Journal of Infection (2016) xx, 1e9

www.elsevierhealth.com/journals/jinf

Diagnosis and management of

undifferentiated fever in children
Sarah S. Long*

Drexel University College of Medicine, Section of Infectious Diseases, St. Christopher’s Hospital for
Children, Philadelphia, PA, USA
Available online - - -

KEYWORDS Summary The incidence and likely causes of fever of unknown origin (FUO) have changed
Chikungunka; over the last few decades, largely because enhanced capabilities of laboratory testing and im-
Vertebral osteomyelitis; aging have helped confirm earlier diagnoses. History and examination are still of paramount
Fatigue of importance for cryptogenic infections. Adolescents who have persisting nonspecific complaints
deconditioning; of fatigue sometimes are referred to Pediatric Infectious Diseases consultants for FUO because
Systemic exertion the problem began with an acute febrile illness or measured temperatures are misidentified as
intolerance disease; “fevers”. A thorough history that reveals myriad symptoms when juxtaposed against normal
Herpes simplex virus findings on examination and simple laboratory testing can suggest a diagnosis of “fatigue of de-
conditioning”. “Treatment” is forced return to school, and reconditioning. The management of
patients with acute onset of fever without an obvious source or focus of infection is dependent
on age. Infants under one month of age are at risk for serious and rapidly progressive bacterial
and viral infections, and yet initially can have fever without other observable abnormalities.
Urgent investigation and pre-emptive therapies usually are prudent. By two months of age,
clinical judgment best guides management. Between one and two months of age, a decision
to investigate or not depends on considerations of the height and duration of fever, the pa-
tient’s observable behavior/interaction, knowledge of concurrent family illnesses, and likeli-
hood of close observation and follow up. Children 6 monthse36 months of age with acute
onset of fever who appear well and have no observable focus of infection can be evaluated
clinically, without laboratory investigation or antibiotic therapy, unless risk factors elevate
the likelihood of urinary tract infection.
ª 2016 The British Infection Association. Published by Elsevier Ltd. All rights reserved.

Unfocussed fever: Five cases, five approaches
to diagnosis and management
Five cases are presented in which patients came to medical
attention because of fever, yet had no clear cause after a
detailed history, physical examination and simple labora-
tory testing. The reader is challenged to choose the most
likely etiology or next step in management. The discussions
are focussed to highlight evidence from recent medical
literature.

* St. Christopher’s Hospital for Children, 160 E. Erie Avenue, Section of Infectious Diseases, USA. Tel.: þ1 215 427 5204; fax: þ1 215 427
8389.
E-mail address: sarah.long@drexelmed.edu

http://dx.doi.org/10.1016/j.jinf.2016.04.025
0163-4453/ª 2016 The British Infection Association. Published by Elsevier Ltd. All rights reserved.

