Clinical Infectious Diseases

MAJOR ARTICLE

Management of Pediatric Pneumonia: A Decade After

the Pediatric Infectious Diseases Society and Infectious
Diseases Society of America Guideline
Lilliam Ambroggio,1 Jillian Cotter,1 Matthew Hall,2 Daniel J. Shapiro,3,4 Susan C. Lipsett,3,4 Adam L. Hersh,5 Samir S. Shah,6,7 Thomas V. Brogan,8,9
Jeffrey S. Gerber,10,11 Derek J. Williams,12 Anne J. Blaschke,5 Jonathan D. Cogen,13 and Mark I. Neuman3,4
1
Sections of Emergency Medicine and Hospital Medicine, Children’s Hospital Colorado, Department of Pediatrics, University of Colorado, Aurora, Colorado, USA; 2Children’s Hospital Association,
Lenexa, Kansas, USA; 3Division of Emergency Medicine, Boston Children’s Hospital, Boston, Massachusetts, USA; 4Departments of Pediatrics and Emergency Medicine, Harvard Medical School,
Boston, Massachusetts, USA; 5Division of Pediatric Infectious Diseases, Department of Pediatrics, School of Medicine, University of Utah, Salt Lake City, Utah, USA; 6Division of Hospital Medicine,

Downloaded from https://academic.oup.com/cid/article/77/11/1604/7206345 by guest on 03 December 2024

Cincinnati Children’s Hospital Medicine Center, Cincinnati, Ohio, USA; 7Department of Pediatrics, University of Cincinnati College of Medicine, Cincinnati, Ohio, USA; 8Division of Critical Care, Seattle
Children’s Hospital, Seattle, Washington, USA; 9Department of Pediatrics, School of Medicine, University of Washington, Seattle, Washington, USA; 10Division of Infectious Diseases, Children’s
Hospital of Philadelphia, Philadelphia, Pennsylvania, USA; 11Department of Pediatrics, Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania, USA; 12Division of
Hospital Medicine, Department of Pediatrics, Monroe Carell Jr Children’s Hospital at Vanderbilt, Vanderbilt University School of Medicine, Nashville, Tennessee, USA; and 13Division of Pulmonary
Medicine and Sleep Medicine, Seattle Children’s Hospital, University of Washington, Seattle, Washington, USA

Background. Incomplete uptake of guidelines can lead to nonstandardized care, increased expenditures, and adverse clinical
outcomes. The objective of this study was to evaluate the impact of the 2011 Pediatric Infectious Diseases Society and Infectious
Diseases Society of America (PIDS/IDSA) pediatric community-acquired pneumonia (CAP) guideline that emphasized
aminopenicillin use and de-emphasized the use of chest radiographs (CXRs) in certain populations.
Methods. This quasi-experimental study queried a national administrative database of children’s hospitals to identify children
aged 3 months–18 years with CAP who visited 1 of 28 participating hospitals from 2009 to 2021. PIDS/IDSA pediatric CAP
guideline recommendations regarding antibiotic therapy, diagnostic testing, and imaging were evaluated. Segmented regression
interrupted time series was used to measure guideline-concordant practices with interruptions for guideline publication and the
Coronavirus Disease 2019 (COVID-19) pandemic.
Results. Of 315 384 children with CAP, 71 804 (22.8%) were hospitalized. Among hospitalized children, there was a decrease in
blood culture performance (0.5% per quarter) and increase in aminopenicillin prescribing (1.1% per quarter). Among children
discharged from the emergency department (ED), there was an increase in aminopenicillin prescription (0.45% per quarter),
whereas the rate of obtaining CXRs declined (0.12% per quarter). However, use of CXRs rebounded during the COVID-19
pandemic (increase of 1.56% per quarter). Hospital length of stay, ED revisit rates, and hospital readmission rates remained stable.
Conclusions. Guideline publication was associated with an increase of aminopenicillin prescribing. However, rates of
diagnostic testing did not materially change, suggesting the need to consider implementation strategies to meaningfully change
clinical practice for children with CAP.
Keywords. diagnostic testing; antibiotics; pneumonia; pediatric.

Community-acquired pneumonia (CAP) is the fifth most prev­ Diseases Society and Infectious Diseases Society of America
alent and second most costly reason for hospitalization among (PIDS/IDSA) jointly published a guideline in 2011 for the man­
children [1]. Additionally, CAP accounts for the most antibiot­ agement of otherwise healthy children with suspected CAP.
ic days of therapy for children who are hospitalized [2]. Prior to The guideline had 92 specific recommendations and was orga­
2011, there was wide variability in the use of diagnostic testing, nized by patient disposition (discharge from the emergency
antibiotic choice, and hospitalization rates among children department [ED] vs hospitalization), diagnostic testing, anti-
with CAP [3, 4]. To standardize care, the Pediatric Infectious infective treatment, discharge criteria, and prevention. These
recommendations were developed by existing evidence, and ex­
pert consensus therefore some had stronger evidence bases (eg,
Received 11 April 2023; editorial decision 13 June 2023; published online 23 June 2023 aminopenicillin prescribing) whereas others had weaker evi­
Correspondence: L. Ambroggio, Sections of Emergency Medicine, and Hospital Medicine, dence bases (eg, macrolide prescribing).
13123 E 16th Ave, B251, Aurora, CO 80045 ([email protected]);
M. Neuman, Division of Emergency Medicine, Boston Children’s Hospital, 300 Longwood Implementation of guidelines is often delayed and/or incom­
Ave, Main South, 0120, Boston, MA 02115 ([email protected]). plete [5–7]. Incomplete uptake of guidelines can lead to non­
Clinical Infectious Diseases® 2023;77(11):1604–11
© The Author(s) 2023. Published by Oxford University Press on behalf of Infectious Diseases
standardized care, increased healthcare expenditures, and
Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@ suboptimal outcomes [8]. Although some studies have evaluat­
oup.com
ed the impact of the PIDS/IDSA guideline on specific diagnostic
https://doi.org/10.1093/cid/ciad385

