1601. Impact of Personalized Audit and Feedback on Management of Pediatric first-line therapy for most children.

for most children. The use of penicillins increased at children’s hos-

Outpatient Community-Acquired Pneumonia pitals after guideline publication, but trends in antibiotic choice for CAP at general
Lori Handy, MD, MSCE1,2; Adriana Cadilla, MD3; Lloyd Werk, MD, MPH4; Maria hospitals have not been evaluated.
Carmen Diaz, MD5; James Franciosi, MD, MSCE6; Joanne Dent, RN, MS, PMP7; Methods.  Retrospective analysis of children 1–17 years admitted from January 2009
Jobayer Hossain, PhD7; James Crutchfield Jr., PhD8 and Timothy Wysocki, PhD 9; through September 2015 to 522 hospitals, captured via the Pediatric Health Information
1
Division of Infectious Diseases, Children’s Hospital of Philadelphia, Philadelphia, System and Premier Perspective databases. Children with CAP were identified by a val-
Pennsylvania, 2Division of Infectious Diseases, Nemours/A.I. duPont Hospital for idated ICD-9 code algorithm, excluding those with complicated pneumonia, complex
Children, Wilmington, Delaware, 3Division of Infectious Diseases, Nemours Children’s chronic conditions, receipt of intensive care, or MRSA infection or colonization. Receipt of
Hospital, Orlando, Florida, 4Department of Pediatrics, Nemours Children’s Hospital, penicillins, cephalosporins, and macrolides was assessed, and trends were modeled using
Orlando, Florida, 5Division of Emergency Medicine, Nemours/Alfred I. duPont segmented logistic regression, adjusting for age, sex, and insurance provider. Standardized
Hospital for Children, Wilmington, Delaware, 6Division of Gastroenterology, Nemours probability of antibiotic receipt was compared between children’s and general hospitals.
Children’s Hospital, Orlando, Florida, 7Nemours/Alfred I. duPont Hospital for Results.  Of 120,238 children hospitalized with CAP, 54% were admitted to
Children, Wilmington, Delaware, 8Lockheed Martin Rotary and Mission Systems, 51 children’s hospitals. After adjustment, penicillin use increased and both ceph-
Orlando, Florida, 9Nemours Children’s Health System, Jacksonville, Florida alosporin and macrolide use decreased in both children’s and general hospitals
(Figure). However, in the final study year, children in general hospitals were less
Session: 169. Stewardship: Pediatric Antimicrobial Stewardship
likely to receive penicillins (standardized probability 0.23, 95% CI [0.17, 0.29] vs.
Friday, October 6, 2017: 12:30 PM 0.57 [0.52, 0.62]) and more likely to receive cephalosporins (0.78 [0.73, 0.82] vs. 0.51
Background.  Community-acquired pneumonia (CAP) is a common infection in [0.45, 0.57]) and macrolides (0.43 [0.38, 0.47] vs. 0.28 [0.25, 0.32]) than children in
children. Guidelines recommend amoxicillin as first line therapy for CAP, while mac- children’s hospitals.
rolides are recommended for school-aged children with atypical pneumonia. Despite Conclusion.  Publication of national guidelines was associated with improved
guidelines, antibiotic choice for CAP varies widely among providers. We aimed to antibiotic selection for CAP at both children’s and general hospitals. However, dispari-
determine the impact of outpatient audit and feedback to individual providers on ties in prescribing between children’s and general hospitals persist.
adherence with published guidelines.
Methods.  We conducted a randomized controlled trial of primary care clinicians Figure.  Trends in standardized probability of receiving select antibiotics for CAP
in a multi-state primary care network from 8/2016–2/2017. Providers received base-
line education. The intervention included personalized feedback from investigators
at 1-month intervals on the provider’s management of a case of CAP identified by
ICD-10-CM codes. Prescription counts of guideline-recommended antibiotic therapy
were compared between groups by Pearson’s chi-squared. Performance scores incor-
porating diagnostic and treatment decisions such as physical examination elements,
antibiotics and medication dosing appropriate for a CAP encounter as defined by
clinical practice guidelines were calculated for each encounter during study intervals.
Results.  Among the 43 providers, the majority were physicians (76% control, 86%
intervention). There were no significant differences in work hours, years since board
certification, sex or race between groups. 316 distinct cases of CAP were diagnosed
(214 control; 102 intervention). In patients <5 years, there was no significant difference
in prescription of amoxicillin between groups (61/103 (59.2%) control; 23/48 (47.9%)
intervention, P = 0.19). In patients ≥5 years, there was a significant difference in pre-
scription of guideline recommended antibiotics of amoxicillin or azithromycin (81/103
(78.6%) control; 48/51 (94.1%) intervention, P < 0.05). There was a small, but apparent
upward trend in mean performance scores in the intervention group (Figure 1).
Conclusion.  Personalized, scheduled audit and feedback in the outpatient setting had a
small but measurable impact on improving physician adherence with guidelines. Audit and
feedback alone is insufficient to substantially improve guideline adherence in the manage-
ment of CAP and should be combined with other antimicrobial stewardship interventions.