Please cite this article in press as: Long SS, Diagnosis and management of undifferentiated fever in children, J Infect (2016), http://
dx.doi.org/10.1016/j.jinf.2016.04.025
2 S.S. Long
Case 1: Fever and back pain in a returned
traveler
Box 1 shows the principal data for this case together with a
question about the most likely diagnosis.
Chikungunya
Chikungunya virus is an alphavirus of the family Togaviri-
dae, which is arthropod-borne, classically transmitted by
Aedes aegypti. Gene mutations in the virus envelope in
certain clades have permitted adaptation to a different
mosquito vector, Ae. albopictus, and expansive transmissi-
bility. For chikungunya virus, the important step to sus-
tained mosquitoehumanemosquito cycle appears to be
related to human behavior driving vector adaptation, and
vector adaptation driving virus adaptation.1 In interepi-
demic periods, chikungunya is maintained in animals, and
in epidemic periods humans are the reservoir and ampli-
fying host. Chikungunya virus appears to have originated
and evolved into distinct clades over centuries in Africa,
with a pandemic involving parts of the Western hemisphere
in the 1820s, and identification confirmed during a 1952
outbreak in southern Tanganyika (now Tanzania). From
Box 1. Case 1. 12-year-old Hispanic girl with
fever and back pain.
History of presenting illness
 Progressive back pain over 1 month e flexion
worse, supine & walking OK.
 2 months previously, had fever and back pain for 2
weeks while in Dominican Republic. Dx of chikun-
gunya made.
Examination findings
 37  C, BMI 29 (97th centile), vital signs normal.
Lying supine without pain.
 No vertebral/paraspinal tenderness. Loss of
normal lordotic curvature.
 Straight leg raising causes pain in lumbar spine.
Investigations
 Hb 12.4 g/dL, Hct 35.8%, WBC 9800 (49% neutro-
phils), Platelets 361,000.
 CRP 1.8 mg/dL, ESR 62 mm/h, protein 7.5 g/dL,
albumin 3.7 g/dL.
Question 1. The most likely diagnosis is
A. Chikungunya
B. Another infectious disease
C. An oncologic diagnosis
D. A rheumatologic diagnosis
E. Nonspecific musculoskeletal pain
Please cite this article in press as: Long SS, Diagnosis and management of undifferentiated fever in children, J Infect (2016), http://
dx.doi.org/10.1016/j.jinf.2016.04.025
the 1960s to 1990s, sporadic outbreaks sometimes involving
>100,000 people occurred due to distinct clades in Africa
and Asia. Spread from a 2005 coastal Kenya epidemic to
the islands of the Indian Ocean was the harbinger of world-
wide distribution. In 2007, Italy had the first inter-tropical
outbreak with indigenous transmission.
An outbreak beginning in late 2013 in the Caribbean led
to >1.5 million clinical cases of chikungunya transmission
documented by mid-2015, with spread to Latin America and
travel-associated cases confirmed in most U.S. states.
Indigenous transmission occurred in Florida, Puerto Rico
and the U.S. Virgin Islands.2 The Caribbean virus appears to
be an Asia lineage, so far unadapted to Ae. albopictus.3,4
Ae. aegyptiacus, however, is an aggressive daytime-biting
mosquito and with high levels of human viremia, transmis-
sion to a na€ıve population occurs in rapid waves, typically
affecting more than one-third of a population.5
The majority (72%e97%) of people infected with chikun-
gunya virus develop symptomatic disease e high fever and
bilateral, symmetrical and severe, debilitating arthralgia.6
Acute symptoms abate over 7e10 days, but >50% of pa-
tients have relapsing or recurring polyarthralgia/arthritis
or tenosynovitis that can persist for months to years.7
The patient’s clinical diagnosis of chikungunya made in
the Dominican Republic two months prior to admission
likely was correct, considering epidemic disease there at
the time. Recent progressive back pain and elevated
sedimentation rate make postinfectious arthritis a consid-
eration, however, isolated non-incapacitating back pain
was not a perfect fit and made further considerations
relevant.
Fever of unknown origin
In a landmark publication by Petersdorf and Beeson in
1961,8 fever of unknown origin (FUO) was codified for pur-
poses of clinical approach to require temperature >38.3  C
(101  F) documented on most days for >3 weeks, and no
diagnosis after 1 week of hospitalization. In the 1990s, de-
Kleijn et al.9 modernized the definition by replacing stay in
hospital with non-diagnostic computed tomography, if any
localizing sign or symptom was present. The availability
of improved diagnostic tests and imaging aided earlier diag-
noses and led to dwindling cases of FUO and shuffling of eti-
ologies. Durack and Street in 199110 grouped FUO by
settings of occurrence, each having distinctive differential
diagnoses: neutropenic host, nosocomial onset, HIV-
associated, and classic (no clear predisposition). Horowitz11
reviewed relative importance of categories of etiology re-
ported in FUO studies, primarily in adults, from 1961 to
2007. Progressively over years, relative importance of
infection fell from 36% to 16% and was replaced with “no
diagnosis” rising from 9% to 51%, and “noninfectious inflam-
matory disease” exchanged place with “neoplasm” (each
accounting for 22% and 7%, respectively in 2007). The fall
in infectious etiologies likely reflects improving capability
for earlier diagnosis rather than changes in disease
incidence.
Studies of FUO in children are limited. Jacobs and
Schutze in 199812 reported 146 children with fever for >2
weeks. Diagnoses were made in 58%, with infectious
Undifferentiated fever in children 3

diseases by far the leading diagnosis followed by autoim-
mune diseases and then malignancy. Among infectious dis-
eases, EpsteineBarr virus infection (EBV), osteomyelitis,
Bartonella infection (cat-scratch disease) and urinary tract
infection (UTI) were noteworthy. Description of an addi-
tional 19 cases of hepatosplenic cat-scratch disease in
1999 solidified bartonellosis as a prominent cause of FUO
in children.13 Enigmatic cases of osteomyelitis and UTI
would seem odd to have escaped more expeditious diag-
nosis. However, delayed diagnosis of osteomyelitis can
occur when a subacute course is related to less virulent or-
ganisms (e.g. Kingella, Actinomyces), partial treatment, or
involvement of flat bones. UTI can lead to FUO when
partially treated, complicated by perinephric infection or
lobar nephronia, or when congenital anomalies (e.g. dupli-
cated ureter) cause intermittent obstruction and recovery
in urine of no bacteria or non-enteric pathogens in low den-
sity. In 2011, Chow and Robinson14 performed a systematic
review of 18 studies (1638 cases) of FUO in children across
continents. Separating reports from developed and devel-
oping countries, infection was diagnosed frequently in
both settings (42% and 56%, respectively). However, compli-
cations of pneumonia, brucellosis, tuberculosis and typhoid
fever dominated in developing countries versus EBV infec-
tion, osteomyelitis, tuberculosis, cat-scratch disease and
UTI in developed countries. Miscellaneous diagnoses were
confirmed in 11% in both settings, predominantly
autoimmune and inflammatory bowel disease in developed
countries, and Kawasaki disease and familial Mediterranean
fever in developing countries. Overall, 384 children (23%)
had no diagnosis, but a new caveat was reported, i.e.
that fever resolved spontaneously in 49% without a diag-
nosis by the time of series’ publication.
Bottieau et al.15 prospectively enrolled 1962 adults who
came to referral in Belgium for fever episodes within 1 year
after visiting a tropical or subtropical area. Malaria was the
most common diagnosis. In studies of children with acute
and classic FUO living in tropical areas, malaria and typhoid
fever were top diagnoses.14,16
Our patient had a tuberculin skin test placed, blood and
urine samples taken for culture, and serologic tests sent for
chikungunya, EBV and Bartonella antibodies. Blood smear
and antigen tests for Plasmodium were negative. Magnetic
resonance imaging of her spine was performed. Images are
shown in Fig. 1. Osteomyelitis of lumbar 4e5 vertebrae with
paravertebral abscess was diagnosed.
Vertebral osteomyelitis
Vertebral osteomyelitis represents <2% of cases of osteo-
myelitis in children. Inoculation of bone occurs hematoge-
nously (or through rich local plexus draining the gut),
directly (e.g., spinal surgery) or from adjacent soft tissue

Figure 1 Sagittal and axial (at level shown by white bar on sagittal image) magnetic resonance images of Case 1 on fluid atten-
uated inversion recovery (FLAIR), and pre and post contrast sequences. Images demonstrate osteomyelitis with lytic lesions and
contrast-enhancing inflammation of lumbar bodies 4 and 5, and enhancing paravertebral soft tissue abscess (arrows). Lower right
image shows CT-guided site of aspiration.