1604 • CID 2023:77 (1 December) • Ambroggio et al

testing practices and antibiotic use [9–13], a broader investiga­ Billing data were used to assess rates of performance of diag­
tion in the past decade has not been reported. Additionally, nostic testing, including blood culture, complete blood count
many recommendations in the guideline are about reducing di­ (CBC), chest radiograph (CXR), and acute phase reactants, in­
agnostic tests that contribute to low-value care. Therefore, the cluding erythrocyte sedimentation rate (ESR), C-reactive pro­
objective of this study was to evaluate the impact of the guide­ tein (CRP), and procalcitonin (PCT). In general, the
line on the management of pediatric CAP using a national ad­ guideline recommends less overall diagnostic testing for chil­
ministrative database of children’s hospitals. dren with CAP unless it will directly inform management.
The guideline strongly recommends narrow-spectrum antibi­
otic use for uncomplicated CAP as first-line therapy for chil­
METHODS dren without penicillin allergy, specifically amoxicillin for
Study Design and Population children treated in the outpatient setting and amoxicillin or
We performed a quasi-experimental study of children 3 ampicillin for children treated in the ED or inpatient setting.
months to 18 years old with an International Classification Throughout the manuscript, “aminopenicillin” is used to indi­

Downloaded from https://academic.oup.com/cid/article/77/11/1604/7206345 by guest on 03 December 2024

of Diseases, Ninth Revision (ICD-9) or Tenth Revision cate receipt of either amoxicillin or ampicillin. We defined
(ICD-10) diagnosis code indicating pneumonia (ICD-9: broad-spectrum antibiotics as any second- or third-generation
480–483, 485–486; ICD-10: J12.x, J13, J14, J15.x, J16.x, J18.0, cephalosporin. We assessed macrolide therapy separately as the
J18.1, J18.8, J18.9) or pleurisy (ICD-9: 511.0, 511.1, 511.9). guideline recommends macrolide therapy when atypical CAP is
Children treated at a Pediatric Health Information System suspected. Children who did not receive antibiotics were also
(PHIS; Children’s Hospital Association, Lenexa, Kansas) par­ included in the study and presumed to have viral CAP.
ticipating hospital between 1 January 2009 and 31 December
2021 were eligible for inclusion. The PHIS database contains Clinical Outcomes

administrative and billing data from 47 US children’s hospi­ To assess the impact of specific recommendations on children
tals. After excluding 19 hospitals that did not have complete with CAP, we evaluated clinical outcomes over time. For chil­
data for the entire study period, 28 hospitals were included dren who were hospitalized, outcomes assessed were hospital
in the analytic dataset. length of stay (LOS) in days, admission to an ICU, death during
Children were excluded if they had a chronic complex con­ hospitalization, and readmission to the hospital within 7, 14,
dition [14], were transferred from an outside hospital, or had a and 30 days. For children discharged from the ED, we assessed
prior diagnosis of CAP within the past 30 days [15]. Children revisits occurring within 7, 14, and 30 days.
who were transferred directly to the intensive care unit (ICU)
from the ED, or had empyema or a chest drainage procedure Statistical Analysis

in the ED or on the first day of hospitalization, were also ex­ Continuous variables were described using median, range, and
cluded to minimize the inclusion of children who were severely interquartile range, and categorical variables were described
ill upon ED presentation. PHIS does not capture urgent care with counts and percentages. Due to the large sample size of
visits or outpatient clinical visits. Results are presented in ag­ the cohort, statistical hypothesis testing was not conducted
gregate in 2 groups: children discharged home from the ED, when comparing settings of care or individual years, as statisti­
and children admitted (under inpatient or observation status) cal significance may not reflect clinical significance. We inves­
to the hospital after ED evaluation. This study was considered tigated overall annual trends using Cochran-Armitage trend
non–human subjects research and exempt as determined by the tests.
Colorado Multiple Institutional Review Board. Interrupted time series (ITS) analysis evaluates population-
based interventions at clearly defined timepoints to understand
the underlying trend before and after the intervention. We used
Exposures segmented regression ITS to assess the impact of the guideline
The main exposures of the study were the specific recommen­ publication in August 2011 on diagnostic testing, antibiotic use,
dations over the decade since the guideline was published. We and clinical outcomes. The segments were categorized by quar­
evaluated recommendations from the guideline that could be ter and because the guideline recommendations varied depend­
reliably obtained from PHIS. Therefore, recommendations ing on the setting of care, we performed separate ITS analyses
that included laboratory results or vital signs, for instance, for ED discharge and hospitalized encounters. We wanted to
were not examined. The full list of selected recommendations evaluate the impact of the guideline a decade after publication;
is presented in Supplementary Table 1. We assessed perfor­ however, due to the unique scenario and changing medical
mance of diagnostic testing, antimicrobial use over the study practices during the Coronavirus Disease 2019 (COVID-19)
period, and in relation to guideline publication and the pandemic, we added a second “interruption” in the ITS analy­
COVID-19 pandemic. ses to investigate trends that may have been affected by the

Guideline for Pediatric Pneumonia • CID 2023:77 (1 December) • 1605

COVID-19 pandemic beginning in March 2020. All ITS models Table 1. Characteristics of Patient Population
were adjusted for age in years, severity (using the
Discharged From
Hospitalization Resource Intensity Scores for Kids [H-RISK]
Hospitalized ED
method) [16], and hospital. Results of the ITS are represented Overall (n = 71 804 (n = 243 580
as adjusted proportions in the figures. A P value <.05 was de­ Characteristic (N = 315 384) [22.8%]) [77.2%])

termined as statistically significant for ITS models. All statisti­ Age, y

cal analyses were performed using SAS statistical software <1 29 684 (9.4) 8672 (12.1) 21 012 (8.6)
1–4 159 643 (50.6) 39 712 (55.3) 119 931 (49.2)
(version 9.4, SAS Institute, Cary, North Carolina).
5–9 81 600 (25.9) 16 013 (22.3) 65 587 (26.9)
10–14 32 015 (10.2) 5401 (7.5) 26 614 (10.9)
15–18 12 442 (3.9) 2006 (2.8) 10 436 (4.3)
RESULTS
Female sex 150 386 (47.7) 34 852 (48.5) 115 534 (47.4)
Study Population Race/ethnicity
A total of 315 384 children were included in the study; 71 804 Non-Hispanic 122 249 (38.8) 30 877 (43) 91 372 (37.5)

