Pediatric Sepsis Research

 Pediatric Sepsis Research | CPCE

Pediatric sepsis is a major cause of mortality of children worldwide. It is a clinical syndrome that complicates an already severe infection or illness. Sepsis can be challenging to recognize because the signs and symptoms are similar to common febrile illnesses of childhood. These challenges are increased further due to the biologic complexity of sepsis. Sepsis is usually caused by an overwhelming bacterial infection that triggers abnormal responses to the body’s circulatory, immune, and energetic systems of the body. Sepsis requires early identification and aggressive treatment to improve chances for prompt recovery and survival. Standard treatment for sepsis is prompt antibiotics, fluid resuscitation and additional aggressive therapies for children with critical illness.

Pediatric sepsis research at CPCE seeks to improve our ability to quickly identify children with sepsis using quality improvement strategies involving clinical data contained within electronic health records, as well as the development of novel biomarkers to identify at-risk patients and their sources of infection.

In addition to the need to improve outcomes for children with sepsis, researchers are also challenged to prevent the development of pathogens’ resistance to antibiotic treatment. Thus, it is important that clinicians can accurately assess whether a patient is at high- or low-risk for a bacterial infection, for which antibiotics are indicated, and curb antibiotic overuse.

Epidemiology of Sepsis and Identification Strategies

Quality Improvement Strategies in Pediatric Sepsis

Biomarkers of Pediatric Sepsis

RNA Expression Profiling in Pediatric Sepsis Recognition

Pediatric Sepsis Published Research


Epidemiology of Sepsis and Identification Strategies

There have been changes in pediatric sepsis management over the last decade with expansion of protocol-based care in many centers. In this multi-center project, we determined trends in pediatric sepsis prevalence, resource utilization, and outcomes over the past decade. We also found important differences in the accuracy of two different identification strategies for pediatric sepsis in administrative data.

The research team performed an analysis using the Pediatric Health Information System database and compared prevalence, resource utilization, and outcomes using two different established patient identification strategies that utilize ICD9 codes. We demonstrated that pediatric sepsis prevalence has increased over the past decade, but that resource utilization and mortality have decreased. 

We also performed medical record review at six centers for patient groups using the two identification strategies above and determined the presence or absence of reference to standard severe sepsis using international consensus definitions. We found important differences in these methods depending on the identification strategy that was used.

In our validation study, we have demonstrated that using sepsis-specific ICD9 codes is more accurate in identifying patients with severe sepsis or septic shock than a combination method using ICD9 codes for infection plus codes for organ dysfunction.

What’s Next: We continue to analyze the data to determine the accuracy of diagnosis codes to identify pediatric severe sepsis and septic shock;

See the Published Research

Please contact Fran Balamuth, MD, PhD, CHOP Attending Physician Emergency Medicine, for more information about this line of research.


Quality Improvement (QI) Strategies in Pediatric Sepsis

Improvements in pediatric sepsis care and outcomes have been demonstrated with the use of standard sepsis protocols which involve prompt fluid resuscitation and antibiotic therapy.

This line of research projects aims to improve pediatric sepsis care by improving our ability to identify which children need aggressive care, and also by improving timeliness of therapy once these children are recognized. We are deploying QI analytic methods using statistical process control as well as serial PDSA (plan-do-study-act) cycles to temporarily test ideas for change and assess their impact. Preliminary results are promising:

For improved identification: We have demonstrated that a retrospectively applied alert to our patient electronic health records improved patient identification compared to bedside physician judgment. We have since implemented a vital sign-based electronic alert for sepsis in the emergency department and have demonstrated improved sepsis recognition.

For timely treatment: We have demonstrated that timely antibiotics decreases mortality in children with severe sepsis and septic shock. We have also demonstrated that patients who are treated on our ED sepsis protocol have improved timeliness of IV fluids and antibiotics and improved morbidity outcomes at 48 hours as measured by organ dysfunction.

See the Published Research

Funding: National Heart, Lung, Blood Institute, CHOP Departments of Emergency Medicine and Critical Care Medicine, Penn and CHOP internal career development awards

Please contact Fran Balamuth MD, PhD, CHOP Attending Physician, Emergency Medicine and Assistant Professor of Pediatrics, for more information about this line of research.