Please cite this article in press as: Long SS, Diagnosis and management of undifferentiated fever in children, J Infect (2016), http://
dx.doi.org/10.1016/j.jinf.2016.04.025
4 S.S. Long

infection. The usual cause is Staphylococcus aureus or Es-

cherichia coli, and occasionally, Salmonella. Zimmerli Box 2. Case
2. A 15 year-old-girl referred for
et al.17 reviewed clinical and laboratory manifestations in prolonged fever and fatigue.
reported cases of vertebral osteomyelitis in adults
(Table 1). Insidious onset, lack of tenderness over the History of presenting illness
spine, and negative blood cultures in a substantial propor- 12 weeks ago: Acute 39.4  C, headache, myalgia,
tion of patients makes the diagnosis difficult until imaging sore throat, congestion
is performed. Vertebral osteomyelitis frequently is compli- Throat culture negative Group
cated by paravertebral (26%) or epidural (17%) extension, A Streptococcus
with resulting lower extremity weakness or paralysis in 38%. Following week: Fever and congestion abated
CT-guided aspiration of our patient’s paraspinal abscess Since then: Persisting temp (37.7  C every
(Fig. 1) revealed purulent material. Gram stain showed only 2e3 days)
WBCs, but Salmonella enterica, subspecies enterica sero- Headache, sore throat, weakness,
type Typhi was isolated from culture. Following a brief hospi- fatigue
talization, cefotaxime was administered intravenously (IV) Poor color and appetite
as an outpatient to complete 4 weeks, followed by ciproflox- 4 weeks ago: Attempted return school
acin orally for 2 weeks. Transition from IV to oral therapy was 2 h in: Reported to school nurse feeling
based on multiple studies of effectiveness, and a 6-week to- faint, chilly, sweaty, unable to
tal course on the 2014 report of a randomized controlled trial stay in class
in 359 French patients with vertebral osteomyelitis, which Since then: No return to school and now is
showed non-inferiority of 6-week versus 12-week therapy.18 home tutored
The patient has been symptom free for 12 months. With the Past history: No medical problems; “A” student;
knowledge of denouement, the patient’s febrile illness two Advanced science; Swim team;
months prior likely was typhoid fever and insidious progres- Church choir; Volunteer elder
sion of vertebral osteomyelitis was classic. The answer to center
Question 1 is B, cued by history of travel.
Question 2a. The most likely diagnosis is
Case 2: Prolonged fever and fatigue in an
A. Classic infectious disease of FUO
adolescent girl
B. Malignancy
C. Periodic fever syndrome
Box 2 shows the principal data for this case together with D. Rheumatologic syndrome
questions about the most likely diagnosis and the most E. None of these
appropriate further management. No diagnosis listed rises
to favored status with the information given (Question 2, Question 2b. What is your next diagnostic step?
Answer E), and obtaining more history is the next best
“diagnostic step” (Question 2a, Answer A). A. More history
Further history delineated these symptoms. Tempera- B. Referral to neurologist
ture peaks between 1600 and 2000 at <100  C. Headache is C. CT of the abdomen
diffuse, is relieved by acetaminophen and does not awaken D. MRI of the brain
her from sleep. Sore throat is only on awakening and is not E. Screening lab tests CBC, ESR, chemistries
worse with swallowing. Weakness has no pattern (e.g.,
Question 2c. The most likely diagnosis is

Table 1 Clinical & laboratory manifestations of vertebral A. Cryptogenic infection

osteomyelitis. B. Rheumatologic/autoinflammatory disease
Clinical findings C. Malignancy
Back pain 86% D. Fatigue of deconditioning
Fever 60% E. Psychiatric disorder
Weakness/sensory deficit/radiculo 30%
Tenderness over the spine 20%
Clinical course
Clinical picture dominated by back pain Most proximal, distal, truncal). She feels fatigued, yet retires at
Extravertebral source found 50% 2300, sleeps through the night, awakens at 0800 and does
Laboratory findings not sleep during the day. She has not lost weight, but rather
Time from onset of pain to diagnosis 42e59 days has gained 2 kg. The patient reports shooting pains (at
Elevated WBC or neutrophilia >80% multiple body areas lasting <10 s), eyes feeling “weird”,
Elevated sedimentation rate 98% heart “jumping”, stomach feeling “weird”, and feeling
Elevated C-reactive protein 100% faint upon arising. Mother has fibromyalgia, grandmother
Blood culture positive 30%e78% is in hospice after failing cancer therapy, and father works
at night. Parents work schedules have been changed to
Data from Zimmerli.17 accommodate the girl’s illness. Findings on complete