Downloaded from https://academic.oup.com/cid/article/77/11/1604/7206345 by guest on 03 December 2024

White
(22.8%) were hospitalized, and 243 580 (77.2%) were dis­
Non-Hispanic 74 964 (23.8) 15 239 (21.2) 59 725 (24.5)
charged from the ED. Half of the children (50.6%) were 1–4 Black
years of age and had government insurance (58%) (Table 1). Hispanic 80 301 (25.5) 16 972 (23.6) 63 329 (26)
The mean number of encounters for CAP prepandemic was Asian 11 510 (3.6) 2767 (3.9) 8743 (3.6)
Other 26 360 (8.4) 5949 (8.3) 20 411 (8.4)
26 766 (range, 22 633–33 936) per year; however, during the
Payor
COVID-19 pandemic, CAP encounters decreased to a mean Government 183 062 (58) 39 100 (54.5) 143 962 (59.1)
of 10 480 (range, 9576–11 384) per year. The annual range of Private 109 235 (34.6) 28 610 (39.8) 80 625 (33.1)
inpatient admissions prepandemic was 5162–7463, and ED en­ Other 23 087 (7.3) 4094 (5.7) 18 993 (7.8)
counters ranged from 17 247 to 26 473 annually. During the Year
2009 33 936 (10.8) 7463 (10.4) 26 473 (10.9)
pandemic, the number of hospitalizations and ED encounters
2010 29 426 (9.3) 6716 (9.4) 22 710 (9.3)
decreased dramatically to 2276–3258 and 7300–8126, respec­
2011 29 556 (9.4) 6306 (8.8) 23 250 (9.5)
tively. During the study period, blood cultures were obtained 2012 28 763 (9.1) 6465 (9) 22 298 (9.2)
in 44.6% of children, CBCs in 24.8%, acute phase reactants in 2013 23 283 (7.4) 5707 (7.9) 17 576 (7.2)
27.6%, and CXR in 83.1%. Among children who received an an­ 2014 23 000 (7.3) 5618 (7.8) 17 382 (7.1)
tibiotic during their encounter, aminopenicillins were given to 2015 26 244 (8.3) 5806 (8.1) 20 438 (8.4)
2016 26 736 (8.5) 5671 (7.9) 21 065 (8.6)
an average 50% of children hospitalized with CAP and in an av­
2017 23 478 (7.4) 5162 (7.2) 18 316 (7.5)
erage of 14.8% of children discharged from the ED with CAP 2018 22 633 (7.2) 5386 (7.5) 17 247 (7.1)
(Table 2). There were 91.1% of children hospitalized with 2019 27 369 (8.7) 5970 (8.3) 21 399 (8.8)
CAP who received antibiotics; 15.1% received an antibiotic oth­ 2020 9576 (3) 2276 (3.2) 7300 (3)
er than aminopenicillin, macrolide, or broad-spectrum cepha­ 2021 11 384 (3.6) 3258 (4.5) 8126 (3.3)

losporin. Among children discharged from the ED with CAP, Abbreviation: ED, emergency department.

27.4% of children received an antibiotic, and 7.4% received

an antibiotic other than those listed.
Inpatient Recommendations
Emergency Department Recommendations Among children hospitalized with CAP, there was a decrease in
Among children discharged home from the ED, guideline pub­ the rate of obtaining blood cultures (0.5% per quarter) and
lication was associated with an increasing trend in aminopeni­ CXRs (0.31% per quarter), and an increase in the rate of amino­
cillin ordered in the ED (0.45% increase per quarter), penicillin use post–guideline publication with an associated de­
performance of acute phase reactants and CBC (0.1% and crease in broad-spectrum therapy (1.1% per quarter) (Figure 1,
0.09% increase per quarter, respectively), and a decline in the Supplementary Table 1). No change in rates of performance of
rate of CXR performance (0.12% decrease per quarter) CBC or acute phase reactants was observed among hospitalized
(Table 3, Figure 1, and Supplementary Table 2). Rates of children relative to guideline publication.
blood culture performance remained stable among children During the COVID-19 pandemic, although there was a sub­
discharged from the ED. The onset of the COVID-19 pandemic stantial decrease in total encounters due to CAP, there was a
was associated with a reduction in blood culture (0.6% per statistically significant increase in aminopenicillin use (0.45%
quarter), acute phase reactant, and CBC performance (0.24% per quarter) and a decrease in obtaining a CBC (0.09% per
and 0.25% per quarter respectively), an increase in CXR perfor­ quarter). Performance of acute phase reactants (0.10% per
mance (1.56%), and no change in aminopenicillin use for chil­ quarter) increased during the COVID-19 pandemic, and
dren discharged home from the ED. CXR performance decreased (0.12% per quarter).

1606 • CID 2023:77 (1 December) • Ambroggio et al

 Table 2. Annual Rates of Recommendations and Outcomes in Children Hospitalized With Community-Acquired Pneumonia

Year and Number of Encounters

Overall 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021
Recommendation/Outcome (71 804) (7463) (6716) (6306) (6465) (5707) (5618) (5806) (5671) (5162) (5386) (5970) (2276) (3258)

