TM

FEBRUARY 2018 • VOL. 39 NO. 2 • PAGES 13-24 aapgrandrounds.org

Febrile Seizure Risk and

Vaccination

Antibiotics, S aureus
Colonization, and Recurrent
Skin Infections

Prenatal Acetaminophen
Exposure and Risk of ADHD in
Children

Third MMR Dose Effective in

Controlling Mumps Outbreak
 Table of Contents
NEUROLOGY
Editorial Board EDITORIAL BOARD
p15 Febrile Seizure Risk and Vaccination
Rebecca Brady, Cincinnati, OH
EDITOR Susan L. Bratton, Salt Lake City, UT
James A. Taylor, Seattle, WA INFECTIOUS DISEASES
Daniel Doherty, Seattle, WA
DEPUTY EDITOR Benjamin Doolittle, New Haven, CT p16 
Antibiotics, S aureus Colonization, and Recurrent
Leslie L. Barton, Tucson, AZ Mike Dubik, Manchester, CT Skin Infections
Patricia Fechner, Seattle, WA
ASSOCIATE EDITOR
William L. Hennrikus, Hershey, PA DEVELOPMENTAL/BEHAVIORAL
Douglas J. Opel, Seattle, WA
Gloria Higgins, Columbus, OH p17 
Prenatal Acetaminophen Exposure and Risk of
EVIDENCE eMENDED EDITOR Mary-Jane Staba Hogan,
Bernhard (Bud) Wiedermann, New Haven, CT ADHD in Children
Washington, DC Daniel Lesser, San Diego, CA
GENERAL PEDIATRICS
CME QUESTION EDITOR Jonathan Mintzer, Stony Brook, NY
Robert Wittler, Wichita, KS Carrie Phillipi, Portland, OR p18 
Third MMR Dose Effective in Controlling Mumps Outbreak
Philip Rosenthal, San Francisco, CA
Pamela Singer, New Hyde Park, NY
INFECTIOUS DISEASES
David Spar, Cincinnati, OH p19 
Developmental Outcomes in Asymptomatic Congenital
Michelle Stevenson, Louisville, KY CMV Infection
David Urion, Boston, MA
Jeffrey Winer, Jacksonville, FL GENERAL PEDIATRICS
Charlene Wong, Philadelphia, PA
p20 
Cardiovascular Risk Factors Influence Carotid
Intima-Media Thickness
Mission: To provide pediatricians with timely synopses and
critiques of important new studies relevant to pediatric practice, HOSPITAL MEDICINE
reviewing methodology, significance, and practical impact, as p21 Rethinking Macrolide Use in Inpatient Pneumonia
part of ongoing CME activity.
NEPHROLOGY
p22 
Outcomes in Rituximab-Treated Nephrotic Syndrome
AAP Grand RoundsTM is published monthly by the American
Academy of Pediatrics, 345 Park Blvd., Itasca, IL 60143. Copyright NEONATOLOGY
©2018 American Academy of Pediatrics. All rights reserved.
p23 
Neonatal Therapeutic Hypothermia: Timing
Postmaster: Send address changes to AAP Grand Rounds,
American Academy of Pediatrics, Customer Service Center, and Effects on Outcomes
345 Park Blvd., Itasca, IL 60143.
Visit www.aappublications.org/blog-posts to read an Evidence
ISSN Numbers eMended post about this article appearing this month.
Print: 1099-6605
Online: 1556-326X

Customer Service and Renewals: (866) 843-2271 CME INFORMATION

Email Address: grandrounds@aap.org AAP Grand Rounds™ is an educational publication.
Subscription Rates: AAP member $102/year; non-member $127/ The American Academy of Pediatrics is accredited by the Accreditation Council for Continuing Medical
year (12 issues), including up to 18 AMA PRA Category 1 credit(s)™. Education (ACCME) to provide continuing medical education for physicians.

The American Academy of Pediatrics designates this enduring material for a maximum of 18 AMA PRA
Managing Editor: Kate Larson
Category 1 Credit(s)™. Physicians should claim only the credit commensurate with the extent of their
Publications Editor: Mark Plemmons participation in the activity.
Production: Michael Hayes This activity is acceptable for up to 18 AAP credits. These credits can be applied toward the AAP CME/CPD
Editorial Associate: Carol Frost Award available to Fellows and Candidate Members of the American Academy of Pediatrics.

The American Academy of Physician Assistants (AAPA) accepts certificates of participation for educational
activities certified for AMA PRA Category 1 Credits™ from organizations accredited by ACCME. Physicians
assistants may receive a maximum of 18 hours of Category 1 credit for completing this program.

The AAP Policy on Disclosure of Financial Relationships and Resolution of Conflicts of Interest for AAP CME Activi- This program is approved for 18 NAPNAP CE contact hours; pharmacology (Rx) contact hours to be
ties is designed to ensure quality, objective, balanced, and scientifically rigorous AAP CME activities by identifying determined per the National Association of Pediatric Nurse Practitioners Continuing Education
and resolving all potential conflicts of interest prior to the confirmation of service of those in a position to influ- Guidelines.
ence and/or control CME content.

All individuals in a position to influence and/or control the content of AAP CME activities, including editorial board
members, authors, and staff, are required to disclose to the AAP and subsequently to learners that the individual
either has no relevant financial relationship or any financial relationships with the manufacturer(s) of any com-
mercial product(s) and/or provider(s) of commercial services discussed in CME activities. None of the editors,
authors, or staff had any relevant financial relationships to disclose for this issue of AAP Grand Rounds unless
noted on the article or below. The AAP has taken steps to resolve any potential conflicts of interest.
Philip Rosenthal, MD (Editorial Board Member) disclosed Research Grants with Bristol-Myers Squibb and Roche.
Robert R. Wittler, MD (Editorial Board Member) disclosed a Speaker’s Bureau with Sanofi Vaccines.
This is a scientific publication designed to present updates and opinion to health care professionals. It
does not provide medical advice for any individual case, and is not intended for the layman.
New minimum performance level requirements:
Per the 2010 revision of the American Medical Association (AMA) Physician’s Recognition Award (PRA) and
credit system, a minimum performance level must be established on enduring material and journal-
based CME activities that are certified for AMA PRA Category 1 Credit™. In order to successfully complete
2017 AAP Grand Rounds quizzes for AMA PRA Category 1 Credit ™, learners must demonstrate a minimum
performance level of 70% or higher on this assessment, which measures achievement of the educational
purpose and/or objectives of this activity.

Starting with 2012 AAP Grand Rounds, AMA PRA Category 1 Credit™ can be claimed only if 70% or more
of the questions are answered correctly. If you score less than 70% on the assessment, you will be given
additional opportunities to answer questions until an overall 70% or greater score is achieved.

No product-specific advertising of any type, or links to product web sites, appear in this educational
activity.