Please cite this article in press as: Long SS, Diagnosis and management of undifferentiated fever in children, J Infect (2016), http://
dx.doi.org/10.1016/j.jinf.2016.04.025
Undifferentiated fever in children
physical and neurologic examinations, including ophthal-
moscopic examination, are normal. Height and weight are
80e90th percentile, heart rate is 110, she is flushed,
animated and helpful with the examination. Complete
blood count, screening chemistry tests and urinalysis show
normal results. Sedimentation rate is 9 mm/h.
The most likely diagnosis is fatigue of deconditioning
(Question 2C, Answer D). The major clue is the remarkable
mismatch of the long list of symptoms yet weight gain,
normal findings on examination and normal laboratory test
results 12 weeks into the problem. The only other possible
differential diagnoses were abuse (excluded by open-ended
private interview with the patient) and depression
(excluded by the patient’s demeanor, expression of feeling
loved and supported, and interest in school, friends and the
future).
Table 2 summarizes typical findings in patients with fa-
tigue of deconditioning. Use of the operational “diagnosis”
of “deconditioning” is preferred by the author over
“chronic fatigue syndrome”, the latter which describes a
serious, complex systemic condition with a guarded prog-
nosis in adults, and which would lead families to much
internet-based misinformation if applied to most patients
referred to pediatric infectious diseases specialists.
Indeed, the Institute of Medicine recently has proposed
that the names for the adult conditions known as “chronic
fatigue syndrome” and “myalgic encephalopathy” be aban-
doned and replaced with “systemic exertion intolerance
disease”.19
The diagnosis of fatigue of deconditioning would be
highly unlikely in a child <10 years of age, and this author
has not made the diagnosis unless the parents’ or patient’s
pre-illness expectations of achievement were high. The
operational dynamic may be that the patient’s
achievement-based self-worth or valued place in the family
is made impossible temporarily because of an acute illness
or injury. The patient loses self-esteem and fears that
declaration of health will lead to resumption of expecta-
tions, and meanwhile captures family and external atten-
tion because of illness. Frequently there is a model of
illness or loss in the family, as in our patient’s case.
In this author’ experience, the vast majority of affected
patients do well with management as summarized in Table 3.
Incremental forced return to school is paramount, and
frequently is prescribed at graduating hours of attendance.
Table 2 Typical findings in patients with fatigue of
deconditioning.
 Age >12 years. Girls > boys
 Pre-illness achievement high
 Family expectations high
 Onset easily dated to an acute febrile illness/injury
 Escalated family and outside attention
 Lengthy list, but vague complaints
 Odd complaints (e.g., “shooting” pains, 30-s “blindness”,
stereotypic jerks, “paralysis”)
 No daytime sleep
 Preserved weight
 Normal physical and neurologic examination
 Normal results of screening lab tests
Please cite this article in press as: Long SS, Diagnosis and management of undifferentiated fever in children, J Infect (2016), http://
dx.doi.org/10.1016/j.jinf.2016.04.025
5
Table 3 Building the therapeutic alliance for manage-
ment of fatigue of deconditioning.
For the patient: Validate symptoms as accurate
Treatment:
Incremental, forced return to school
Exercise program
Predict exacerbation of symptoms
No school absence is permitted without visit
to medical provider
No expectation of performance is allowed
Focus is to be back at school among classmates
and friends
For the family:
Validate their concern
Validate their medical pursuit
Justify your diagnosis and lack of need for further
medical pursuit
Express confidence in the diagnosis and good prognosis
Treatment:
Change focus to health
Avoid asking “How do you feel?”
No expectation of performance is allowed
For the primary provider:
Take control and be firm on the rehabilitation plan
Discourage further subspecialist referrals
Encourage frequent visits/examinations and assure
continuous oversight
Any day that the patient “can’t go” or “must” be picked up
earlier than scheduled is coupled with a visit to the primary
provider. Since patients truly are deconditioned, incremen-
tal forced exercise also is key. Using a trainer is a highly suc-
cessful approach, as this person is outside the family
dynamic, and the patient usually responds to coach-
ingeinterestepraise of such a person. Many deconditioned
patients have orthostatic intolerance. In a controlled study
of military recruits, exercise ameliorated symptoms,20 and
in a small pediatric study of patients with orthostatic hypo-
tension, static handgrip coupled with rising abolished symp-
toms and minimized cardiovascular changes.21 Validating a
real (not imaginary or contrived) problem, engaging parents
and the patient to promise “no expectation of performance”,
and re-focusing attention on health rather than illness are
essential parts of a therapeutic alliance.
Cases 3: Acute febrile illness in a 14-day-old
and a 6-week-old infant without clinical focus
of infection
Box 3 shows the principal data with questions about the
most likely diagnosis and the appropriate further manage-
ment of fever in two cases: a 14-day-old infant and a 6-