Blood culture 32 024 (44.6) 3763 (50.4) 3405 (50.7) 3112 (49.3) 2956 (45.7) 2597 (45.5) 2210 (39.3) 2309 (39.8) 2210 (39) 2028 (39.3) 2303 (42.8) 2652 (44.4) 1024 (45) 1455 (44.7)
Acute phase reactants 19 847 (27.6) 1953 (26.2) 1381 (20.6) 1397 (22.2) 1453 (22.5) 1229 (21.5) 1242 (22.1) 1442 (24.8) 1425 (25.1) 1459 (28.3) 1716 (31.9) 2481 (41.6) 1005 (44.2) 1664 (51.1)
ESR 5564 (7.7) 421 (5.6) 369 (5.5) 459 (7.3) 466 (7.2) 377 (6.6) 378 (6.7) 442 (7.6) 426 (7.5) 477 (9.2) 451 (8.4) 581 (9.7) 261 (11.5) 456 (14)
CRP 17 821 (24.8) 1925 (25.8) 1320 (19.7) 1325 (21) 1393 (21.5) 1171 (20.5) 1184 (21.1) 1395 (24) 1385 (24.4) 1399 (27.1) 1470 (27.3) 1869 (31.3) 765 (33.6) 1220 (37.4)
PCT 2879 (4) 2 (0) 12 (0.2) 28 (0.4) 16 (0.2) 24 (0.4) 32 (0.6) 67 (1.2) 86 (1.5) 110 (2.1) 377 (7) 854 (14.3) 407 (17.9) 864 (26.5)
CBC 17 821 (24.8) 1925 (25.8) 1320 (19.7) 1325 (21) 1393 (21.5) 1171 (20.5) 1184 (21.1) 1395 (24) 1385 (24.4) 1399 (27.1) 1470 (27.3) 1869 (31.3) 765 (33.6) 1220 (37.4)
Chest imaging 59 647 (83.1) 6532 (87.5) 5919 (88.1) 5550 (88) 5575 (86.2) 4868 (85.3) 4614 (82.1) 4827 (83.1) 4587 (80.9) 4220 (81.8) 4307 (80) 4578 (76.7) 1714 (75.3) 2356 (72.3)
Antimicrobial therapy
Aminopenicillin 35 913 (50) 2095 (28.1) 2031 (30.2) 2129 (33.8) 2627 (40.6) 2747 (48.1) 2929 (52.1) 3533 (60.9) 3632 (64) 3325 (64.4) 3539 (65.7) 3833 (64.2) 1403 (61.6) 2090 (64.1)
2nd-/3rd-generation 10 215 (14.2) 2506 (33.6) 1957 (29.1) 1358 (21.5) 990 (15.3) 703 (12.3) 594 (10.6) 477 (8.2) 391 (6.9) 338 (6.5) 314 (5.8) 331 (5.5) 113 (5) 143 (4.4)
cephalosporins
Macrolides 20 807 (29) 2339 (31.3) 2263 (33.7) 2334 (37) 2573 (39.8) 1894 (33.2) 1836 (32.7) 1638 (28.2) 1521 (26.8) 1228 (23.8) 1069 (19.8) 1387 (23.2) 469 (20.6) 256 (7.9)
No antibiotics 17 265 (24) 1874 (25.1) 1819 (27.1) 1692 (26.8) 1591 (24.6) 1349 (23.6) 1220 (21.7) 1152 (19.8) 1088 (19.2) 1070 (20.7) 1276 (23.7) 1441 (24.1) 641 (28.2) 1052 (32.3)
Hospital LOS, d, mean (SD) 1.7 (1.8) 1.8 (1.8) 1.8 (1.8) 1.8 (1.8) 1.7 (1.8) 1.7 (1.8) 1.7 (1.8) 1.7 (1.8) 1.7 (1.8) 1.7 (1.8) 1.7 (1.8) 1.7 (1.7) 1.7 (1.8) 1.7 (1.7)
ICU admission 302 (0.4) 12 (0.2) 22 (0.3) 20 (0.3) 17 (0.3) 20 (0.4) 30 (0.5) 35 (0.6) 37 (0.7) 31 (0.6) 21 (0.4) 28 (0.5) 16 (0.7) 13 (0.4)
Any revisit—all cause
7d 2351 (3.3) 240 (3.2) 211 (3.1) 198 (3.1) 197 (3) 175 (3.1) 156 (2.8) 202 (3.5) 195 (3.4) 177 (3.4) 189 (3.5) 205 (3.4) 106 (4.7) 100 (3.1)
14 d 3746 (5.2) 367 (4.9) 331 (4.9) 321 (5.1) 310 (4.8) 284 (5) 281 (5) 316 (5.4) 317 (5.6) 291 (5.6) 292 (5.4) 330 (5.5) 148 (6.5) 158 (4.8)
30 d 6560 (9.1) 650 (8.7) 614 (9.1) 556 (8.8) 530 (8.2) 511 (9) 490 (8.7) 532 (9.2) 547 (9.6) 509 (9.9) 526 (9.8) 568 (9.5) 223 (9.8) 304 (9.3)
Readmission—all cause
7d 1129 (1.6) 119 (1.6) 86 (1.3) 102 (1.6) 83 (1.3) 84 (1.5) 85 (1.5) 91 (1.6) 99 (1.7) 91 (1.8) 93 (1.7) 88 (1.5) 57 (2.5) 51 (1.6)
14 d 1598 (2.2) 168 (2.3) 126 (1.9) 140 (2.2) 120 (1.9) 115 (2) 130 (2.3) 135 (2.3) 141 (2.5) 122 (2.4) 133 (2.5) 123 (2.1) 72 (3.2) 73 (2.2)
30 d 2557 (3.6) 254 (3.4) 219 (3.3) 218 (3.5) 199 (3.1) 191 (3.3) 209 (3.7) 210 (3.6) 223 (3.9) 195 (3.8) 209 (3.9) 201 (3.4) 98 (4.3) 131 (4)
Data are presented as No. (%) unless otherwise indicated.
Abbreviations: CBC, complete blood count; CRP, C-reactive protein; ESR, erythrocyte sedimentation rate; ICU, intensive care unit; LOS, length of stay; PCT, procalcitonin; SD, standard deviation.