Disclosures.  All authors: No reported disclosures.

1602. Antibiotic Prescribing for Pediatric Community-Acquired Pneumonia at

Children’s Hospitals and General Hospitals Following National Guideline Release
Alison Tribble, MD1; Rachael Ross, MPH2 and Jeffrey S. Gerber, MD, PhD2
1
Department of Pediatrics and Communicable Diseases, Division of Infectious
Diseases, University of Michigan Medical School, Ann Arbor, Michigan, 2Department
of Pediatrics, Division of Infectious Diseases, Children’s Hospital of Philadelphia,
Disclosures.  All authors: No reported disclosures.
Philadelphia, Pennsylvania
Session: 169. Stewardship: Pediatric Antimicrobial Stewardship
1603. Diagnosis and Management of Pediatric Community Acquired Pneumonia
Friday, October 6, 2017: 12:30 PM Requiring Hospitalization: How Well Are We Following National Guidelines?
Background.  The 2011 IDSA/PIDS guidelines for pediatric community-acquired Andrew Shieh, BA and Jessica E Ericson, MD; Penn State University College of
pneumonia (CAP) recommend penicillin, amoxicillin, or ampicillin (penicillins) as Medicine, Hershey, Pennsylvania

Poster Abstracts  •  OFID 2017:4 (Suppl 1) • S501

Downloaded from https://academic.oup.com/ofid/article-abstract/4/suppl_1/S501/4295404/Diagnosis-and-Management-of-Pediatric-Community