week-old infant.
Bacterial infection in very young infants
Management of the very young febrile infant must take into
consideration the infant’s limited ability to localize and
6 S.S. Long
Box 3. Case 3a. A 14-day-old infant with
fever for 6 h. Case 3b. A 6-week-old infant
with fever for 6 h.
History of presenting illness
 Born at term to 20-year-old mother by vaginal
delivery.
 No maternal complications during pregnancy or
delivery.
 Infant was well until acute onset fever 38.5  C and
decreased feeding.
Examination findings
 Examination is reassuring.
 No exanthem, rhinorrhea.
Question 3. My management plan would be to/to
perform
A. Observe at home with follow up <24 h
B. Blood culture þ observe at home with follow up
<24 h
C. Blood, urine and CSF tests/cultures þ if reassur-
ing, ceftriaxone IM þ observe at home with follow
up <24 h
D. Blood, urine and CSF tests/cultures þ admit þ IV
antibiotics
E. D above þ HSV PCR on CSF/plasma þ acyclovir
control infection, and inability of parents and providers to
determine degree of illness because of undeveloped behav-
ioral milestones. Although likeliness of a serious bacterial
or disseminated viral infection is predictable by high fever
or hypothermia, ill appearance (e.g., poor color, tone,
attention to a face) and abnormally high or low white blood
cell count, lack of such findings does not exclude serious
infection. Attempts to identify febrile young infants at low
risk for serious bacterial infection (SBI) led to distillation of
infant characteristics/findings with high negative predic-
tive values (NPVs). The “Rochester criteria” published in
1990 concluded a NPV of >95% if the infant was well
appearing with no risk factors other than age, had no
suspected site of infection, had a total WBC count >5000
and <15,000 cells/mm3, urinalysis with 10 WBCs/hpf and
stool (if diarrheal) with 5 WBCs/hpf.22 The “Pittsburgh
criteria” published in 2001 concluded a NPV approaching
100% by adding a negative CSF evaluation and enhanced uri-
nalysis showing <10 cells/mm3 and negative Gram stain.23
In 1993, several pediatric infectious diseases and emer-
gency medicine subspecialists developed “practice guide-
lines” for management of febrile infants <90 days of age
who appeared well and had no source determined by exam-
ination.24 Their consensus opinion was that all infants <28
days of age should be evaluated (including blood, urine
and CSF tests and cultures) and hospitalized e with discre-
tionary use of empiric antibiotic therapy pending culture
results. Options for management were proposed for febrile
Please cite this article in press as: Long SS, Diagnosis and management of undifferentiated fever in children, J Infect (2016), http://
dx.doi.org/10.1016/j.jinf.2016.04.025
infants 28 through 89 days of age. They favored full evalu-
ation and if low risk criteria were met, administration of
one dose of ceftriaxone, discharge from the Emergency
Department (ED) and re-evaluation by a provider 24 h later.
These “guidelines” were not derived or vetted through
formal processes as guidelines would be currently, and
were not endorsed by the American Academy of Pediatrics
(AAP) or an Infectious Diseases Society. The approach was
adopted by Emergency Medicine physicians in the U.S. but
not by pediatricians practicing in the community. In 2004,
the AAP’s Pediatric Research in the Office Setting (PROS)
collaborative group published their case series using a clin-
ical prediction model, and confirmed similar prevalence of
SBI/meningitis in the community as in the ED.25 PROS data
showed an odds ratio for SBI/meningitis of 5.5 for age 30
days compared with 61e90 days, and incidence of 0.4% in
minimally ill appearing infants 25e60 days of age with tem-
perature <38.6  C. Although only 46% of PROS practitioners
adhered to the 1993 “guidelines” for evaluation and hospi-
talization of all febrile infants 30 days of age, and only
42% performed any testing of minimally ill appearing infants
31e90 days of age, their management choices had 97%
sensitivity (i.e., similar to ED’s testing approach) for treat-
ment of infants with SBI/meningitis at the first encounter.
The dichotomy of approach in EDs and practitioners’ offices
has persisted in the U.S., and may be appropriate depend-
ing on age. There remains for this author urgent concern for
the febrile infant <28 days of age, who may have serious
infection but no other observable signs of illness. By 6
weeks of age, the infant is able to smile meaningfully and
to show observable interest in the environment when
well. Such reassuring findings in a newly, modestly febrile
infant can permit safe management without testing or
treatment when constant family observation and repeated
medical evaluations can be performed. Although there is
not a single correct answer, this author would choose hos-
pitalization (D or E) for Case 3a and observation (A) for
Case 3b. Remembering that a blood culture is not therapeu-
tic, and bacteremia is not benign in young infants, one
should not be “observing” a patient at home with a blood
culture in the laboratory (Cases 3a and 3b, answer B is
incorrect).
Nonspecific presentation of perinatally acquired
herpes simplex virus infection
Although perinatally acquired herpes simplex virus (HSV)
infection is uncommon, and at least one-third of patients
have characteristic skin or mucous membrane lesions, HSV