Guideline for Pediatric Pneumonia • CID 2023:77 (1 December) • 1607

Downloaded from https://academic.oup.com/cid/article/77/11/1604/7206345 by guest on 03 December 2024
 Table 3. Annual Rates of Recommendations and Outcomes in Emergency Department Discharges for Children With Community-Acquired Pneumonia

Year and Number of Encounters

Recommendation/ Overall 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021
Outcome (243 580) (26 473) (22 710) (23 250) (22 298) (17 576) (17 382) (20 438) (21 065) (18 316) (17 247) (21 399) (7300) (8126)

Blood culture 19 153 (7.9) 2803 (10.6) 2214 (9.7) 1845 (7.9) 1686 (7.6) 1308 (7.4) 1138 (6.5) 1395 (6.8) 1435 (6.8) 1165 (6.4) 1120 (6.5) 1535 (7.2) 618 (8.5) 891 (11)
CBC 10 874 (4.5) 1117 (4.2) 737 (3.2) 703 (3) 717 (3.2) 621 (3.5) 638 (3.7) 921 (4.5) 952 (4.5) 923 (5) 829 (4.8) 1280 (6) 568 (7.8) 868 (10.7)
ESR 4244 (1.7) 271 (1) 225 (1) 259 (1.1) 262 (1.2) 235 (1.3) 214 (1.2) 374 (1.8) 384 (1.8) 355 (1.9) 373 (2.2) 543 (2.5) 254 (3.5) 495 (6.1)

1608 • CID 2023:77 (1 December) • Ambroggio et al

CRP 10 874 (4.5) 1117 (4.2) 737 (3.2) 703 (3) 717 (3.2) 621 (3.5) 638 (3.7) 921 (4.5) 952 (4.5) 923 (5) 829 (4.8) 1280 (6) 568 (7.8) 868 (10.7)
PCT 989 (0.4) 5 (0) 1 (0) 7 (0) 3 (0) 8 (0) 10 (0) 11 (0.1) 23 (0.1) 94 (0.5) 307 (1.4) 190 (2.6) 330 (4.1)
Acute phase reactants 11 927 (4.9) 1151 (4.3) 785 (3.5) 744 (3.2) 763 (3.4) 664 (3.8) 674 (3.9) 967 (4.7) 991 (4.7) 968 (5.3) 924 (5.4) 1523 (7.1) 700 (9.6) 1073 (13.2)
Chest imaging 197 448 (81.1) 22 739 (85.9) 19 208 (84.6) 19 337 (83.2) 18 248 (81.8) 14 105 (80.3) 13 744 (79.1) 15 954 (78.1) 16 711 (79.3) 14 392 (78.6) 13 599 (78.8) 17 153 (80.2) 5660 (77.5) 6598 (81.2)
Antimicrobial therapy
Aminopenicillin 360 22 (14.8) 2609 (9.9) 2294 (10.1) 2567 (11) 2626 (11.8) 2397 (13.6) 2458 (14.1) 3019 (14.8) 3611 (17.1) 3323 (18.1) 3360 (19.5) 4310 (20.1) 1530 (21) 1918 (23.6)
2nd-/3rd-generation 3936 (1.6) 473 (1.8) 427 (1.9) 504 (2.2) 352 (1.6) 249 (1.4) 216 (1.2) 271 (1.3) 315 (1.5) 278 (1.5) 274 (1.6) 344 (1.6) 101 (1.4) 132 (1.6)
cephalosporins
Macrolides 11 209 (4.6) 1062 (4) 990 (4.4) 1271 (5.5) 1217 (5.5) 792 (4.5) 801 (4.6) 908 (4.4) 1022 (4.9) 872 (4.8) 738 (4.3) 954 (4.5) 346 (4.7) 236 (2.9)
No antibiotics 196 315 (80.6) 22 383 (84.6) 19 126 (84.2) 19 144 (82.3) 18 342 (82.3) 14 296 (81.3) 14 110 (81.2) 16 479 (80.6) 16 423 (78) 14 102 (77) 13 430 (77.9) 16 689 (78) 5656 (77.5) 6135 (75.5)
Any revisits—all cause
7d 18 808 (7.7) 2047 (7.7) 1839 (8.1) 1760 (7.6) 1682 (7.5) 1291 (7.3) 1357 (7.8) 1663 (8.1) 1628 (7.7) 1404 (7.7) 1329 (7.7) 1701 (7.9) 552 (7.6) 555 (6.8)
14 d 23 939 (9.8) 2498 (9.4) 2297 (10.1) 2232 (9.6) 2168 (9.7) 1628 (9.3) 1765 (10.2) 2082 (10.2) 2117 (10) 1798 (9.8) 1721 (10) 2170 (10.1) 713 (9.8) 750 (9.2)
30 d 34 922 (14.3) 3654 (13.8) 3278 (14.4) 3247 (14) 3181 (14.3) 2378 (13.5) 2563 (14.7) 2999 (14.7) 3110 (14.8) 2679 (14.6) 2559 (14.8) 3141 (14.7) 963 (13.2) 1170 (14.4)
Readmission—all
cause
7d 6474 (2.7) 718 (2.7) 562 (2.5) 547 (2.4) 548 (2.5) 422 (2.4) 467 (2.7) 602 (2.9) 579 (2.7) 485 (2.6) 477 (2.8) 655 (3.1) 193 (2.6) 219 (2.7)
14 d 7337 (3) 808 (3.1) 651 (2.9) 621 (2.7) 627 (2.8) 471 (2.7) 521 (3) 673 (3.3) 658 (3.1) 551 (3) 530 (3.1) 738 (3.4) 224 (3.1) 264 (3.2)
30 d 9020 (3.7) 1002 (3.8) 782 (3.4) 761 (3.3) 767 (3.4) 584 (3.3) 651 (3.7) 811 (4) 824 (3.9) 691 (3.8) 663 (3.8) 873 (4.1) 270 (3.7) 341 (4.2)
Data are presented as No. (%).
Abbreviations: CBC, complete blood count; CRP, C-reactive protein; ESR, erythrocyte sedimentation rate; PCT, procalcitonin.

Downloaded from https://academic.oup.com/cid/article/77/11/1604/7206345 by guest on 03 December 2024

 Downloaded from https://academic.oup.com/cid/article/77/11/1604/7206345 by guest on 03 December 2024
Figure 1. Quarterly trends in guideline recommendations. *“Aminopenicillin” is used to indicate receipt of either amoxicillin or ampicillin. Abbreviations: COVID-19,
Coronavirus Disease 2019; ED, emergency department; PIDS/IDSA, Pediatric Infectious Diseases Society and Infectious Diseases Society of America; Q4, quarter 4.

