactors and their association with mortality.  After controlling for age, race, severity of illness, presence of mechanical ventilation, and a chronic medical condition, presenting to a VPS emergency department was independently  associated with lower odds of mortality in multivariable logistic regression. When we evaluated our secondary outcomes in multivariable linear regressions, IC length  of stay, hospital length of stay, and ventilator-free days were not independently associated with patient origin after controlling for covariates identified in univariate analysis.  We evaluated the need for organ support as a surrogate for organ dysfunction due to inadequate reporting of PILOD scoring in our cohort.  In simple comparisons, a higher percentage of children presenting to a non-VPSED required CPR or extracorporeal life support.  However, there was no difference between the two groups in those requiring renal replacement therapy.  There are plausible explanations for the differences in mortality seen in our study. First, non-VPS subjects more likely required CPR and ECMO, but not renal replacement therapy  and unadjusted analysis. And this remains a plausible explanation for the difference we saw in our study.  PILOD scoring was reported infrequently as a non-mandatory data entry point in our data set, and we could not include it as a confounder.  And it's possible that the increased mortality seen in our study was due to the presence of multi-organ dysfunction among transferred patients.  Bundled care for the management of severe sepsis and septic shock is becoming the standard of care in how we streamline the evaluation and management of these patients, and we know  this improves care. However, this infrastructure may not be present at non-pediatric-specific hospitals.  In our study, we found that patients who traveled by air were sicker, which makes sense. However, there was not enough information provided around the decision to which a particular  patient received a mode of transport, whether that was weather, availability of immediate transport, fixed wing versus helicopter, or pediatric-specific transports.  And we could not evaluate it as a confounder in our study. When arriving from outlying facilities, there may be a lack of access to pediatric subspecialists,  providers who have comfort for caring for pediatric patients and individuals who are PALS trained. Our study has several strengths.  We evaluated patients along the entire spectrum of sepsis in large numbers using a robust,  quality-controlled data set and applied strict inclusion criteria to avoid including patients  whose diagnosis came much later in their hospital course, thus evaluating the role of resuscitation  at community-based hospitals and, to our knowledge, represents the largest study to evaluate the  impact of initial location of resuscitation on clinical outcomes along the spectrum of pediatric sepsis in multiple PICUs across the United States.  Our study has several limitations. First, this was a retrospective study. Second, we evaluated for potential causes of increased mortality in our study in evaluating  need for organ support systems like CRRT, ECMO, and CPR as a surrogate for organ dysfunction.  However, due to model stability and low numbers, we could not include it in our robust regression analysis.  As previously mentioned, we could not include transportation as a confounder as well. Our study lacked the clinical granularity to evaluate the mortality difference seen  in our study, such as type of volume, amount of volume, vasoactive inotrope scores, and nutritional status, to name a few.  The clinical guidelines for the management of septic shock changed in 2017, and it's unclear to what degree this affected the magnitude of outcome seen in our study.  We did not control for hospital-acquired infections that may prolong ICU or hospital length of It's becoming increasingly common to use non-invasive support modes of ventilation.  However, in our data set, there were insufficient numbers where this was recorded to reasonably evaluate this in our study.  We also had insufficient numbers to evaluate functional outcomes, and we included the entire spectrum of sepsis in our study.  We did this to maximize the capture of patients and to account for institutional differences in how sepsis may be documented.  The fact that we found a difference despite this remained significant. In conclusion, in our retrospective cohort study in consecutive PICU patients, presentation  to a VPSCD was associated with improved survival. However, patient origin was not associated with hospital, ICU length of stay, and ventilator-free days.  Overall mortality in our study was 6.3%, which is lower than in previously reported studies of severe sepsis and septic shock, and we attribute this to including patients with  sepsis. And severity of illness was higher for non-VPS subjects, suggesting either inadequate resuscitation  efforts or a failure to recognize the severity of illness in these patients. A planned future study is to evaluate the patients we contributed to this data set to  evaluate with more clinical granularity the differences we saw in our current study, and to evaluate the impact of pediatric transport systems on sepsis outcomes in states such  as ours, with large population but low density. And to identify areas geographically where mortality may be occurring more frequently,  and use this to advocate for education and outreach programs. Thank you for your time, and I'd like to thank my mentor, Dr. Allen, and our collaborators in this study.