14 aapgrandrounds.org
NEUROLOGY
Febrile Seizure Risk and Vaccination
Source: Duffy J, Hambidge SJ, Jackson LA, et al. Febrile seizure risk after
vaccination in children one to five months of age. Pediatr Neurol. 2017
Nov;76:72–78. doi: 10.1016/j.pediatrneurol.2017.08.005
Investigators from multiple institutions conducted a retrospective
cohort study to estimate the risk of febrile seizures in the day
following vaccination. The study population included all children
aged 1–5 months who belonged to a Vaccine Safety Datalink health
care organization. Data of all vaccines received by the study pop-
ulation from 2006–2011 were abstracted from the medical record.
Cases of febrile seizures were ascertained by identifying children
over the study period who had an ED visit or inpatient admission
that was associated with an ICD-9 code for convulsion. Identified
children were excluded if their visit or admission occurred at a
time other than on the day of or day after the administration of
one or more vaccinations. Medical charts of all identified children
were reviewed to determine if they experienced a febrile seizure,
defined as a seizure in a child with a fever (≥38°C) within the prior
24 hours, excluding patients with CNS infection or a history of a
febrile seizure. Febrile seizures were considered probable if the
child received a febrile seizure diagnosis by the treating physician
and its clinical characteristics were described.
All probable febrile seizure cases were included in analyses. The
investigators compared the incidence of febrile seizure within the
first day of vaccination (the risk interval) with a control interval,
defined as 14–20 days post-vaccination, to calculate the incidence
rate ratio (IRR). This control interval was chosen because it repre-
sented a period during which inactivated or live-attenuated vac-
cines should not affect the risk of fever or febrile seizures. The
investigators also calculated the excess number of cases that oc-
curred due to exposure to vaccination (attributable risk) using the
incidence rates during the risk and control intervals.
There were 585,342 vaccination visits over the study period and 15
febrile seizure cases. Children with febrile seizures had received a
median of 4 vaccines (range 2–6), with the most common combi-
nation of vaccines received simultaneously being rotavirus, pneu-
mococcal conjugate, Haemophilus influenzae type b, and diphthe-
ria-tetanus-acellular pertussis-hepatitis B-inactivated poliovirus
vaccines. The IRR for febrile seizures was 23 (95% CI: 5.12, 100.8)
per 100,000 children vaccinated, with an attributable risk of 3.92
(95% CI: 1.68, 6.17) per 100,000 children vaccinated.
The investigators conclude that although there is a large relative
risk of febrile seizure on the day of or day after vaccination in
children age 1–5 months, the absolute risk is small.
COMMENTARY BY
David K. Urion, MD, FAAN, Child Neurology and Neurodevelopmental
Disabilities, Boston Children’s Hospital, Boston, MA
Dr Urion has disclosed no financial relationship relevant to this commentary. This com-
mentary does not contain a discussion of an unapproved/investigative use of a commer-
cial product/device.
While the risk of autism after vaccination is probably the larg-
est source of parental concern a pediatrician faces in the space
around vaccinations—despite substantive research, which has not
• FEBRUARY 2018
TM
been able to show a relationship between vaccination and au-
tism—the occurrence of seizures around the time of vaccination
is probably a close second. The current study, multicentered and
well-powered, examined the latter issue.
The current investigators did not, and by its design could not,
include families who stayed home, with or without such counsel
from their pediatrician. Thus, the study may represent a lower
number of febrile seizures after vaccination than actually oc-
curred. Data from other assessments of febrile seizures, with or
without proximate vaccination, suggest that this number is a rel-
atively small percentage of such cases in infants <6 months old.1,2
While the relative risk of febrile seizures after vaccinations was
high, the absolute risk was quite low. It is also important to note
that the AAP defines febrile seizures as occurring in children >6
months old; thus, the seizures reported in the current study do
not technically meet this definition.3
Most of us in daily life probably use the concept of “risk” more
aligned with absolute risk rather than relative risk.4,5 For example,
the relative risk of acquiring a norovirus infection from eating food
served in a common bowl at a gathering may be higher during
certain months of the year compared to others—for instance, July
versus November. The absolute risk, however, remains low in both
instances, and we don’t forego the condiment either time.
This increased relative risk may lead to anticipatory manage-
ment of infants with known individual or family histories of fe-
brile seizures.
Bottom Line: Vaccinations in infants <6 months of age have an
increased relative risk of associated febrile seizure, but low ab-
solute risk.
EDITORS’ NOTE
There are few things more frightening to a parent of a young child
than witnessing a febrile seizure. Although pediatricians may clas-
sify these events as “benign,” they are anything but that to par-
ents. The results of this study present a dilemma when discussing
the risks of vaccines to parents. Although the increased risk is very
small (approximately 4 excess seizures per 100,000 vaccinated in-
fants), it’s not zero.
References
1. Nilsson G, et al. Neuropediatrics. 2016;47(06):368–373; doi: 10.1055/s-0036-
1586731
2. Patel N, et al. BMJ. 2015;351:h4240; doi: 10.1136/bmj.h4240
3. Subcommittee on Febrile Seizures. Pediatrics. 2011;127(2):389–394; doi:
10.1542/peds.2010-3318
4. Breakwell G. The Psychology of Risk. 2nd ed. Cambridge, UK: Cambridge
University Press; 2014:15–23
5. Slovic P. The Perception of Risk. New York, NY: Routledge Press; 2013:27–32
15
INFECTIOUS DISEASES

Antibiotics, S aureus Colonization,

and Recurrent Skin Infections
Source: Hogan PG, Rodriguez M, Spenner AM, et al. Impact of systemic
antibiotics on Staphylococcus aureus colonization and recurrent skin
infection [published online ahead of print August 24, 2017. Clin Infect
Dis. doi: 10.1093/cid/cix754
Investigators from Washington University, St. Louis, MO, and
Southern Illinois University, Springfield, IL, assessed the effect of
systemic antibiotics prescribed at the time of incision and drain-
age (I&D) for Staphylococcus aureus skin and soft tissue infections
(SSTI) on rates of future colonization and recurrent infections in
children. The investigators combined data from a series of pro-
spective studies, conducted in patients <21 years old with S aureus
SSTI since 2008. At the time of I&D (baseline), swabs of the anteri-
or nares, inguinal folds, and axillae of study participants were ob-
tained. For the current study, only children who had S aureus col-
onization at one or more of these sites were included. Additional
data collected at baseline included demographics, decolonization
measures (intranasal mupirocin, dilute bleach, or chlorhexidine
body washes), whether the infecting S aureus was methicillin-sen-
sitive (MSSA) or resistant (MRSA), and whether the child received
antibiotics recommended by the Infectious Diseases Society of
America (eg, clindamycin or trimethoprim-sulfamethoxazole
[TMP-SMZ]).1 Repeat swabs to determine colonization status were
done up to 3 months after baseline. In addition, follow-up data
were collected for up to a year after baseline to assess for re-
current SSTI. Cox regression was used to assess the association
between use of a guideline-recommended antibiotic at baseline
on future colonization or recurrent SSTI, after adjusting for age,
race, SSTI isolate (MRSA or MSSA), and number of anatomic sites
colonized at baseline.
COMMENTARY BY
Mike Dubik, MD, FAAP, Portsmouth, VA
Dr Dubik has disclosed no financial relationship relevant to this commentary. This com-
mentary does not contain a discussion of an unapproved/investigative use of a commer-
cial product/device.
The results of this study join others supporting the use of empiric
systemic antibiotics as an adjunct to I&D.2–5 I&D is the primary
treatment for simple abscesses with the addition of empiric anti-
biotics recommended for complex abscesses or purulent cellulitis.1
As S aureus is the most common cause of SSTIs, the recommended
empiric antibiotics include TMP-SMX, clindamycin, a tetracycline,
or linezolid.1 Although not recommended for decolonization per
se, systemic antibiotics, and clindamycin in particular, appear to
offer some advantages.
It might seem rather intuitive that eradication of S aureus colo-
nization would reduce the incidence of SSTI recurrence. However,
this apparent benefit must be balanced against the cost, potential
side effects, and the potential development of resistance. (See re-
lated article in AAP Grand Rounds February 2015;33[2]:246).
Bottom Line: Systemic antibiotics in combination with I&D de-
crease colonization and recurrence of S aureus SSTIs.
EDITORS’ NOTE
Attempts to eradicate S aureus colonization, including chlor-
hexidine body washes and intranasal mupirocin, have met with
variable success.1 This important investigation demonstrates the
superior efficacy of systemic clindamycin as an adjunct to I&D of
purulent SSTIs. However, note that S aureus eradication was nei-
ther complete (56%) nor its duration beyond 3 months defined.

