Keynote Speaker presentation at the Third Annual Barsan Emergency Medicine Research Forum

1. Story to Sepsis Screen
A Journey in the Details
Elizabeth R. Alpern, MD, MSCE
Professor of Pediatrics
Ann & Robert H. Lurie
Children’s Hospital of Chicago
Division of Emergency Medicine

2. Disclosures
• I have no relevant financial relationships with the
manufacturer(s) of any commercial product(s) or provider(s)
of commercial services discussed in this CME activity
• I do not intend to discuss an unapproved/investigative use of
a commercial product/device in my presentation

3. Project work supported by:
AHRQ R01HS020270
NICHD R01HD087363
PECARN infrastructure support by:
Health Resources and Services Administration
(HRSA), Maternal and Child Health Bureau (MCHB),
Emergency Medical Services for Children (EMSC)
through the following grants: U03MC00008,
U03MC00003, U03MC22684, U03MC00007,
U03MC00001, U03MC22685, U03MC00006

4. • Emergency care for children is variable with
significant opportunities for improvement.
• Basic administrative data are not adequate for
reporting and improving quality of care.
• Advances in health information technology to
access patient-centric clinical data (natural
language processing and penetrance of EHR)
provide opportunity.
Rationale

5. Aims:
•Develop an emergency care visit registry for
pediatric patients from EHR.
•Collect and determine benchmarks for stakeholder-
prioritized emergency care performance at
Emergency Department (ED) and clinician level.
•Report performance to individual ED clinicians and
sites while evaluating change using a staggered time-
series study.
PECARN Registry

6. The Pediatric Emergency Care Applied
Research Network (PECARN)
www.pecarn.org

7. 60 balanced, stakeholder-endorsed quality
performance measures
•https://emscimprovement.center/resources/toolboxes/emergency
Performance Measures

8. PECARN Registry

9. Patient
Electronic Health
Record
All ED Visits
from sites
Monthly Data
Transmission
validation
De-identification
Data
Coordinating
Center
Natural Language
Processing (NLP)
Performance
Measures
Benchmarks
Site specific
Provider specific
Timely
PECARN
Registry
Improved
Patient
Care

10. • Site
• Patient identifiers:
►Patient number, encounter
number
• Demographics
►Date of birth (DOB), sex,
race, ethnicity, zip, payer
• Visit information
►Triage category, chief
complaint, arrival mode
►Date/Time: notification, ED
door, sort/triage, discharge
• Providers
► Provider ID, provider role,
provider D/T
• Vitals
► Vitals D/T
► T, HR, RR, SBP, DBP
► oxygen saturation
► weight
• Medications
► Current; ED (D/T); discharge
Variables

11. • Clinical assessments
►Asthma score, pain score,
Glasgow Coma Scale
(GCS)
• Narrative
►Narrative D/T, author
type, narrative
• Radiology
►Order D/T, start D/T,
avail D/T, report D/T,
report
• Labs (including Micro)
► Lab D/T result
• Procedures
► CPT, ICD9, ICD10
• Diagnosis
► ICD9, e-codes, ICD10
• Disposition
► ED disposition
► Hospital discharge D/T
► Vital status
Variables

12. Children’s Hospital Colorado (main & satellite)
Children’s Hospital of Philadelphia
Children’s National Medical Center (main & satellite)
Cincinnati Children’s Hospital (main & satellite)
Lurie Children’s Hospital of Chicago
Milwaukee Children’s Hospital
Nationwide Children’s Hospital
University of Michigan
University of California - Davis
Since 2012
Since 2016
Since 2018
PECARN Registry

13. Scope of the PECARN Registry
PECARN Registry
2012-2016
N
Encounters 2,250,444
Patients 911,239
Diagnoses 6,619,901
Lab Results 13,920,477
Medication Orders 2,648,548
Radiology Tests 800,730
Narrative Documents 13,707,590
Providers receiving report card 1023

14. QI TENETS Upheld
Measurement of outcomes
•Defined; effectively & efficiently
measured
Transparency of data
•Rigorously determined & validated
Iterative data
•Informs processes of care
•Allows systems to adjust