must be considered in very young infants who appear ill or
have fever alone. In a retrospective series of 5817 neonates
without underlying conditions admitted to Texas Children’s
Hospital for any reason, 4.6% had serious bacterial infection
and 8.4% had viral infection confirmed (prior to availability
of respiratory panel polymerase chain reaction (PCR)
tests).26 For febrile neonates, 14.2% had bacterial infection
and 0.3% had HSV confirmed. HSV infection was more
frequent than bacterial infection, however, among infants
8e14 days of age (0.6% vs. 0.2%, respectively), in neonates
with hypothermia (1.1% vs. none) and in neonates with fe-
ver and CSF mononuclear pleocytosis (1.6% vs. 0.8%). In a
Undifferentiated fever in children
case series of 32 infants in Philadelphia brought for illness
confirmed to be perinatally acquired HSV, 50% had only
nonspecific symptoms or signs at presentation (i.e., absent
skin or mucous membrane lesions or neurologic signs/sei-
zures), which was fever in 75%.27 Of the 32 infants, 75%
had CNS HSV confirmed (predominantly by CSF PCR testing);
CNS HSV was confirmed in 40% of infants presenting with
mucocutaneous lesions, 83% with seizures, and 94% with
only nonspecific symptoms/signs. Importantly, 94% with
HSV diagnosis who presented with nonspecific symptoms/
signs had onset of illness at 21 days of age. A similarly
collected case series from St. Louis and Salt Lake City re-
ported 49 infants with HSV diagnosed before 6 weeks of
age. Sixteen percent (8 patients) had only nonspecific
symptoms/signs on admission; 7 patients had fever, 7
were <28 days of age at presentation, and 7 had positive
CSF PCR test.28 Among w7000 febrile infants 60 days of
age evaluated in Salt Lake City from 2004 to 2015, 0.3%
had HSV; incidence was 0.9% in those 14 days and 1.6%
in those with CSF pleocytosis.28
Nonspecific/febrile manifestations of perinatally ac-
quired HSV likely are early presentations of CNS infection.
Differences across studies may be related to differences in
incidence of HSV and use of EDs for acute care, where
complete testing and hospitalization of mildly ill neonates
is standard. Shah et al.29 reviewed an administrative data-
base of 41 freestanding children’s hospitals, collected 1098
cases of confirmed perinatal HSV infection treated with
acyclovir, and found odds ratio of death of 2.63 (95% CI,
1.36e5.08) for infants in whom commencement of acyclovir
was delayed until hospital day 2 or 3 compared with hospi-
tal day 1. In a 2011e2013 retrospective cohort study of ED
management of 35,000 febrile infants at 37 U.S. children’s
hospitals, 32% of 7712 infants 28 days of age were given
acyclovir empirically.30 Universal empiric acyclovir therapy
has not been recommended by any official body in the U.S.
as standard of care at this time.
Table 4 shows this author’s approach to management of
young infants who appear well and have fever without a
focus of infection. Melvin et al.31 published findings on 63
Table 4 An approach to management of the well-
appearing young infant with unfocussed fever.
 “All” infants <28 days should be hospitalized
e Perform usual tests and cultures plus CSF/plasma
for PCR testing for HSV  EV, HPeV (depending
on season, seizures)
e Administer acyclovir 60 mg/kg/day IV if onset
illness 21 days of age, or for clarity of algorithms,
admission <28 days of age
 Infants 60 days of age can be evaluated clinically
 Individualize management for infants 28e59 days of age
e Consider family illness, height and duration of
fever, sex, circumcision
e Use laboratory tests, hospitalization, antibiotics
selectively
e Minimize use of pathway of full evaluation
followed by administration of ceftriaxone and
observation at home for those meeting low-risk criteria
e Do not pursue HSV as etiology of unfocussed fever
Please cite this article in press as: Long SS, Diagnosis and management of undifferentiated fever in children, J Infect (2016), http://
dx.doi.org/10.1016/j.jinf.2016.04.025
7
Seattle infants with HSV diagnosis in whom plasma PCR
testing was performed at diagnosis. Overall, 83% had HSV
DNA detected in plasma: 100% of those categorized with
disseminated disease, 64% with CNS disease and 78% with
skin or mucous membrane disease. It is our practice to
perform PCR testing for HSV (and other viruses as appro-
priate) on CSF and plasma samples and to administer
acyclovir empirically to infants with onset of illness at
21 days of age (Case 3a, E) but not for hospitalized infants
28 days of age (Case 3b) because of extreme rarity of
nonspecific presentation of HSV in the latter group. With re-
striction of empiric use to infants <28 days of age with
nonspecific illness, number-needed-treat to treat one in-
fant with HSV might be 300e400 infants.
Case 3a was evaluated fully and treated with acyclovir
as well as antibiotics empirically. CSF examination showed:
WBCs 7/mm,3 RBCs 13/mm,3 glucose 56 mg/dL, protein
92 mg/dL; PCR testing was positive for HSV. The infant
developed seizures on hospital day 3, but had negative
CSF PCR on day 5, an otherwise uncomplicated 3-week
course of therapy and normal magnetic resonance imaging
study of the brain at discharge.
Case 4: Fever in a 6-month-old infant
Box 4 shows the principal data for this case together with a
question about the appropriate further management.
In a prospective study of 620 infants 4 weeks to 24 months
of age with rectal temperature of 38.3  C evaluated at the