Data were analyzed on 383 children with a median age of 3 years. References
Overall, 81% of baseline SSTI were caused by MRSA. Baseline col- 1. Liu C, et al. Clin Infect Dis. 2011;52(3):285–292; doi: 10.1093/cid/cir034
2. Duong M, et al. Ann Emerg Med. 2010;55(5):401–407; doi: 10.1016/j.
onization was most often detected in inguinal folds, followed by
annemergmed.2009.03.01
anterior nares; 46% of study participants were colonized at mul-
3. Talan DA, et al. N Engl J Med. 2016;374:823–832; doi: 10.1056/NEJMoa1507476
tiple sites. At the time of I&D, 93% of children were prescribed
4. Schmitz GR, et al. Ann Emerg Med. 2010;56(3):283–270; doi: 10.1016/j.
a guideline-recommended antibiotic (clindamycin or TMP-SMZ). annemergmed.2010.03.002
Follow-up colonization data were collected on 357 participants. 5. Holmes I, et al. J Pediatr. 2016;169:128–134; doi: 10.1016/j.jpeds.2015.10.044
The risk of S aureus colonization at follow-up was significantly 6. Stevens DL, et al. Clin Infect Dis. 2014;59(2):e10-52; doi: 10.1093/cid/ciu444
lower for children receiving recommended antibiotics compared
to those receiving non-recommended or no antibiotics at baseline
(colonization rates 48% and 77%, respectively, adjusted hazard ra-
tio [aHR] 0.49, 95% CI 0.30, 0.79). Children receiving clindamycin
at baseline were significantly less likely to be colonized at fol-
low-up than those receiving TMP-SMZ (44% and 57%, respectively;
AAP Journal CME
P = .03). Recurrent SSTI occurred in 40% of children receiving a You can complete and claim credit for all of your
recommended antibiotic at baseline vs 64% of those receiving a quizzes online. Visit www.aapgrandrounds.org.
non-recommended or no antibiotic (aHR = 0.57, 95% CI 0.34, 0.94).
The authors conclude that use of clindamycin or TMP-SMZ at the
time of I&D in children with an SSTI caused by S aureus reduces
the risk of future colonization and rates of recurrent SSTI.