Biomarkers of Pediatric Sepsis

Overuse of antibiotics contributes to the rise of multi-drug resistant organisms. While approximately 50 percent of inpatient antibiotic use is inappropriate, efforts to curb antibiotic overuse have not been successful, largely due to the inability to identify patients with low likelihood of bacterial infection. Biomarkers have been proposed as a tool to inform decisions to stop antibiotics in patients for which they are not needed.

Our research focuses on optimizing antibiotic use in ICU patients with signs of pediatric sepsis by using biomarkers to differentiate patients with versus without invasive bacterial infection. In other words, we seek to reduce unnecessary antibiotic use in patients with signs of infection.

This research project is using biomarkers to derive a negative predictive model to identify ICU patients with signs of sepsis but at low risk for invasive bacterial infection. We are doing this in several phases:

Phase I: We enrolled ICU patients with signs of infection and started on antibiotics. We collected blood samples at baseline, 24, 48 and 72 hours to measure biomarker values. We were able to identify which patients had versus did not have a bacterial infection, as well as use differences in biomarker values to develop the negative predictive biomarker model. Based on Phase I data, the biomarker results will be presented at 24 hours after infectious signs first appear.

Phase II: We need to determine whether clinicians will use the biomarker model to augment decision making surrounding discontinuation of antibiotics for patients with low likelihood of bacterial infection. In this phase we will provide education to clinicians about the use of biomarkers to identify infection and present biomarker results to clinicians treating ICU patients with signs of infection placed on new antibiotic therapy and observe clinician decision-making patterns.

What’s  Next: We will next study the potential impact and cost-effectiveness of biomarker-based algorithms on reducing unnecessary antibiotic use in critically ill patients.

Funding: Centers for Disease Control and Prevention’s Prevention Epicenter Program

Please contact Susan Coffin, MD, MPH, CHOP Attending Physician and Clinical Director of Infectious Diseases, for more information about this line of research.


RNA Expression Profiling in Pediatric Sepsis Recognition

This line of research aims to use RNA expression profiling, a technique which allows evaluation of the expression of many genes, to evaluate both illness severity and source pathogen in children with suspected sepsis.

To do this, we need to develop novel biomarkers to assist physicians in identifying which pediatric patients with suspected sepsis are at risk of developing critical illness in the first two days of hospitalization, as well as in identifying which children have a bacterial source of infection.

We are collecting blood specimens from children in the emergency department who are treated on the CHOP sepsis protocol and enrolled in the the study. RNA is extracted from the blood and RNA expression profiles are determined.  These results will be evaluated in the context of rich epidemiologic data such as vital signs and laboratory results. Enrollment is currently underway.

Funding: National Heart, Lung, Blood Institute, National Institute of Child Health and Human Development, University of Pennsylvania

Please contact Fran Balamuth, MD, PhD, CHOP Attending Physician, Emergency Medicine and Assistant Professor of Pediatrics, for more information about this line of research.


Pediatric Sepsis Published Research

Epidemiology of Sepsis and Identification Strategies

Balamuth F, Weiss SL, Neuman MI, Scott H, Brady PW, Paul R, Ferris RWD,McClead R, Hayes, K, Gaieski D, Hall M, Shah SS, and Alpern ER. Pediatric Severe Sepsis in US Children's Hospitals. Pediatric Critical Care Medicine. 2014 Nov;15(9):798-805.

Quality Improvement Strategies in Pediatric Sepsis

Weiss SL, Fitzgerald JC, Balamuth F, Alpern ER, Lavelle J, Chilutti M, Grundmeier RW,  Nadkarni V, Thomas N. Delayed Antimicrobial Therapy Increases Mortality and Organ Dysfunction Duration in Pediatric Sepsis. Critical Care Medicine. 2014 Nov;42(11):2409-2417.

Balamuth F, Elizabeth R. Alpern, Grundmeier RW, Chilutti M, Weiss SL, Fitzgerald JC, Hayes K, Bilker W, Lautenbach E. Comparison of Two Sepsis Recognition Methods in a Pediatric Emergency Department. Acad Emerg Med. 2015 Nov;22(11):1298-306.