Boston City Hospital in 1973 and 1974, 3.2% unexpectedly had
Box. 4. Case 4. A 6-month-old boy with fever
for 36 h.
History of presenting illness
 A previously healthy infant boy was well until
acute onset of high fever.
 Except for decreased feeding there are no other
symptoms or signs.
Examination findings
 Temperature is 39  C.
 Physical examination is reassuring.
 The boy is circumcised.
Question 5. My management plan would be to/to
perform
A. Observe at home with follow up <24 h
B. Urinalysis with further management pending
results
C. CBC with further management pending results
D. CBC, urinalysis & culture, blood
culture þ ceftriaxone IM and observe at home
with follow up <24 h
E. Tests of D above þ CSF þ admit to hospital
8 S.S. Long
a positive blood culture, which was Streptococcus pneumo-
niae in w80% of cases.32 Fever 39  C, age 7e18 months,
and WBC 15,000 cells/mm3 were associated with positive
blood culture. Called “benign pneumococcal bacteremia”
because the vast majority of infections resolved spontane-
ously, the “practice guidelines” of 1993 and U.S. Pediatric
Emergency Medicine community, nonetheless, subscribed
to performance of WBC count in children 3e36 months of
age with unfocussed high fever, and if WBC was
15,000 cells/mm3, obtaining a blood culture and adminis-
tering a dose of ceftriaxone prior to ED discharge.24,33
“Glitches” in the “guidelines” were noted: variable
incidence of pneumococcal bacteremia and meningitis
across studies; knowledge that in the vast majority,
bacteremia is self-limited; lack of evidence that the
approach reduced meningitis; risk of entering ill children
into the management pathway or potential dismissal from
concern of children with WBC <15,000 (both typical of
Haemophilus influenzae b); or consideration that >90% of
patients given a parenteral antibiotic had no bacterial
infection but could have clinical follow up clouded, leading
to additional testing.34 Finally, community practitioners
did not pursue well-appearing febrile infants and toddlers
with testing or treatment e to no adverse outcome.
Although this “practice guideline” also was not endorsed
by the AAP or an Infectious Diseases Society, it was com-
mon practice in U.S. EDs for the next two decades.35
With universal implementation of pneumococcal conjugate
vaccines (PCVs) such approach now seems indefensible.
The efficacy trial that led to licensure of PCV7 showed
>97% reduction in cases of vaccine serotype pneumococcal
bacteremia. Early studies of impact of PCV13 in young chil-
dren across nine countries in the Americas, Europe and
Israel show substantial reduction of invasive disease due
to the added vaccine serotypes as well as evidence of indi-
rect community protection.36e46 Pneumococcal bacter-
emia in a well appearing, previously healthy infant/
toddler in 2016 would be extremely rare. In fact, children
with invasive pneumococcal disease in the PCV13 era
rarely have occult illness or unfocussed fever, rather
they have underlying conditions and obvious tissue sites
of infection.47 A UTI is the most relevant, treatable infec-
tion in young children with unfocussed fever. A danger of
continuing use of the WBC as pivotal to action in EDs is
that infection due to non-pneumococcal pathogens, espe-
cially Gram-negative bacillary urinary tract pathogens
and S. aureus, frequently is not associated with WBC count
>15,000/mm3.
The recent ability to detect many respiratory tract
viruses by PCR testing of nasopharyngeal specimens (or
plasma or both) confirms what was assumed. In a U.S.
prospective study of >200 children 2e36 months of age
with acute onset of fever without a focus, 75% had virus
detected. Somewhat surprisingly, approximately one-third
of positive PCR tests were from plasma specimens only.48
Although the Case 5 boy has high fever, he is circumcised
and lacks other risk factors for UTI. His risk of UTI is <1%.49
The best answer to Question 5 is A. The patient had no tests
performed. Fever abated after three days, when a faint
pinpoint macular, non-coalescent exanthem appeared on
the face and trunk, typical of roseola.
Please cite this article in press as: Long SS, Diagnosis and management of undifferentiated fever in children, J Infect (2016), http://
dx.doi.org/10.1016/j.jinf.2016.04.025
Conflict of interest
Dr. Long has no conflict of interest to disclose.
References
1. Morens DM, Fauci AS. Chikungunya at the door e déja  vu all
over again? N Engl J Med 2014;371:885e7.
2. www.cdc.gov/chikungunya/geo [accessed 8.01.16].
3. Weaver SC. Arrival of chikungunya virus in the new world: pros-
pects for spread and impact on public health. PLoS Negl Trop
Dis 2014;8:e2921.
4. Weaver SC, Osorio JE, Livengood JA, Chen R, Stinchcomb DT.
Chikungunya virus and prospects for a vaccine. Expert Rev Vac-
cines 2012;11:1087e101.
5. Staples JE, Fischer M. Chikungunya virus in the Americas e
what a vectorborne pathogen can do. N Engl J Med 2014;
371:887e9.
6. Staples JE, Breiman RF, Powers AM. Chikungunya fever: an
epidemiological review of a re-emerging infectious disease.
Clin Infect Dis 2009;49:942e8.
7. Jaffar-Bandjee MC, Ramful D, Gauzere BA, Hoarau JJ, Krej-
bich-Trotot P, Robin S, et al. Emergence and clinical insights
into the pathology of chikungunya virus infection: chronic
arthralgia, pain and arthritis post-CHIK. Expert Rev Anti Infect
Ther 2010;8:987e96.