16 aapgrandrounds.org
DEVELOPMENTAL/BEHAVIORAL
Prenatal Acetaminophen Exposure
and Risk of ADHD in Children
Source: Ystrom E, Gustavson K, Brandlistuen RE, et al. Prenatal exposure
to acetaminophen and risk of ADHD. Pediatrics. 2017;140(5):e20163840;
doi:10.1542/peds.2016-3840
An international group of investigators assessed the association
between prenatal maternal acetaminophen use and ADHD in chil-
dren. The investigators focused on controlling for potential con-
founding variables, such as parental ADHD symptoms and indica-
tions for acetaminophen use and pre-pregnancy use. Data from
the Norwegian Mother and Child Cohort Study (MoBa) for children
born between 1999 and 2009 were abstracted to collect informa-
tion on maternal acetaminophen use beginning 6 months prior to
conception and 6 months postpartum. As part of MoBa, mothers
completed questionnaires at multiple points during and following
pregnancy. Data on acetaminophen use were stratified by peri-
od (including pre-pregnancy and during each trimester) and by
indication (fever and infections, pain conditions, or unspecified);
the number of days of acetaminophen use was also recorded. In-
formation regarding anxiety and depression symptoms and other
medication use (including tobacco and alcohol) during pregnancy
was also recorded. Fathers completed questionnaires that includ-
ed pre-pregnancy acetaminophen use, and both parents reported
ADHD symptoms using a standardized scale.
MoBa data were linked to the Norwegian Patient Registry to deter-
mine if study children had received an ADHD diagnosis between
2008 and 2014. Cox regression was used to estimate hazard ratios
(HR) for associations between acetaminophen use and ADHD in
offspring.
The analysis included 112,973 children, of whom 2,246 had ever
been diagnosed with ADHD. After controlling for multiple con-
founders, including pre-pregnancy use, ADHD symptoms, mater-
nal anxiety and depression symptoms, and other medication use,
the risk of ADHD was significantly higher in children of mothers
who used acetaminophen during pregnancy (HR = 1.12, 95% CI 1.02,
1.24). The HRs for use during 1, 2, or 3 trimesters were 1.07 (95%
CI 0.96, 1.29), 1.22 (95% CI 1.07, 1.38), and 1.27 (95% CI 0.99, 1.63),
respectively. Compared to those who used no acetaminophen,
mothers who used 1–7 days were at decreased risk of having a
child with ADHD (HR = 0.90, 95% CI 0.81, 1.00), while those who
used >29 days were at significantly increased risk (HR = 2.20, 95%
CI 1.50, 2.34). The magnitude of risk did not vary across indications
for use: HRs of 2.40 (95% CI 0.34, 16.78), 2.56 (95% CI 1.54, 4.25), and
2.13 (95% CI 0.88, 5.15) for fever and infections, pain conditions, and
unspecified indication, respectively. Paternal pre-pregnancy acet-
aminophen use was also associated with increased risk of ADHD
in offspring (HR = 1.27, 95% CI 1.08, 1.49), and increasing usage was
associated with higher risk. However, maternal pre-pregnancy use
was not statistically associated with an increased risk of ADHD.
The authors conclude that long-term acetaminophen use during
pregnancy by mothers, and pre-pregnancy use by fathers, is asso-
ciated with increased risk of ADHD in their children.
• FEBRUARY 2018
TM
COMMENTARY BY
Dan Doherty, MD, PhD, FAAP, Genetic and Developmental Medicine,
Pediatrics, University of Washington and Seattle Children’s Hospital,
Seattle, WA
Dr Doherty has disclosed no financial relationship relevant to this commentary. This com-
mentary does not contain a discussion of an unapproved/investigative use of a commer-
cial product/device.
Determining the long-term risks of medications during pregnancy
to outcomes in children is challenging, because large, prospective
studies relating standardized outcome measures to pregnancy
exposures are rare. Such studies cannot collect data on all expo-
sures and outcomes, and data on paternal health and exposures
are particularly sparse. Nonetheless, the current study is a good
example of how large datasets can be leveraged to evaluate as-
sociations between common exposures, such as acetaminophen,
and common outcomes, such as ADHD.
Prior to this study, the Society for Maternal-Fetal Medicine pub-
lished a statement acknowledging that observational studies had
identified associations between acetaminophen exposure during
pregnancy and adverse child outcomes, but that the existing stud-
ies should be interpreted with caution due to methodological is-
sues, including recall bias, lack of dose information, residual con-
founding, and multiple hypothesis testing. They concluded that
“acetaminophen be considered a reasonable and appropriate
medication choice for the treatment of pain and/or fever during
pregnancy.”1 The current investigators found that acetaminophen
use during, but not prior to, pregnancy is associated with ADHD
diagnosis, decreasing the likelihood that the association is due
to genetic or other possible confounders. In addition, longer du-
ration of use was associated with higher risk. Appropriately, the
investigators note that they cannot determine whether the asso-
ciation is causal. Their finding that paternal acetaminophen use is
also associated with ADHD diagnosis indicates that there may be
more than one link between acetaminophen and ADHD risk.
Even before the current data, Olsen and Liew2 argued for changing
the recommendations about acetaminophen use during pregnan-
cy, but additional data will be required to reach a final conclusion.
Regardless, the results of the current study support the possible
risks of acetaminophen use (particularly prolonged use) during
pregnancy, which should be weighed against the potentially ben-
eficial analgesic and antipyretic effects.
Bottom Line: Mounting observational evidence indicates that ac-
etaminophen exposure during pregnancy may be related to ADHD
risk, although no definitive causal link has been proven.
References
1. Society for Maternal-Fetal Medicine (SMFM) Publications Committee. Am J
Obstet Gynecol. 2017;216(3):B14–B15; doi: 10.1016/j.ajog.2017.01.021
2. Olsen J, et al. Expert Opin Drug Saf. 2017;16(12):1395–1398; doi:
10.1080/14740338.2017.1384812
17
GENERAL PEDIATRICS
Third MMR Dose Effective in
Controlling Mumps Outbreak
Source: Cardemil CV, Dahl RM, James L, et al. Effectiveness of a third
dose of MMR vaccine for mumps outbreak control. N Engl J Med. 2017
;377(10):947–956; doi:10.1056/NEJMoa1703309
Investigators from multiple institutions conducted an observa-
tional study to assess the effectiveness of a third dose of MMR in
the control of a mumps outbreak. The mumps outbreak occurred
at the University of Iowa from August 2015–May 2016. The Univer-
sity of Iowa requires all students upon enrollment to submit pro-
vider-signed vaccination records with verification of the dates of
receipt of 2 doses of MMR. During the mumps outbreak, a vacci-
nation campaign was launched in which a third dose of MMR was
offered to students 18–24 years old.
Study participants were students enrolled at the university for the
2015–2016 academic year. The primary exposure was vaccination
dose status, categorized as 0, 2, or 3 doses as determined by re-
view of university vaccination records. The primary outcome was
contraction of mumps (yes/no), using an accepted definition of
probable or confirmed mumps. Demographics were obtained from
university records.
Investigators compared attack rates of mumps according to dose
status and years since receipt of the second dose of MMR. The in-
cremental vaccine effectiveness of the third MMR dose compared
to 2 doses was estimated using multivariable regression models.
The investigators also assessed vaccine effectiveness of 2 versus
no MMR doses stratified by years since receipt of the second dose.
There were 20,496 students included in analysis. Most students
(98.1%) had received at least 2 MMR doses before the outbreak.
There were 4,783 students who received a third dose during the
outbreak.
The overall attack rate was 12.6 per 1,000. The attack rate was sig-
nificantly lower among those with 3 versus 2 MMR doses (6.7 vs
14.5 per 1,000; P<.001) and varied significantly by years since re-
ceipt of the second dose: it increased from 1.6 per 1,000 if the sec-
ond dose was received within 2 years to 17.6 per 1,000 if received
16–23 years earlier. Receipt of a third MMR dose reduced the risk
of mumps by 78.1% compared to receipt of 2 doses (95% CI 61%,
88%). The effectiveness of 2 (vs 0) MMR doses varied by year since
receipt of the second dose: 89.4% if received <13 years earlier and
31.8% if received ≥13 years earlier (P=.002).
The investigators conclude that waning immunity likely contrib-
uted to the mumps outbreak, and a third MMR dose is effective in
outbreak control.
COMMENTARY BY
Carrie Phillipi, MD, PhD, FAAP, Oregon Health & Science University,
Portland, Oregon
Dr Phillipi has disclosed no financial relationship relevant to this commentary. This com-
mentary does not contain a discussion of an unapproved/investigative use of a commer-
cial product/device.
Mumps is characterized by fever, respiratory symptoms, and pa-
rotid gland swelling. Highly contagious, mumps is spread by respi-
ratory droplets even before apparent symptoms. Affected contacts
18
exhibit symptoms 2–3 weeks after exposure, and although mumps
is typically a mild disease, serious complications, including deaf-
ness, encephalitis, and orchitis, are not rare. Vaccination with a
2-dose series of MMR has been extremely effective in reducing
cases of mumps, but outbreaks still occur in groups of unvaccinat-
ed individuals and on college campuses.1,2 Although outbreaks are
larger in areas with lower vaccine coverage rates, many mumps
outbreaks have occurred on college campuses where 2-dose cov-
erage with MMR has exceeded 90%.3,4 (See AAP Grand Rounds June
2017;37[6]:64.5)
The University of Iowa enrolls nearly 22,000 undergraduates, many
of whom live on campus in dormitory settings. Because students
submit vaccination records to comply with a 2-dose MMR vaccine re-
quirement to enroll in classes, University of Iowa students provided
an ideal population in which to study the epidemiology of a mumps
outbreak and the effectiveness of a third dose of MMR vaccine. Stu-
dents who received only 2 doses were more likely to suffer from
mumps than those who received a third dose, and the risk of mumps
increased with time from receipt of the second dose of MMR. The
strength of these data convinced the Advisory Committee on Immu-
nization Practices to unanimously recommend a third dose of MMR
vaccine for individuals at risk from a mumps outbreak.6
This study was observational; receipt of the third dose could be
due to perceived risk and/or other reasons for seeking health
care, though this would more likely underestimate vaccine ef-
fectiveness. Students may have also sought care apart from the
university-sponsored health clinics. The authors conducted ad-
ditional sensitivity analyses to determine if vaccine effectiveness
varied by student age and were unable to find evidence for this.
Bottom Line: Waning immunity likely contributes to mumps out-
breaks on college campuses. A third dose of MMR vaccine is help-
ful in controlling outbreaks.
EDITORS’ NOTE
The CDC MMR vaccination recommendation for the general pop-
ulation remains unchanged. The 2-dose schedule has been esti-
mated to be 88% effective in preventing mumps outside of high-
risk (ie, intense exposure) settings.7
References
1. Mumps. Centers for Disease Control and Prevention. Available at www.cdc.
gov/mumps/hcp.html#clinical
2. The History of Vaccines. Mumps. Available at www.historyofvaccines.org/
content/articles/mumps
3. Albertson JP, et al. MMWR Morb Mortal Wkly Rep. 2016;65(29):731–734; doi:
10.15585/mmwr.mm6529a2
4. Centers for Disease Control and Prevention. MMWR Morb Mortal Wkly Rep.
2012 Dec 7;61(48):986–989
5. Patel LN, et al. Clin Infect Dis. 2017;15;64(4):408–412; doi: 10.1093/cid/ciw762
6. Jenco M. ACIP: Give 3rd mumps vaccine dose during outbreaks. AAP News,
Oct. 26, 2017 Available at: http://www.aappublications.org/news/2017/10/26/
Mumps102617
7. American Academy of Pediatrics. Mumps. In: Kimberlin DW (ed) Red Book:
2015 Report of the Committee on Infectious Diseases. 30th ed. Elk Grove
Village, IL: American Academy of Pediatrics; 2015:564–568
aapgrandrounds.org
INFECTIOUS DISEASES
Developmental Outcomes in
Asymptomatic Congenital CMV Infection
Source: Lopez AS, Lanzieri TM, Claussen AH, et al. Intelligence and ac-
ademic achievement with asymptomatic congenital cytomegalovirus
infection. Pediatrics. 2017 Nov;140(5). e20171517. doi: 10.1542/peds.2017-
1517
Investigators from the CDC and Texas Children’s Hospital, Houston,
TX, conducted a longitudinal case control study to assess intelligence
and academic achievement in children with asymptomatic cytomeg-
alovirus (CMV) infections. Cases were children born between 1982
and 1992 at Women’s Hospital of Texas who were found to be positive
on newborn screening for CMV (based on a urine culture within the
first 3 weeks of life). Only newborns with no signs of congenital CMV
disease (ie, no purpura, petechiae, jaundice, hepatosplenomegaly,
microcephaly, elevated liver enzymes, hemolytic anemia, or throm-
bocytopenia) were included as cases. Controls were children who
screened negative for CMV at birth and matched with a case on date
of birth (within 6 days). Study participants were assessed with stan-
dardized, age-appropriate, full-scale intelligence, receptive and ex-
pressive language, and academic achievement tests at multiple peri-
ods during their childhood, including infancy, preschool, elementary
school, middle school, and high school. All study children had audi-
ological testing to detect sensorineural hearing loss (SNHL) before
the age of 2 years. The investigators used “growth curve” modeling to
account for trends in scores over time and to adjust for potentially
confounding variables such as maternal education levels. Compar-
isons were made between cases with normal hearing and controls
and between cases with SNHL and control children.
During the study period, 32,543 newborns were screened for CMV; 135
(0.4%) tested positive. Study data were collected on 78 cases with
normal hearing, 11 cases with SNHL, and 40 controls (no control child
was diagnosed with SNHL before the age of 2 years). Mean (95% CI)
full-scale intelligence test scores at 5 and 18 years of age, respective-
ly, were 108 (105, 110) and 111 (108, 114) for cases with normal hearing,
101 (95, 106) and 104 (98, 110) for cases with SNHL, and 108 (104, 111)
and 111 (107, 114) for controls. Overall, there were no significant differ-
ences between cases with normal hearing and controls (P=.96), but
cases with SNHL had scores that were 7 points lower than controls
(P<.05). There were no significant differences between cases with
normal hearing and controls for expressive and receptive language
scores (P=.36); cases with SNHL had significantly lower receptive,
but not expressive, language scores than controls. Math and reading
scores were not significantly different across all 3 groups.
The authors conclude that infants with asymptomatic CMV infection
with normal hearing by age 2 years do not appear to have differ-