Clinical Outcome Balancing Measures CAP is based on strong previous etiological studies demon­
Among children who were discharged home from the ED, only strating that the most common serious bacterial cause of
9.8% of children revisited the ED within 14 days, and 3% were CAP is Streptococcus pneumoniae [17–19]. This uptake in ami­
admitted upon ED revisit for any cause within 14 days of their nopenicillin prescribing has been described in other studies,
index ED encounter. There were no clinically significant differ­ many of which were published shortly after guideline publica­
ences in clinical outcomes in children discharged from the ED tion [11, 20]. A previous study conducted at 3 institutions only
following guideline publication (Tables 2 and 3). 9 months after guideline publication found that aminopenicil­
Among children who were hospitalized with CAP, the mean lin prescribing among children hospitalized with CAP had an
LOS was 1.7 days (standard deviation, 1.8 days), 0.4% of chil­ absolute increase of 11.3% and that cephalosporin prescribing
dren were admitted to the ICU, 5.2% of children had a revisit was reduced by 12.4% [21]. This study also found that actively
to the ED or hospital within 14 days, and 2.2% of children disseminating the guideline had greater increases in aminope­
were readmitted within 14 days for any cause (Table 2). Since nicillin prescribing (absolute difference of 20.4% for hospitals
the guideline was published, there were no clinically significant with active dissemination compared with 2.0% in the hospital
differences in the outcomes of LOS, ICU admission, any revisit, with passive dissemination of the guideline) [21]. Active dis­
or all-cause readmission (Table 2). semination of the guideline may be imperative to increasing ad­
herence to the recommendation, as was demonstrated in 1
study using quality improvement science where appropriate
DISCUSSION
first-line prescribing increased from 0% to 100% in the ED
More than a decade following publication of the PIDS/IDSA and from 30% to 100% on hospital medicine resident teams
guidelines, this study of >300 000 children diagnosed with in 6 months [11]. In another study, 86.4% of ED providers an­
CAP in the ED and inpatient settings found high uptake of ami­ swered that they would use narrow-spectrum antibiotics for pe­
nopenicillin prescribing, a decrease in CBC and CRP ordering diatric CAP; however, in practice only 46% of children with
in the ED, and minimal changes in the ordering of CXRs. No CAP received narrow-spectrum antibiotics [22]. These studies
clinically significant differences in clinical outcomes were ob­ suggest that although there is strong evidence for the use of
served following guideline publication, indicating that de­ aminopenicillins as first-line therapy in children with CAP, ac­
creased diagnostic testing did not adversely change the tive dissemination of the recommendation results in a quicker
overall outcomes in children with CAP [17]. and more sustained change in practice [23].
The most noticeable change since guideline publication is The PIDS/IDSA guideline strongly recommends obtaining a
the increased use of aminopenicillins. The evidence for the rec­ CXR for children who are hospitalized with moderate to severe
ommendation for aminopenicillins as first-line therapy for CAP; however, the guideline recommends against performance

Guideline for Pediatric Pneumonia • CID 2023:77 (1 December) • 1609

of CXR in children with CAP who are well enough to be man­ ordering from 30% to 19% for children in the ED without
aged as outpatients [17]. In our study, 81% of children dis­ changing the frequency of antibiotic use, LOS, or ED reutiliza­
charged from the ED received a CXR. Since the guideline was tion [33]. These previous studies may have found a larger chan­
published, studies have shown that CXR performance in chil­ ge in diagnostic ordering as they were single-center studies with
dren managed as outpatients does not change overall diagnosis focused interventions.
or antibiotic prescribing [24]. Two studies found that indepen­ Although the current evidence supports the decreased use of
dent of CXR findings, children discharged from the ED with CBC, the evidence to support the use of PCT and CRP has been
CAP were more likely to receive antibiotics for CAP if the clin­ more varied. Ordering of PCT for CAP varied across hospitals
ical provider had an a priori plan to prescribe antibiotics [25, between 1% and 44% compared with CRP, which varied from
26]. Additionally, a study in the outpatient setting including 1% to 89% [34]. This variation indicates the medical uncertain­
42 primary care pediatricians across 10 primary care practices ty of using these diagnostic tests for CAP. Some studies have
demonstrated that CXR were obtained in only 5% of the 1906 suggested that PCT use can be helpful in differentiating a bac­
children diagnosed with CAP, with only 10% of the total cohort terial versus viral etiology for CAP [35, 36]. Additionally, CRP

Downloaded from https://academic.oup.com/cid/article/77/11/1604/7206345 by guest on 03 December 2024