by guest
on 04 October 2017
 Session: 169. Stewardship: Pediatric Antimicrobial Stewardship
Friday, October 6, 2017: 12:30 PM
Background.  Guidelines for the diagnosis and treatment of pediatric community
acquired pneumonia (CAP) were updated in 2011. It is unknown how well guidelines
are used by physicians to manage CAP for children who require hospitalization.
Methods.  Diagnosis codes were used to identify patients from 4  months to
18  years old with a diagnosis of CAP between January 2012 and December 2015.
Hospital records were reviewed to confirm the diagnosis of CAP and to determine
patient demographics, risk factors, clinical characteristics, and treatment outcomes.
Patients who were immunocompromised for any reason, had cystic fibrosis, a current
tracheostomy, or other concurrent bacterial illnesses were excluded. Factors for chil-
dren who were treated according to guideline recommendations and those who were
not were compared using Fisher’s exact test or Mann–Whitney test. A multivariable
logistic regression analysis evaluated the relationship between patient factors, clinical
characteristics, and guideline adherence. Data analysis was performed using Stata 14.
Results.  Of the 154 children with CAP, 90 (58%) were treated according to the
guidelines. In non-adherent cases, antibiotic coverage was too broad in 23 (36%)
patients, included unnecessary MRSA coverage in 11 (17%), and was of prolonged
duration in 20 (31%) patients. Only 10 (16%) had antibiotic coverage that was too nar-
row. The adherent group had a 1-day shorter length of stay (LOS) (P = 0.05) and 2-day
shorter duration of antibiotic therapy (P < 0.01). There was no significant difference
in the number of chest X-rays performed, complications, duration of fever, supple-
mental oxygen use, and need for intensive care unit admission. On regression analysis,
older age and age-appropriate immunization status were significantly associated with
decreased adjusted odds of guideline adherence, odds ratio (OR) = 0.8, 95% confidence
interval (CI) [0.66, 0.95], and 0.09, [0.01, 0.59], respectively.
Conclusion.  Guideline adherence is associated with similar outcomes, shorter LOS,
and duration of treatment compared with non-adherence. Further studies should investi-
gate why older children are less likely to receive recommended antibiotic therapy for CAP.
Disclosures.  All authors: No reported disclosures. 
specifically from wound specimens and in acute osteoarticular infections (OAI) were
compared with available antibiograms at a freestanding children’s hospital.
Methods.  Encounter, billing, and electronic microbiology surveillance data were
utilized to identify SA wound cultures and acute OAI (osteomyelitis and septic arth-
ritis) cases at Women and Children’s Hospital of Buffalo, from 2013 to 2016. OAI cases’
medical records were reviewed to ensure diagnostic accuracy. SA wound and OAI
specific data were tabulated and compared with published institutional antibiograms.
General pediatric locations were defined as community clinics, the emergency room,
and general pediatric wards, with intensive-care and oncology units excluded.
Results.  Significant discordance existed between general pediatrics SA suscepti-
bilities in the aggregate antibiograms, with both wound cultures and OAI cases, for all
4 years: Figures 1 and 2. The proportion of SA that was methicillin-susceptible (MSSA)
was consistently higher in wound specimens than in aggregate data (e.g., 63% vs. 53%
in 2016; p≤ 0.01), and is increasing: 63% in 2015–16 vs. 53% in 2013–14, p ≤0.01.
Clindamycin (clinda) susceptibility for all SA (MSSA + MRSA) was higher in wound
cultures than aggregate data, 89% vs. 82% for 2013–2015 (p ≤ 0.01). For OAI cases,
the proportion of MSSA was consistently ~20% higher than in aggregate data (2016:
79% vs. 53%, P = 0.05), and clinda susceptibility for all SA in this group appears to be
decreasing: 83% in 2015–16 vs. 96% in 2013–14, P = 0.13.
Conclusion.  While our institutional antibiograms created uncertainty, a wound
culture review indicated that clinda remains an appropriate empiric choice for commu-
nity-onset skin and soft-tissue infections. Conversely, an OAI specific analysis revealed a
predominance of MSSA and higher rates of clinda-resistant SA, leading us to reappraise the
empiric use of clinda in this subset. Pediatric facilities should emphasize stratified, speci-
men and clinical-context specific—rather than aggregate—antibiograms, especially for SA.