ences in intelligence, language development, or academic achieve-
ment during childhood compared with uninfected children.
COMMENTARY BY
Rebecca Brady, MD, FAAP, Infectious Diseases, Cincinnati Children’s
Hospital, Cincinnati, OH
Dr Brady has disclosed no financial relationship relevant to this commentary. This com-
mentary does not contain a discussion of an unapproved/investigative use of a commer-
cial product/device.
• FEBRUARY 2018
TM
Testing for congenital CMV infection is not part of routine new-
born screening in the United States.1 However, Utah2 and Connecti-
cut3 have enacted legislation mandating that all infants who fail
their newborn hearing screen be tested for CMV infection within
the first 3 weeks of life unless the parent declines the test. The
goal of this hearing-targeted congenital CMV screening program
is to provide early interventions for those with SNHL to improve
long-term language outcomes.
The results from the first 2 years of the Utah program were recent-
ly published.2 Sixty-two percent of infants who never passed their
newborn hearing screen were tested for CMV infection. Fourteen
(6%) of 234 infants tested within the first 3 weeks of life were CMV
positive; 6 (42.9%) had hearing loss (3 SNHL, 2 conductive hearing
loss, and 1 mixed).
Hearing-targeted congenital CMV screening programs raise import-
ant concerns.4 First, although monitoring of children with congen-
ital CMV infection for SNHL is endorsed by the Joint Committee on
Infant Hearing,5 guidance on the performance of this monitoring is
not provided. Second, the benefit and safety of antiviral therapy
(valganciclovir) in infants with asymptomatic congenital CMV infec-
tion, including those with isolated SNHL, is not currently known.4
Third, the diagnosis of congenital CMV infection raises questions
about long-term neurodevelopmental outcomes. The results of
this study provide some reassurance that infants with asymptom-
atic congenital CMV infection who maintain normal hearing by age
2 years have similar intelligence, language development, and aca-
demic achievement during childhood and adolescence compared
to uninfected children. However, the number of infants was small,
and most were born to women of medium or high socioeconomic
status.6 Larger studies in more diverse populations are needed.
Bottom Line: Infants diagnosed with congenital CMV infection re-
quire follow-up hearing tests. Those who maintain normal hearing
by age 2 years may have good neurodevelopmental outcomes.
EDITORS’ NOTE
While the findings of the current investigation are encouraging,
it is important to note that most infants with congenital CMV in-
fection are probably not detected by hearing-targeted screen-
ing.4 Universal screening for congenital CMV infection, however, is
fraught with multiple challenges; its feasibility and affordability
are being evaluated.
References
1. Grosse SD, et al. J Clin Virol. 2009;46S:S32–S36; doi: 10.1016/j.jcv.2009.08.019
2. Diener ML, et al. Pediatrics. 2017;139(2):e20160789; doi: 10.1542/peds.2016-0789
3. Connecticut Committee on Public Health. An Act Concerning Newborn Screen-
ing for Cytomegalovirus and Establishing a Public Education Program for
Cytomegalovirus. H.B. 5147 (2014)
4. Grosse SD, et al. Pediatrics. 2017;139(2):e20163837; doi: 10.1542/peds.2016-3837
5. American Academy of Pediatrics, Joint Committee on Infant Hearing. Pediat-
rics. 2007;120(4):898–921; doi: 10.1542/peds.2007-2333
6. Boppana SB, et al. Pediatrics. 2017;140(5):e20172526; doi: 10.1542/peds.2017-
2526
19
GENERAL PEDIATRICS
Cardiovascular Risk Factors Influence
Carotid Intima-Media Thickness
Source: White D, Place R, Michael T, et al. The relationship between cor-
onary artery disease risk factors and carotid intima-media thickness in
children. J Pediatr. 2017;190:38–42; doi:10.1016/j.jpeds.2017.07.034
Investigators from multiple institutions conducted a cross-section-
al study to assess the association between coronary artery disease
risk factors and carotid intima-media thickness in children. Partic-
ipants were fifth graders attending schools in southern Maine. For
the study, families of participating children completed a question-
naire regarding history of early coronary vascular disease (CVD) in
either parents or grandparents, and average physical activity in the
child. The following data were collected on each study participant:
height and weight (to calculate BMI), blood pressure (BP), total cho-
lesterol, high-density lipoprotein cholesterol (HDL-C), and fasting
glucose. Cut-off values were established to determine the presence
of a coronary artery disease risk factor, including family history of
early CVD (occurring at <55 years in male family members and <65
years in females), physical inactivity (<60 minutes/day), cholesterol
(>170 mg/dL), HDL-C (<40 mg/dL), glucose (>100 mg/dL), overweight
(BMI >85th percentile), and BP (>130/85). Ultrasound examinations
of both the right and left carotid arteries were conducted in study
children and carotid intima-media thickness measured. The associ-
ation between the number of risk factors and carotid intima-media
thickness was assessed with ANOVA. In addition, regression analy-
sis was used to assess the association of individual risk factors with
left and right average carotid intima-media thickness.
A total of 123 children were enrolled in the study. Results were ob-
tained on 119 with a mean age of 10.5 + 0.52 years; 51% of partici-
pants were female. Overall, 65% of study children were overweight,
67% were physically inactive, 30% had high cholesterol levels, 25%
low HDL-C, 13% high glucose levels, and 5% elevated BP. Compared
to children with 0 coronary artery disease risk factors, those with
2 or 3 risk factors had significantly higher left carotid intima-me-
dia thickness (P<.001 for both), and those with 3 risk factors had
significantly higher left and right average thickness (P<.03). There
was no statistical association between risk factors and right ca-
rotid intima-media thickness. Among individual risk factors, high
glucose levels, elevated systolic BP, and overweight were statisti-
cally associated with increased carotid intima-media thickness. In
multiple analyses, overweight status was consistently the stron-
gest variable associated with carotid intima-media thickness.
The authors conclude that the presence of 2 or more coronary
artery disease risk factors is associated with carotid intima-media
thickness.   3. Ren L, et al. PLoSOne. 2015;10:e0144182; doi: 10.1371/journal.pone.0144182
COMMENTARY BY
Benjamin R. Doolittle, MD, M Div, FAAP, FACP, Internal Medicine &
Pediatrics, Yale University School of Medicine, New Haven, CT
Dr Doolittle has disclosed no financial relationship relevant to this commentary. This
commentary does not contain a discussion of an unapproved/investigative use of a com-
mercial product/device.
While tackling CVD risk factors is an important element of adult
primary care, addressing these same issues among children is
20
not often regarded as a priority. The results of the current study
challenge us to reconsider and suggest that a more aggressive ap-
proach toward modifiable risk factors, even among children, may
be important toward preventing future cardiovascular morbidity
and mortality. In the 1990s, the Bogalusa Heart study revealed
that atherosclerotic plaques were present even among children
and young adults at autopsy.1 These findings demonstrated that
atherosclerosis—the number one cause of adult mortality—begins
in childhood. The current investigators went one step further and
examined factors that affect vascular health that can be changed:
BMI, lipids, BP. Cardiovascular health is no longer only the domain
of adult providers.
This study begs several provocative questions: Does increased ca-
rotid intima thickness in childhood lead to increased cardiovascu-
lar mortality in adulthood? The only way to answer this is to create
a population database similar to the Framingham Heart study. Is
it not time for such a longitudinal study? The other provocative
question: Should we begin to intervene more aggressively with
statins and BP lowering agents among children? We have been
hesitant to do so for a number of reasons (cost, side-effects, lack
of controlled trials, lack of familiarity with these agents, conflict-
ing guidelines). (See AAP Grand Rounds, July 2015;34[1]:1–2.2) And
yet, perhaps the time is right for an intervention trial to address
this question.
There is one important study limitation. Carotid artery intima
thickness is used as a proxy for future CVD; the thicker the ca-
rotid artery intima, the greater the likelihood for atherosclerotic
plaque. This correlation is imperfect.3 And yet, measuring carotid
artery intima thickness via ultrasound is the best objective mea-
sure of vascular health short of biopsy.
The challenge for providers now is to use the information at
hand—BMI, lipids, glucose, BP—and emphasize healthy lifestyles,
especially among children with risk factors.
Bottom Line: Even among children, cardiovascular risk factors
are associated with increased carotid intima-media thickness, a
marker for poor cardiovascular health. Pediatricians need to ad-
dress modifiable cardiovascular risk factors in their patients.
References
1. Berenson GS, et al. N Engl J Med. 1998;338:1650–1656; doi: 10.1056/
NEJM199806043382302
2. Gooding HC, et al. JAMA Pediatr. 2015;169(6):569–574; doi: http://dx.doi.
org/10.1001/jamapediatrics.2015.0168
aapgrandrounds.org
HOSPITAL MEDICINE
Rethinking Macrolide Use in
Inpatient Pneumonia
Source: Williams DJ, Edwards KM, Self WH, et al. Effectiveness of β-lac-
tam monotherapy vs macrolide combination therapy for children
hospitalized with pneumonia. JAMA Pediatr. 2017;171(12):1184–1191; doi:
10.1001/jamapediatrics.2017.3225
Investigators from multiple institutions conducted a multicenter,
prospective, observational study to compare the effectiveness
of β-lactam monotherapy (BL) versus β-lactam plus macrolide
combination therapy (BL+) in the treatment of pneumonia in hos-
pitalized children. Children <18 years of age were eligible if they
were hospitalized at 1 of 3 US children’s hospitals with signs or
symptoms of acute infection, acute respiratory illness, and ra-
diographic evidence of pneumonia. Study staff conducted chart
reviews of all enrolled participants for demographic and clinical
information, including respiratory virus tests and blood cultures.
The primary exposure variable was empirical antibiotic use, de-
fined as antibiotics received during the first 2 calendar days of
hospitalization. BL participants received β-lactam monotherapy,
defined as oral or parenteral penicillin, ampicillin, ampicillin-sul-
bactam, amoxicillin, amoxicillin-clavulanate, or 2nd- or 3rd-gen-
eration cephalosporins. BL+ participants received a β-lactam plus
an oral or parenteral macrolide (azithromycin or clarithromycin).
The primary outcome was length of stay (LOS). The investigators
tested the independent association of antibiotic use group and
LOS after adjusting for demographic and clinical characteristics.
They also conducted several propensity score-matched analyses
to address the possibility of residual confounding.
There were 1,418 participants included in analysis, of whom 1,019
(71.9%) received BL. Among BL+ participants, 98% received azith-
romycin. Overall, 73.7% of participants had a virus detected with
or without bacterial co-detection. Only 8.8% had atypical bacte-
ria detected, with 95.2% of these being Mycoplasma pneumoniae.
Median LOS was 55 hours for children receiving BL alone versus
59 hours among BL+ participants (adjusted hazard ratio 0.87; 95%
CI 0.74–1.01). There remained no significant difference in LOS be-
tween antibiotic groups in the propensity analyses.
The investigators conclude that empirical use of macrolides in
children hospitalized with pneumonia confers no benefit in LOS.
COMMENTARY BY
Jeffrey Winer, MD, MA, MSHS, FAAP, Pediatric Hospital Medicine,
Wolfson Children’s Hospital, Jacksonville, FL
Dr Winer has disclosed no financial relationship relevant to this commentary. This com-
mentary does not contain a discussion of an unapproved/investigative use of a commer-
cial product/device.
The Pediatric Infectious Diseases Society/Infectious Diseases
Society of America guidelines for the management of communi-
ty-acquired pneumonia (CAP) recommend macrolides when atyp-
ical causes of pneumonia are “significant considerations.”1 The
question remains, which patients will benefit from the addition
of macrolides for coverage of typical causes of CAP? The current
• FEBRUARY 2018
TM
investigation, to the extent that any single, non-randomized study
can, seems to answer this question with an emphatic answer: very
few, if any.
There was no benefit demonstrated for empirical therapy with
a macrolide in combination with a β-lactam as compared to
β-lactam monotherapy. Results were similar in children with
atypical pathogens admitted to the ICU or with wheezing. LOS, ICU
admission, rehospitalization, and recovery at follow-up among
children in these 2 groups were not significantly different. The
findings of the current study differ from previous studies that
used administrative data,2,3 which inherently lack detailed etio-
logic and clinical information. While the benefit of macrolides in
pediatric CAP may be unclear, there is ample evidence of the pub-
lic health risk of their overuse. With the imperfect, real-time infor-
mation available to clinicians, it may become increasingly difficult
to justify the addition of a macrolide in most instances.
Bottom Line: Pediatric inpatients with CAP have equivalent clini-
cal outcomes whether they are treated with β-lactam monothera-
py or β-lactam plus macrolide therapy.
EDITORS’ NOTE
A virus was detected in >70% and bacteria in <10% of the pa-
tients in this study. Perhaps many children hospitalized with
CAP, especially those whose immunizations are up to date, need
no antibiotic therapy at all.4 (See AAP Grand Rounds September
2017;38[3]:29.5)
References
1. Smith MJ. JAMA Pediatr. 2017;171(12):1157–1148; doi: 10.1001/jamapediat-
rics.2017.3828
2. Bradley JS, et al. Clin Infect Dis. 2011;53(7):e25–76; doi: 10.1093/cid/cir531
3. Ambroggio L, et al. J Pediatr. 2012;161(6):1097–1103; doi: 10.1016/j.
jpeds.2012.06.067
4. Leyenaar JK, et al. Pediatr Infect Dis. 2014;33(4):387–392; doi: 10.1097/
INF.0000000000000119
5. Olarte L, et al. Clin Infect Dis. 2017;64(12):1699–1704; doi: 10.1093/cid/cix115
Continuing Thoughts — Evidence eMended*
*emend — from the Latin (c. 1400), “to free
from fault”; to improve by critical editing
www.aappublications.org/blog-posts
Come visit our blog, Evidence eMended, hosted by
Bud Wiedermann, MD, MA, FAAP. Bud is a former
member of our Editorial Board who now serves as
Consulting Editor for New Media.
The goal of the blog is to add value (and fun) for our readers by
creating an easy way to enter the discussion of specific studies.
Discuss the perceived weight of the evidence as applied to your
practice situations and the patient populations you serve, and in
the process learn more about critical appraisal.
Come talk with Bud as he leads the way.
21
NEPHROLOGY