Video Summary

A study was conducted to evaluate the impact of initial location of resuscitation on clinical outcomes in pediatric intensive care unit (PICU) patients with sepsis. The study compared patients who received their initial resuscitation at a community-based hospital to those who received it at a virtual pediatric systems (VPS) participating hospital. The hypothesis was that patients who received their initial resuscitation at a VPS participating hospital would have improved clinical outcomes. The study found that presenting to a VPS emergency department was independently associated with lower odds of mortality. However, there were no significant differences in ICU length of stay, hospital length of stay, and ventilator-free days. Further study is planned to evaluate the differences seen in the current study.

Asset Subtitle

Sepsis, Worldwide Data, Pediatrics, 2022

Asset Caption

INTRODUCTION: Sepsis, severe sepsis and septic shock are potentially life-threatening conditions. Given that early recognition, timely resuscitation, and appropriate antibiotics are associated with improved outcomes in pediatric sepsis, we investigated the association between sepsis outcomes and location of initial evaluation and resuscitation (Virtual Pediatric System (VPS) participating children’s hospital versus a non-VPS participating hospital). METHODS: Using the VPS database (a quality-controlled registry of PICU patients from over 200 PICU’s worldwide), we performed a 10-year retrospective cohort study of children (aged 30 days to 18 years) with a primary diagnosis of sepsis, severe sepsis or septic shock who were admitted to a VPS participating PICU from either a referring hospital emergency department (non-VPS ED) or a VPS hospital ED in the United States. Subjects were included if a diagnosis of sepsis was made within 24 hours of PICU admission and had a length of stay ≥ 1 day. Mechanical ventilation (MV) duration was assessed using ventilator free days (VFD) for patients who were intubated within 48 hours of PICU admission. RESULTS: Data were obtained on 64,510 subjects; 9,833 subjects were included for analysis of mortality and intensive care unit length of stay (ICU LOS). 5,507 subjects were included in analysis of VFD. 26% (n = 2,562) of subjects were admitted from a non-VPS ED and 74% (n = 7271) were admitted from a VPS ED. Initial presentation to a VPS-ED was associated with improved survival (OR 0.81, p = 0.045, 95% CI 0.66-0.99) while controlling for severity of illness, comorbidities, presence of MV, age and race. Non-VPS ED subjects had a higher probability of death based on PIM scoring [1.2 IQR (0.8, 3.9) vs 1.0 IQR (0.8, 3.5) p = < 0.0001], longer ICU LOS [3 days IQR (1,8) vs 2 days IQR (1,7) p = < 0.0001] and were more likely to require mechanical ventilation (50% vs 35%, p = < 0.0001). However, VFD were not different between the two groups [23.9 IQR (20.2, 26.2) vs 24 IQR (20.2, 26.2) p = >0.9]. CONCLUSIONS: In our retrospective study of pediatric patients with sepsis, severe sepsis and septic shock, initial management at a VPS participating hospital compared to a non-VPS hospital was associated with improved survival and decreased ICU LOS.

Meta Tag

Content Type Presentation

Knowledge Area Sepsis

Knowledge Area Pediatrics

Knowledge Area Worldwide Data

Knowledge Level Intermediate

Knowledge Level Advanced

Learning Pathway Sepsis Resources

Membership Level Select

Tag Sepsis

Tag Epidemiology Outcomes

Tag Pediatrics

Tag Shock

Tag Mortality

Tag Outcomes Research

Tag Length of Stay

Year 2022

Keywords

resuscitation

clinical outcomes

pediatric intensive care unit

mortality

emergency department

Sepsis

Pediatrics

Sepsis

Epidemiology Outcomes

Pediatrics

Shock

Outcomes Research

Length of Stay

Presentation

Select

2022

Worldwide Data

Intermediate

Advanced

Sepsis Resources

	Society of Critical Care Medicine 500 Midway Drive Mount Prospect, IL 60056 USA Phone: +1 847 827-6888 Fax: +1 847 439-7226 Email: support@sccm.org	Contact Us About SCCM Newsroom Advertising & Sponsorship DONATE	MySCCM LearnICU Patients & Families Surviving Sepsis Campaign Critical Care Societies Collaborative	GET OUR NEWSLETTER
© Society of Critical Care Medicine. All rights reserved. \| Privacy Statement \| Terms & Conditions The Society of Critical Care Medicine, SCCM, and Critical Care Congress are registered trademarks of the Society of Critical Care Medicine.