returning for an unscheduled follow-up visit. Of those who had or PCT may have some benefit in distinguishing overall severity
unscheduled visits, <0.5% were ultimately admitted to the hos­ for CAP [28, 37]. Therefore, CRP and PCT may be useful in
pital for CAP [10]. certain situations.
Although blood culture is not routinely recommended in Notably, our study found no statistical differences in hospital
children with CAP, they were obtained in 7.9% of children LOS, ED return visits, or readmissions over the decade following
with CAP in the ED and 44.6% of children hospitalized with guideline publication. This observation is important given the
CAP. Blood cultures are recommended in children requiring major shift toward aminopenicillin prescribing, which was not
hospitalization for presumed bacterial CAP that is moderate associated with worse outcomes in children with CAP.
to severe; however, no published pediatric severity score has Additionally, the reduction in the use of diagnostic testing
been applied clinically to help providers distinguish mild such as blood cultures was not associated with worse outcomes.
from moderate or severe CAP [27, 28]. Thus, some variability There are several limitations to this study. First this was a
in practice may be related to interpretation of this recommen­ quasi-experimental study design investigating the influence of
dation. One study interpreted the recommendation as suggest­ a national guideline on the management of children with
ing that any child hospitalized with CAP should have a blood CAP across US children’s hospitals and may not be generaliz­
culture obtained. The study demonstrated an increase in ob­ able to all settings. Other factors, such as the presence of any
taining blood cultures from 53% to 90% in 6 months with a pos­ institution-specific guidelines, quality improvement initiatives,
itivity rate of 2.5% [29]. In a separate study, blood cultures were or medical-legal considerations may have contributed to the
obtained on 2143 children, only 2.2% of cultures were positive, observed changes in this study [12, 13, 38]. Second, as PHIS
with S. pneumoniae being the most common pathogen detected uses billing data, the study cannot investigate whether results
[30]. A multicenter national study of 7509 children hospitalized from diagnostic tests could have changed over time (eg, more
with CAP had similar findings, in which only 2.5% of children positive blood cultures over time; however, published studies
had a positive blood culture with S. pneumoniae being the most suggest this is not the case), nor was the study able to investigate
prevalent [31]. As the guideline recommends aminopenicillins, all recommendations from the guideline. Third, it is unknown
which are effective against S. pneumoniae, as first-line therapy, whether illness severity has changed over time, impacting the
and the high prevalence of viral etiologies, this may suggest ob­ utilization of the ED or hospitalization among children diag­
taining blood cultures in all children who are hospitalized with nosed with CAP. Fourth, although the recommendations are
CAP may be of low yield as a diagnostic tool. to reduce diagnostic testing for children with suspected CAP,
Diagnostic tests, such as ESR, CRP, PCT, and CBC are not there may be other diagnoses as part of the differential; there­
routinely recommended for children with uncomplicated fore, it is difficult using administrative data to determine the in­
CAP; however, recent publications indicate that these tests tent of ordering the specific diagnostic test or medication.
are still obtained in most hospitals. In our study, use of some Therefore, further studies investigating the clinical decision
of these tests statistically increased in the ITS analysis since making behind these recommendations is a critical step in
the guideline publication with a substantial increase during changing behavior. Fifth, vaccine history and/or immunization
the COVID-19 pandemic. Our findings are in contrast to pre­ status was not available in PHIS, therefore we were not able to
vious studies, as a nationally representative study found that assess “appropriate” antibiotic therapy for CAP, but rather
CBCs were ordered in a weighted average of 20% of children overall trends in antibiotic prescriptions.
with CAP who visited the ED between 2008 and 2015, with In conclusion, we observed that the PIDS/IDSA guideline
no statistical difference before and after the guideline was pub­ published in 2011 was associated with a meaningful reduction
lished [32]. A single-site study demonstrated a decrease of CBC in broad-spectrum antibiotic prescribing among children with

1610 • CID 2023:77 (1 December) • Ambroggio et al

pneumonia, without observed adverse outcomes. However, 13. McLaren SH, Mistry RD, Neuman MI, Florin TA, Dayan PS. Guideline adherence
in diagnostic testing and treatment of community-acquired pneumonia in chil­
rates of diagnostic testing including the use of CXR in the ED dren. Pediatr Emerg Care 2021; 37:485–93.
setting was not substantially impacted, suggesting the need 14. Feudtner C, Feinstein JA, Zhong W, Hall M, Dai D. Pediatric complex chronic
conditions classification system version 2: updated for ICD-10 and complex med­
for either stronger evidence or direct implementation efforts
ical technology dependence and transplantation. BMC Pediatr 2014; 14:199.
to change practice. Future guidelines should incorporate newer 15. Thomson J, Hall M, Ambroggio L, et al. Antibiotics for aspiration pneumonia in
evidence and explicitly consider how to enhance uptake and neurologically impaired children. J Hosp Med 2019; 14:E1–8.
16. Richardson T, Rodean J, Harris M, Berry J, Gay JC, Hall M. Development of
implementation of the recommendations to improve the care Hospitalization Resource Intensity Scores for Kids (H-RISK) and comparison
of children. across pediatric populations. J Hosp Med 2018; 13:602–8.
17. Bradley JS, Byington CL, Shah SS, et al. The management of community-acquired
pneumonia in infants and children older than 3 months of age: clinical practice
Supplementary Data
guidelines by the Pediatric Infectious Diseases Society and the Infectious
Supplementary materials are available at Clinical Infectious Diseases online. Diseases Society of America. Clin Infect Dis 2011; 53:e25–76.
Consisting of data provided by the authors to benefit the reader, the posted 18. Michelow IC, Olsen K, Lozano J, et al. Epidemiology and clinical characteristics of
materials are not copyedited and are the sole responsibility of the authors, community-acquired pneumonia in hospitalized children. Pediatrics 2004; 113:
so questions or comments should be addressed to the corresponding 701–7.

Downloaded from https://academic.oup.com/cid/article/77/11/1604/7206345 by guest on 03 December 2024