Outcomes in Rituximab-Treated
Nephrotic Syndrome
Source: Kamei K, Ishikura K, Sako M, et al. Long-term outcome of child- Grand Rounds, the use of lower steroid dosing was evaluated (AAP
hood-onset complicated nephrotic syndrome after a multicenter, dou- Grand Rounds, April 2017;37[4]:391). Other researchers have exam-
ble-blind, randomized, placebo-controlled trial of rituximab. Pediatr ined the use of second-line immunosuppressive agents. While
Nephrol. 2017;32(11):2071–2078; doi:10.1007/s00467-017-3718-0 oral agents such as tacrolimus and mycophenolate are frequently
used, these long-term, twice-daily medications pose problems for
Investigators from multiple Japanese institutions conducted a
patient compliance and tolerability. As an IV agent, rituximab is
retrospective observational study to assess the long-term clini-
a promising alternative. In an earlier study conducted by the in-
cal course among children with frequently relapsing nephrotic
vestigators in the present study, use of rituximab was found to be
syndrome (FRNS) and steroid-dependent nephrotic syndrome
superior to placebo in patients with FRNS and SDNS, with longer
(SDNS) treated after treatment with rituximab. Patients with child-
time to relapse (median of 267 days) and lower rate of steroid use
hood-onset FRNS or SDNS, who had received 375 mg/m2 rituximab
post-treatment.2
once weekly for 4 weeks as part of a clinical trial conducted from
2008–2010, were eligible to participate. FRNS was defined as ≥4 re- The current study included both patients who had received only
lapses of nephrotic syndrome within any 12-month period. SDNS steroids prior to rituximab, as well as those who were on other
was defined as 2 consecutive relapses during tapering or within 2 prior immunosuppressive agents. Similarly, a study comparing re-
weeks of discontinuation of steroid treatment. sponses to various second-line agents in FRNS found rituximab
to be superior to both mycophenolate and tacrolimus with regard
Patients were followed clinically until 2014. Patients were as-
to complete remission rates (83% vs 63% and 78%, respectively),3
sessed for disease relapse, need for additional rituximab treat-
although numbers were small. These data suggest that rituximab
ment, immunosuppressive agent treatment, renal function, and
is effective for management of nephrotic syndrome. The current
rituximab-related adverse events. Demographics and age of onset
study expands on previous data to provide long-term follow-up,
of disease were also collected. Investigators used Kaplan-Meier
with a median follow-up time of 59 months. While most patients
survival curves to determine relapse-free remission time (mea-
required additional treatment during the follow-up period, most
sured in days) after rituximab treatment.
relapses occurred after B-cell repletion. This is promising, as the
There were 51 patients included in analysis. Most patients were higher rate of remission combined with longer relapse-free dura-
male (75%) and had a median age of 11 years. Median age at onset tion suggest that rituximab could help reduce the burden of ill-
of FRNS or SDNS was 3 years. Overall, 94% of patients (N=48) devel- ness in patients with FRNS.
oped relapses during the study period, with a median relapse-free
While this study did not define a treatment regimen for post-rit-
remission time of 261 days. Among those who relapsed, 86% un-
uximab relapse, future research may focus on the establishment
derwent re-administration or continuation of immunosuppressive
of maintenance rituximab regimens for management of nephrotic
agents, and 43% received additional rituximab treatment. Among
syndrome.
13 patients who did not require immunosuppressive agents at the
time of initial rituximab treatment, 5 did not require further ritux- Bottom Line: Rituximab is a promising agent for frequently relaps-
imab or immunosuppressive agents and 3 did not relapse during ing and steroid-dependent nephrotic syndrome, providing benefit
the study period. No patients developed chronic renal insufficien- for efficacy and patient tolerability.
cy, and there were no rituximab-related adverse events. References
The investigators conclude that most patients with FRNS and 1. Raja K, et al. Pediatr Nephrol. 2017;32(1):99–105
SDNS treated with rituximab suffer relapse and require additional 2. Iijima K, et al. Lancet. 2014;384(9950):1273–1281; doi: 10.1016/S0140-
6736(14)60541-9
rituximab treatment or long-term immunosuppressive agents.
3. Kim J, et al. Nephron Extra. 2014;4(1):8–17; doi: 10.1159/000357355
COMMENTARY BY
Pamela Singer, MD, FAAP, Donald and Barbara Zucker School of
Medicine at Hofstra/Northwell, New Hyde Park, NY
Dr Singer has disclosed no financial relationship relevant to this commentary. This com-
mentary does not contain a discussion of an unapproved/investigative use of a commer-
cial product/device.