1604. Use of MRSA Nasal Swab to Guide Empiric Antibiotic Treatment of

Hospital Acquired or Community Acquired Pneumonia in a Pediatric Population
Daniel Linfesty, MD1 and John Manaloor, MD2; 1Internal Medicine and Pediatrics,
Indiana University, Indianapolis, Indiana, 2Pediatrics, Indiana University School of
Medicine, Indianapolis, Indiana
Session: 169. Stewardship: Pediatric Antimicrobial Stewardship
Friday, October 6, 2017: 12:30 PM
Background.  Current PIDS/IDSA guidelines recommend the use of MRSA
empiric coverage in the case of hospital acquired pneumonia (HAP) and in community
acquired pneumonia (CAP) if patients have risk factors or clinical characteristics con-
sistent with MRSA infection. Retrospective studies in adult patients have shown the
MRSA PCR nasal swab to have a negative predictive value of 99% in patients treated
for pneumonia in the inpatient setting, making the MRSA nasal swab a potential tool
to guide de-escalation of empiric antibiotics. No published studies to date have exam-
ined the sensitivity and specificity of MRSA PCR nasal swab in pediatric populations
Methods.  A cohort of patients was identified by cross-matching internal physician
and laboratory billing data from the past 10 years at IU hospitals in the Indianapolis area
for pneumonia. An initial pool of 550 patients were identified. Patients less than 25 years
of age were eligible. Patients were excluded if they had medical conditions such as Cystic
Fibrosis, Chronic Lung Disease, or cavitary pneumonia seondary to IV drug use. Chart
review identified a total of 28 patients that met diagnostic criteria for pneumonia, had
culture data and had a MRSA PCR nasal swab performed during treatment.
Results.  In the cohort, 5 patients had positive MRSA nasal swab and positive cul-
tures for MRSA. Two patients had positive MRSA swab with negative cultures. Twenty- Disclosures.  All authors: No reported disclosures.
one patients had negative MRSA nasal swabs and cultures without MRSA growth. No
patients were identified with a negative MRSA nasal swab that grew MRSA in cultures.
In this population the MRSA nasal swab had a sensitivity of 100%, specificity of 91%,
positive predictive valve of 71%, and a negative predictive value of 100%. The patients 1606. Adverse Effects from Antibiotics for Acute Respiratory Tract Infections in
with negative MRSA swab and negative cultures, 66% were treated with vancomycin. Children: Comparison of Two Data Sources
If MRSA nasal PCR was used to guide treatment, 71% of the patients with a negative Jeffrey S. Gerber, MD, PhD1; Rachael Ross, MPH1; Matthew Bryan, PhD2; A.
nasal swab had potential to discontinue MRSA empiric antibiotics sooner. Russell Localio, PhD3; Theoklis Zaoutis, MD, MSCE4; Richard Wasserman, MD5
Conclusion.  The MRSA PCR nasal swab has a high negative predictive valve in and Alexander Fiks, MD, MSCE6; 1Department of Pediatrics, Division of Infectious
this pediatric population of inpatients treated for HAP/CAP. This is consistent with Diseases, Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania, 2Children’s
results from adult studies. The high negative predictive valve makes the MRSA PCR Hospital of Philadelphia, Philadelphia, Pennsylvania, 3University of Pennsylvania,
nasal swab a potential tool as a rapid diagnostic test to guide empiric antibiotic therapy. Philadelphia, Pennsylvania, 4Division of Infectious Diseases, Children’s Hospital
Disclosures.  All authors: No reported disclosures.  of Philadelphia, Philadelphia, Pennsylvania, 5University of Vermont, Burlington,
Vermont, 6Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania
Session: 169. Stewardship: Pediatric Antimicrobial Stewardship
1605. Consistent Differences between Wound Culture and Osteoarticular
Infection Staphylococcus aureus Susceptibilities and Institutional Antibiograms at Friday, October 6, 2017: 12:30 PM
a Children’s Hospital Background.  Outpatient acute respiratory tract infections (ARTIs) account
Shamim Islam, MD, DTM&H1; Neel Shah, MD2 and Nora Biary, MD3; 1University for the majority of antibiotic exposure in children. Thus, is is essential to under-
at Buffalo, State University of New York, Buffalo, New York, 2Pediatrics Residency stand the outcomes and adverse effect profiles of different therapeutic approaches to
Program, University At Buffalo, State University of NY, Buffalo, New York, 3University managing these common infections. In a study comparing the effectiveness of nar-
At Buffalo, State University of New York, Buffalo, New York row- and broad-spectrum antibiotics for treatment of ARTIs, we compared rates of
Session: 169. Stewardship: Pediatric Antimicrobial Stewardship adverse effects reported by patients to rates obtained by the electronic health record.
Methods.  We used a retrospective cohort and a prospective cohort, both of which
Friday, October 6, 2017: 12:30 PM
included children treated with antibiotics for an ARTI (acute otitis media, Group
Background.  Aggregate hospital and unit-based antibiograms guide empiric A streptococcal pharyngitis, acute sinusitis) in a network of 31 pediatric primary care
antibiotic decision making but may not best inform the microbiology of certain practices. In the retrospective cohort, adverse drug effects including diarrhea, can-
important presentations. In this analysis, Staphylococcus aureus (SA) susceptibilities didiasis, non-candida rash, other allergic reaction, vomiting, and unspecified adverse

S502 • OFID 2017:4 (Suppl 1) • Poster Abstracts

Downloaded from https://academic.oup.com/ofid/article-abstract/4/suppl_1/S501/4295404/Diagnosis-and-Management-of-Pediatric-Community

by guest
on 04 October 2017