15. Performance Measures Derived from PECARN
Registry
Expert Panel determination of
Ideal Benchmark
Report Cards

16. • Visits attributable to provider (or site)
• Not identifiable to anyone but provider
• Stringent case identification for cohorts
• Number of cases involved in the measure provided
►Monthly or rolling quarter count
• Graphic representation of performance
• Trends over time
• Comparisons of performance for:
►Site (proportion or median)
►Network sites together
►Achievable Benchmark of Care (ABC)
Report Cards

17. • Measurable level of excellence
• Objective, reproducible, and predetermined
• Providers with high performance define
achievable level of excellence
• Providers with a small number of relevant
visits will not have high influence on
benchmark
Kiefe CI, et al., 1998, 2001
Achievable Benchmark of Care

18. Provider Report Card
Provider
performance
Site
performanceOverall network
performance
Achievable benchmarks
of care Variability of
providers within site
Trend of single
provider over time
Trends of site and
network over time

19. Site Report Card

20. Illustrative Case:
Antibiotic Use in Viral Illness
Goyal et al, Pediatrics 2017

24. Use of Antibiotics in Viral Illnesses
N = 39,445
Overall Antibiotic Use 2.6%
Main ED 2.6% Satellite ED 2.6%
PEM 2.8%
Pediatrician 2.8%
APN 2.1%

31. Antibiotics by Race/Ethnicity
*Adjusted for age, insurance status, acuity level, ED site, ED type, provider type
Referent=White, NH

32. Antibiotics by Race/Ethnicity
*Adjusted for age, insurance status, acuity level, ED site, ED type, provider type
Referent=White, NH

33. ED antibiotic use for acute ARTI
• Overall ED antibiotic provision for viral ARTI
low
• Differences in antibiotic provision by
patient race/ethnicity exist
– NH-whites more likely to receive unnecessary
antibiotics than minority patients
– Differences persisted after adjustment for confounding

34. • Can we use the rich clinical
data of the EHR to evaluate
diagnostics, therapeutics,
and outcomes?
• Can we improve care?
• Can we expand to all components of
care (pre-hospital, ED, inpatient)?

36. What is sepsis?
S
E
P
S
I
S

37. Sepsis: A continuum of illness

38. • Sepsis is life-threatening organ dysfunction caused by a
dysregulated host response to infection
• Septic shock is subset with circulatory (hypotension) &
cellular/metabolic abnormalities (lactate >2 mmol/L)
JAMA Feb 2016

39. ~4% of total pediatric
hospitalizations
PCCM 2014

40. Mortality 16.9%
Hospital LOS 16 days
ICU LOS 5 days
Hospital Costs / Pt $65K

41. Increased mortality with
> 3 hr delay to antibiotics
OR 3.92 (1.27, 12.06)
CCM 2014

42. Pediatric patients with severe sepsis treated according to a
protocolized care guideline and order set in the ED
•Improved Resolution of OD at day 2 OR 4.2 (1.7,10.4)
PCCM 2016

43. Improving Pediatric Sepsis Outcomes (IPSO)
• Reduce sepsis
mortality by 75%
• Reduce hospital-
onset severe
sepsis by 75%
Addresses all stages of sepsis across the care continuum:
emergency department, intensive care, general care,
hematology/oncology and bone marrow transplant units
initially, and NICU, pre-hospital, and ambulatory settings.
https://www.childrenshospitals.org/Programs-and-
Services/Quality-Improvement-and-

44. • 16 year old in triage with the
following vital signs:
– Temp: 39.8
– HR: 140
– RR: 38
– BP: 100/67
– Pox: 97% RA
• 3 yo patient in triage with the
following vital signs:
– Temp: 39.8
– HR: 140
– RR: 38
– BP: 100/67
– Pox: 97% RA
Challenges in sepsis recognition

45. Sepsis Recognition: A fine balance

46. Annals EM 2016

47. MissedCasesofPICUSepsis
Electronic
Sepsis Alert
+
Clinician ID
Annals EM 2016

48. PED Screen
Overall Goal: Improve the detection and
treatment of pediatric sepsis in the ED setting
–Expanded multicenter EHR registry
–Automate determination of organ dysfunction
directly from multicenter EHR registry
–Derive and validate prediction model of
pediatric sepsis that predicts organ
dysfunction within 48 hours using ED EHR
data from the first 4 hours of care

49. ealpern@luriechildrens.org