The development of alternatives to corticosteroids for the man-

agement of nephrotic syndrome remains an area of great interest
for pediatric nephrologists. In a study recently reviewed in AAP

22 aapgrandrounds.org
NEONATOLOGY

Neonatal Therapeutic Hypothermia: Timing

and Effects on Outcomes
Source: Laptook AR, Shankaran S, Tyson JE, et al. Effect of therapeu- COMMENTARY BY
tic hypothermia initiated after 6 hours of age on death or disability Jonathan Mintzer, MD, FAAP, Stony Brook Children’s Hospital, Neonatal-
among newborns with hypoxic-ischemic encephalopathy: A random- Perinatal Medicine, Stony Brook, NY
ized clinical trial. JAMA. 2017 Oct 24;318(16):1550–1560. doi: 10.1001/ Dr Mintzer has disclosed no financial relationship relevant to this commentary. This com-
jama.2017.14972 mentary does not contain a discussion of an unapproved/investigative use of a commer-
cial product/device.
Investigators from multiple institutions conducted a random-
Among full-term and late premature neonates, HIE is associated
ized controlled trial to estimate the effectiveness of hypother-
with substantial short- and long-term sequelae, including death
mia treatment initiated 6 to 24 hours after birth in term and late
and neurodevelopmental impairment. In numerous studies, the
preterm newborns with hypoxic–ischemic encephalopathy (HIE).
advent of therapeutic hypothermia has represented a critical
Study participants were neonates born at >36 weeks’ gestation
advance in the management of HIE, with significant benefits on
with moderate to severe HIE, as defined using standardized cri-
long-term outcomes.1–3 In these original studies, cooling was initi-
teria, who were treated between 2008–2014 at institutions within
ated within a 6-hour window following ischemic injury based on
the National Institutes of Child Health and Human Development
sheep models.4 However, hypothermia initiation before 6 hours
Neonatal Research Network. Neonates randomized to the inter-
after birth may present logistic difficulties when infants require
vention arm were cooled to a target esophageal temperature of
hospital transfer. Moreover, HIE may not be recognized clinically
33.5° C for 96 hours; those in the control arm were maintained
until after 6 hours. Though cooling following a longer post-isch-
at an esophageal temperature of 37.0°C. Study infants underwent
emia interval has been hypothesized to be potentially neuropro-
developmental testing using the Bayley Scales of Infant Develop-
tective, no studies to date have confirmed the clinical benefit of
ment III (Bayley) and the Gross Motor Function Classification Score
this approach.
(GMFCS, range of scores 0–5, with higher scores indicative of more
disability) at 18–22 months of age. The primary outcome was death The investigators in the current study conducted a randomized
or moderate to severe disability. Moderate or severe disability was controlled trial to evaluate outcomes following therapeutic hypo-
defined by specific Bailey score and GMFCS criteria, blindness, thermia initiated 6–24 hours following ischemic injury among ne-
hearing deficits, or presence of an active seizure disorder. onates ≥36 weeks’ gestational age. A reduced rate of the compos-
ite outcome of death or moderate-severe disability was observed
Because HIE is a rare diagnosis, the investigators realized that it
in the hypothermia group compared to controls. Though these
would be difficult to enroll a sample large enough to have statisti-
results appear promising, ongoing research into this management
cal power to detect clinically relevant differences. Thus, they used
strategy is required to fully ascertain its effectiveness. Additional
a Bayesian analysis using a “neutral prior,” assuming no treatment
therapies with adjuvant neuroprotective agents (eg, antiepileptic
effect (relative risk [RR] 1.0, with a 95% credible interval ranging
drugs, erythropoietin) are currently being investigated.5
from an RR of 0.5–2.0). Log binomial models, which included pos-
sible confounding variables, were used to measure RR. This anal- Bottom Line: Therapeutic hypothermia for hypoxic-ischemic en-
ysis plan allowed for an estimate of the overall probability of a cephalopathy initiated 6–24 hours following ischemic injury may
treatment effect, and the probability that death or disability was have a favorable effect on death or moderate-severe disability
at least 1%, 2%, or 3% less among hypothermia-treated infants among infants born ≥36 weeks’ gestational age.
than controls. References
A total of 168 newborns were enrolled (83 randomized to hypo- 1. Azzopardi D, et al. N Engl J Med. 2014;371(2):140–149; doi:10.1056/NEJ-
thermia) with a mean gestational age of 39 weeks; complete fol- Moa1315788
2. Shankaran S, et al. N Engl J Med. 2005;353(15):1574–1584; doi:10.1056/NE-
low-up was obtained on 157 (93.5%). There were 9 deaths among
JMcps050929
infants in each treatment group. Death or disability occurred in
3. Shankaran S, et al. N Engl J Med. 2012;366(22):2085–2092; doi:10.1056/NEJ-
19 of 78 infants randomized to hypothermia (24.4%) and 22 of 79
Moa1112066
(27.9%) among noncooled controls (adjusted RR was 0.86, 95% 4. Gunn AJ, et al. J Clin Invest. 1997;99(2):248–256; doi:10.1172/JCI119153
credible interval 0.58, 1.29). Based on these results, there was a 5. Papile LA, et al. Pediatrics. 2014;133(6):1146–1150; doi:10.1542/peds.2014-0899
76% probability of reduced death and disability among newborns
treated with hypothermia compared to control infants. The prob-
ability that hypothermia treatment reduced death or disability by
at least 1%, 2%, or 3% was 71%, 64%, and 56%, respectively.
The authors conclude that hypothermia initiated at 6–24 hours of
life in infants born at >36 weeks’ gestation with moderate-to-se-
vere HIE likely reduces the risk of death and disability, but there
remains uncertainty regarding its effectiveness.