author. 19. Jain S, Williams DJ, Arnold SR, et al. Community-acquired pneumonia requiring
hospitalization among U.S. children. N Engl J Med 2015; 372:835–45.
20. Ross RK, Hersh AL, Kronman MP, et al. Impact of Infectious Diseases Society of
Notes America/Pediatric Infectious Diseases Society guidelines on treatment of
Data sharing. De-identified individual participant data will not be made community-acquired pneumonia in hospitalized children. Clin Infect Dis 2014;
available. 58:834–8.
Potential conflicts of interest. L. A.’s institution receives funding from 21. Williams DJ, Edwards KM, Self WH, et al. Antibiotic choice for children hospi­
Pfizer Inc on her behalf for an unrelated study and also reports support talized with pneumonia and adherence to national guidelines. Pediatrics 2015;
from the National Institute of Allergy and Infectious Diseases, unrelated 136:44–52.
to this work (grant numbers K01AI125413 and R21AI154239). D. J. S. re­ 22. Shaw R, Popovsky E, Abo A, et al. Improving antibiotic prescribing in the emer­
gency department for uncomplicated community-acquired pneumonia. World J
ports a grant or contract from the Agency for Healthcare Quality and
Emerg Med 2020; 11:199–205.
Research, unrelated to this work (T32HS000063-28). J. C. reports grants
23. Williams DJ, Hall M, Gerber JS, et al. Impact of a national guideline on antibiotic
or contracts unrelated to this work from Pfizer Inc. All other authors report selection for hospitalized pneumonia. Pediatrics 2017; 139:e20163231.
no potential conflicts. 24. Geanacopoulos AT, Porter JJ, Monuteaux MC, Lipsett SC, Neuman MI. Trends in
All authors have submitted the ICMJE Form for Disclosure of Potential chest radiographs for pneumonia in emergency departments. Pediatrics 2020;
Conflicts of Interest. Conflicts that the editors consider relevant to the con­ 145:e20192816.
tent of the manuscript have been disclosed. 25. Lipshaw MJ, Florin TA, Krueger S, et al. Factors associated with antibiotic pre­
scribing and outcomes for pediatric pneumonia in the emergency department.
References Pediatr Emerg Care 2021; 37:e1033–8.
26. Nelson KA, Morrow C, Wingerter SL, Bachur RG, Neuman MI. Impact of chest
1. Keren R, Luan X, Localio R, et al. Prioritization of comparative effectiveness re­
radiography on antibiotic treatment for children with suspected pneumonia.
search topics in hospital pediatrics. Arch Pediatr Adolesc Med 2012; 166:1155–64.
Pediatr Emerg Care 2016; 32:514–9.
2. Gerber JS, Kronman MP, Ross RK, et al. Identifying targets for antimicrobial
27. Florin TA, Ambroggio L, Lorenz D, et al. Development and internal validation of
stewardship in children’s hospitals. Infect Control Hosp Epidemiol 2013; 34:
a prediction model to risk stratify children with suspected community-acquired
1252–8.
pneumonia. Clin Infect Dis 2021; 73:e2713–21.
3. Florin TA, French B, Zorc JJ, Alpern ER, Shah SS. Variation in emergency depart­
28. Williams DJ, Zhu Y, Grijalva CG, et al. Predicting severe pneumonia outcomes in
ment diagnostic testing and disposition outcomes in pneumonia. Pediatrics 2013;
children. Pediatrics 2016; 138:e20161019.
132:237–44.
29. Murtagh Kurowski E, Shah SS, Thomson J, et al. Improvement methodology in­
4. Ambroggio L, Taylor JA, Tabb LP, Newschaffer CJ, Evans AA, Shah SS.
creases guideline recommended blood cultures in children with pneumonia.
Comparative effectiveness of empiric β-lactam monotherapy and
Pediatrics 2015; 135:e1052–9.
beta-lactam-macrolide combination therapy in children hospitalized with
30. Fritz CQ, Edwards KM, Self WH, et al. Prevalence, risk factors, and outcomes of
community-acquired pneumonia. J Pediatr 2012; 161:1097–103.
5. Balas EA, Boren SA. Managing clinical knowledge for health care improvement. bacteremic pneumonia in children. Pediatrics 2019; 144:e20183090.
Yearb Med Inform 2000; 1:65–70. 31. Neuman MI, Hall M, Lipsett SC, et al. Utility of blood culture among children
6. Westfall JM, Mold J, Fagnan L. Practice-based research—“blue highways” on the hospitalized with community-acquired pneumonia. Pediatrics 2017; 140:
NIH roadmap. JAMA 2007; 297:403–6. e20171013.
7. Davies B, Edwards N, Ploeg J, Virani T. Insights about the process and impact of 32. Florin TA, Byczkowski T, Gerber JS, Ruddy R, Kuppermann N. Diagnostic testing
implementing nursing guidelines on delivery of care in hospitals and community and antibiotic use in young children with community-acquired pneumonia in the
settings. BMC Health Serv Res 2008; 8:29. United States, 2008–2015. J Pediatric Infect Dis Soc 2020; 9:248–52.
8. Prior M, Guerin M, Grimmer-Somers K. The effectiveness of clinical guideline 33. Rogers AJ, Lye PS, Ciener DA, Ren B, Kuhn EM, Morrison AK. Using quality im­
implementation strategies—a synthesis of systematic review findings. J Eval provement to change testing practices for community-acquired pneumonia.
Clin Pract 2008; 14:888–97. Pediatr Qual Saf 2018; 3:e105.
9. Heine D, Cochran C, Moore M, Titus MO, Lintzenich Andrews A. The prevalence 34. Cotter JM, Hardee I, Moss A, Dempsey A, Ambroggio L. Procalcitonin use: var­
of bacteremia in pediatric patients with community-acquired pneumonia: guide­ iation across hospitals and trends over time. Hosp Pediatr 2021; doi:10.1542/
lines to reduce the frequency of obtaining blood cultures. Hospital Pediatrics hpeds.2021-005992.
2013; 3:92–6. 35. Baer G, Baumann P, Buettcher M, et al. Procalcitonin guidance to reduce antibi­
10. Ambroggio L, Mangeot C, Murtagh Kurowski E, et al. Guideline adoption for otic treatment of lower respiratory tract infection in children and adolescents
community-acquired pneumonia in the outpatient setting. Pediatrics 2018; 142: (ProPAED): a randomized controlled trial. PLoS One 2013; 8:e68419.
e20180331. 36. Stockmann C, Ampofo K, Killpack J, et al. Procalcitonin accurately identifies hos­
11. Ambroggio L, Thomson J, Murtagh Kurowski E, et al. Quality improvement pitalized children with low risk of bacterial community-acquired pneumonia. J
methods increase appropriate antibiotic prescribing for childhood pneumonia. Pediatric Infect Dis Soc 2018; 7:46–53.
Pediatrics 2013; 131:e1623–31. 37. Florin TA, Ambroggio L, Brokamp C, et al. Biomarkers and disease severity in
12. Parikh K, Hall M, Blaschke AJ, et al. Aggregate and hospital-level impact of na­ children with community-acquired pneumonia. Pediatrics 2020; 145:e20193728.
tional guidelines on diagnostic resource utilization for children with pneumonia 38. Neuman MI, Hall M, Hersh AL, et al. Influence of hospital guidelines on manage­
at children’s hospitals. J Hosp Med 2016; 11:317–23. ment of children hospitalized with pneumonia. Pediatrics 2012; 130:e823–30.

Guideline for Pediatric Pneumonia • CID 2023:77 (1 December) • 1611