• FEBRUARY 2018 23
TM
CME QUESTIONS
The following continuing medical education questions cover the content
of the February 2018 issue of AAP Grand Rounds. Please keep this issue.
Each year’s material is worth up to 18 AMA PRA Category 1 Credit(s)TM.
Complete and claim credit online at www.aapgrandrounds.org.
Need username and password? Contact customer service at 866-843-2271.
CME OBJECTIVES
1.  Contrast the relative risk from the absolute risk of febrile seizures
associated with vaccination.
2.  Understand the association of autism with maternal and paternal
acetaminophen use.
3.  Describe the developmental outcomes in children with asymptomatic
congenital CMV infection.
1. A previously healthy 4-month-old infant is brought to the pediatrician’s office for
a well-child check. He is scheduled to receive the recommended vaccinations (ro-
tavirus, pneumococcal conjugate, Haemophilus influenza type b, diphtheria-teta-
nus-acellular pertussis, hepatitis B, and inactivated poliovirus vaccines). The mother
asks about potential complications of vaccinations. The most accurate estimate of
the incidence rate ratio of a febrile seizure on the day of and the day following im-
munizations is which of the following based on the study by Duffy et al?
a. 23 per 100,000 children
b. 2 per 100 children
c. 0.4 per 100 children
d. 5 per 100 children
e. 28 per 1,000 children
2. A 14-month-old boy is admitted to the hospital for incision and drainage of a left
buttock abscess that has increased in size. He has been on oral cefdinir the past
3 days. A culture of drainage from the site 3 days ago grew MRSA susceptible to
clindamycin, vancomycin, and TMP-SMX (trimethoprim-sulfamethoxazole). He is
started on IV clindamycin. Cultures from his nares and inguinal fold on admission
grows MRSA. Which of the following is the most accurate conclusion of the study
by Hogan et al on the impact of systemic antibiotics on S aureus colonization and
recurrent skin infection?
a. Systemic antibiotics (clindamycin or TMP-SMX) compared to no antibiotics
resulted in decreased colonization but had no effect on the rate of recurrent
skin infection.
b. Use of IV vancomycin was associated with a significantly decreased risk of
recurrent infection.
c. TMP-SMX was not associated with a decreased risk of remaining colonized.
d. 81% of baseline skin infections were caused by methicillin-susceptible
S aureus.
e. Clindamycin was more effective than TMP-SMX in eradicating S aureus
colonization.
3. At the well-child visit for a 3-year-old boy, his pregnant mother asks you whether
it is OK for her to take acetaminophen for pain from an ankle sprain. Based on
the recent study by Ystrom et al, you counsel her that increased risk of ADHD was
associated with acetaminophen use by which of the following?
a. Mothers before pregnancy
b. Neonates
c. Fathers before pregnancy and long-term use by mothers during pregnancy
d. Fathers during pregnancy
e. Postpartum breastfeeding mothers
4. An asymptomatic 18-year-old male living in the dorms is a student at a university
with an outbreak of mumps. He received 2 doses of MMR at 12 months and 4 years
of age. The university provided MMR vaccine to all students, and he received a
third dose of MMR. Which of the following is most accurate concerning the effec-
tiveness of the MMR vaccine in protecting him from contracting mumps based on
the study by Cardemil et al?
a. Circulating wild type virus exposure from childhood will help protect him from
mumps.
b. High rates of receipt of at least 2 doses of MMR will prevent future campus
outbreaks.
c. A third dose of MMR vaccine provides protection due to capturing prior
non-responders.
d. Time since the 2-dose MMR series was completed is unrelated to the risk of
acquiring mumps.
e. MMR vaccination with a third dose provided increased protection to those
with 2 prior doses.
24
5. A term, healthy newborn infant is diagnosed with congenital CMV infection as part
of a research study. She passed her newborn hearing screen. The pediatrician rec-
ommends frequent monitoring of her hearing and neurodevelopment. Her parents
ask if there are any studies that provide insights into long-term neurodevelop-
mental outcomes of infants with asymptomatic congenital CMV infection. Which of
the following is the most accurate conclusion of the study by Lopez et al?
a. 25% of infants diagnosed with congenital CMV infection developed sensori-
neural hearing loss (SNHL) by age 2 years.
b. Infants with asymptomatic congenital CMV infection and SNHL had significant-
ly lower expressive language scores than uninfected children.
c. Infants with asymptomatic congenital CMV infection who maintained normal
hearing by age 2 years had similar intelligence during childhood compared to
uninfected children.
d. Math scores were lower in infants with congenital CMV infection than in unin-
fected children.
e. Reading scores were lower in infants with congenital CMV infection than in
uninfected children.
6. An 11-year-old girl presents to the office for a routine well-child exam. The par-
ent’s only current concern is that her weight is at the 97th percentile. The paternal
grandfather was obese, and he recently died from a stroke. Based on the study
by White et al, which of the following was most strongly associated with a higher
carotid intima-medial thickness based on multivariate regression analysis?
a. Elevated total cholesterol
b. Elevated blood glucose
c. Increased HDL-C
d. Elevated diastolic blood pressure
e. High BMI (overweight or obese)
7. A previously healthy 40-month-old girl is admitted to the hospital from the ED with
community-acquired left lower lobe pneumonia confirmed on chest x-ray. The resi-
dent on service confers with the attending, and the patient is started on IV ampicillin.
The medical student asks if azithromycin should be added for coverage of atypical
bacteria. Which of the following is the most accurate conclusion of the study by
Williams et al on the effectiveness of ß-lactam monotherapy (BL) vs ß-lactam plus
macrolide combination therapy (BL+) for children hospitalized with pneumonia?
a. B L+ treated subjects had a significantly shorter length of hospital stay in the
unmatched and the propensity-matched cohorts.
b. Most participants in the study received BL+ therapy.
c. 36% of participants had atypical bacteria detected.
d. BL+ combination therapy conferred no benefit over BL monotherapy.
e. The unmatched cohort showed no benefit with BL+ therapy over monotherapy,
but the propensity-matched analysis showed a significantly shorter length of
hospital stay for the BL+ treated group.
8. You are caring for a 6-year-old patient with nephrotic syndrome who has now pre-
sented with his fourth relapse in 9 months. In discussing the use of rituximab as a
second agent with the parents, which of the following based on the study by Kamei
et al would best predict that the patient would not need additional immunosuppres-
sive agents or additional rituximab after receiving the initial course of rituximab?
a. N
 ot being on other immunosuppressive agents at the time of rituximab
administration
b. Male gender
c. Duration of B-cell depletion after rituximab
d. Age at onset of nephrotic syndrome
e. Development of chronic renal insufficiency
9. A 36 6/7–week gestational age neonate with suspected hypoxic-ischemic enceph-
alopathy is being transferred to your institution from a rural hospital for consid-
eration of whole body cooling. The baby was born 4 hours ago following urgent
Cesarean delivery due to decreased fetal movement, with a fetal heart rate tracing
demonstrating multiple prolonged bradycardic events. Apgars were 2, 3, and 6 at
1, 5, and 10 minutes of life, respectively. The birthing hospital is located 2 hours
away by air transport. Based on the article by Laptook et al, which of the following
statements concerning therapeutic hypothermia is most accurate?
a. T herapeutic hypothermia initiated beyond 6 hours following ischemic injury is
associated with severe fat necrosis and should be avoided.
b. Given the likely delay in initiating therapeutic hypothermia, this form of thera-
py should only be considered if seizure activity is observed.
c. Among infants born at ≥36 weeks’ gestation, therapeutic hypothermia started
between 6 and 24 hours post-ischemia was associated with 76% probability of
reduced death or moderate to severe disability.
d. Delayed therapeutic hypothermia initiated beyond 6 hours of ischemic injury
had a 93% probability of reducing death or disability by at least 3%.
e. Therapeutic hypothermia started between 6 and 24 hours post-ischemic injury
was associated with benefits for long-term outcomes only when applied as
selective head cooling rather than whole body cooling.
8. a 6. e 2. e 4. e
9. c 7. d 5. c 1. a 3. c
Answers:
aapgrandrounds.org