8. Petersdorf RG, Beeson PB. Fever of unexplained origin: report
of 100 cases. Medicine (Baltimore) 1961;40:1e30.
9. deKleijn EM, van Lier HJ, van der Meer JWfor The Netherlands
FUO Study Group. Fever of unknown origin (FUO) II: diagnostic
procedures in a prospective multicenter study of 167 patients.
Medicine 1997;76:401e4.
10. Durack DT, Street AC. Fever of unknown origin e re-examined
and redefined. Curr Clin Top Infect Dis 1991;11:35e51.
11. Horowitz HW. Fever of unknown origin or fever of too many or-
igins? N Engl J Med 2013;368:197e9.
12. Jacobs RF, Schutze GE. Bartonella henselae as a cause of pro-
longed fever and fever of unknown origin in children. Clin
Infect Dis 1998;26:80e4.
13. Arisoy ES, Correa AG, Wagner ML, Kaplan SL. Hepatosplenic
cat-scratch disease in children: selected clinical features and
treatment. Clin Infect Dis 1999;28:778e84.
14. Chow A, Robinson JL. Fever of unknown origin in children: a
systematic review. World J Pediatr 2011;7:5e10.
15. Bottieau E, Clerinx J, van den Enden E, van Esbroeck M,
Colebunders R, van Gompel A, et al. Fever after a stay in the
tropics. Diagnostic predictors of the leading tropical condi-
tions. Medicine 2007;86:18e25.
16. Akpede GO, Abiodun PO, Sykes RM. Acute fevers of unknown
origin in young children in the tropics. J Pediatr 1993;122:
79e81.
17. Zimmerli W. Vertebral osteomyelitis. N Engl J Med 2010;362:
1022e8.
18. Bernard L, Dinh A, Ghout I, Simo D, Zeller V, Issartel B, et al.
Antibiotic treatment for 6 weeks versus 12 weeks in patients
with pyogenic vertebral osteomyelitis: an open-label, non-
inferiority, randomized, controlled trial. Lancet 2015;385:
875e82.
19. Institute of Medicine. Beyond myalgic encephalomyelitis/-
chronic fatigue syndrome: redefining an illness. Report Guide
for Clinicians. Institute of Medicine of the National Academies.
www.iom.edu/MECFS. [accessed 8.01.16].
20. Winker R, Barth A, Bidmon D, Ponocny I, Weber M, Mayr O,
et al. Endurance exercise training in orthostatic intolerance:
a randomized, controlled trial. Hypertension 2005;45:391e8.
Undifferentiated fever in children
21. Clarke DA, Medow MS, Taneja I, Ocon AJ, Stewart JM. Initial
orthostatic hypotension in the young is attenuated by static
handgrip. J Pediatr 2010;156:1019e22.
22. Powell KR. Evaluation and management of febrile infants
younger than 60 days of age. Pediatr Infect Dis J 1990;9:
153e7.
23. Herr SM, Wald ER, Pitetti RD, Choi SS. Enhanced urinalysis im-
proves identification of febrile infants ages 60 days and
younger at low risk of serious bacterial illness. Pediatrics
2001;108:866e71.
24. Baraff LJ, Bass JW, Fleisher GR, Klein JO, McCracken Jr GH,
Powell KR, et al. Practice guideline for the management of in-
fants and children 0 to 36 months of age with fever without a
source. Ann Emerg Med 1993;22:1198e210.
25. Pantell RH, Newman TB, Bernzweig J, Bergman DA,
Takayama JI, Segal M, et al. Management and outcomes of
care of fever in early infancy. JAMA 2004;291:1203e12.
26. Caviness AC, Demmler GJ, Almendarez Y, Selwyn BJ. The prev-
alence of neonatal herpes simplex virus infection compared
with serious bacterial illness in hospitalized neonates. J Pe-
diatr 2008;153:164e9.
27. Long SS, Pool TE, Vodzak J, Daskalaki I, Gould JM. Herpes sim-
plex virus infection in young infants during 2 decades of
empiric acyclovir therapy. Pediatr Infect Dis J 2011;30:
556e61.
28. Curfman A, Glissmeyer EW, Ahmad FA, Korgenski EK,
Blashke AJ, Byington CL, et al. Initial presentation of neonatal
herpes simplex virus infection. J Pediatr 2016;172:121e6.
29. Shah SS, Aronson PL, Mohamad Z, Lorch SA. Delayed acyclovir
therapy and death among neonates with herpes simplex virus
infection. Pediatrics 2011;128:1153e60.
30. Aronson PL, Thurm C, Alpern ER, Alessandrini EA, Williams DJ,
Shah SS, et al. Variation in care of the febrile young infant <90
days in US pediatric emergency departments. Pediatrics 2014;
134:667e76.
31. Melvin AJ, Mohan KM, Schiffer JT, Drolette LM, Magaret A,
Corey L, et al. Plasma and cerebrospinal fluid herpes simplex
virus levels at diagnosis and outcome of neonatal infection. J
Pediatr 2015;166:827e33.
32. Teele DW, Pelton SI, Grant MJA, Herskowitz J, Rosen DJ,
Allen CE, et al. Bacteremia in febrile children under 2 years
of age: results of cultures of blood of 600 consecutive febrile
children seen in a “walk-in” clinic. J Pediatr 1975;87:227e30.
33. Fleisher GR, Rosenberg N, Vinci R, Steinberg J, Powell K,
Christy C, et al. Intramuscular versus oral antibiotic therapy
for the prevention of meningitis and other bacterial sequelae
in young, febrile children at risk for occult bacteremia. J Pe-
diatr 1994;124:504e12.
34. Long SS. Antibiotic therapy in febrile children: “best-laid
schemes.”. J Pediatr 1994;124:585e8.
35. Jhaveri R, Byington CL, Klein JO, Shapiro ED. Management of
the non-toxic-appearing acutely febrile child: a 21st century
approach. (Medical Progress). J Pediatr 2011;159:181e5.
36. Steens A, Bergsaker MA, Aaberge IS, Ronning K, Vestrheim DF.
Prompt effect of replacing the 7-valent pneumococcal conju-
gate vaccine with the 13-valent vaccine on the epidemiology
Please cite this article in press as: Long SS, Diagnosis and management of undifferentiated fever in children, J Infect (2016), http://
dx.doi.org/10.1016/j.jinf.2016.04.025
9
of invasive pneumococcal disease in Norway. Vaccine 2013;
31:6232e8.
37. Guevara M, Ezpeleta C, Gil-Setas A, Torroba L, Beristain X,
Aguinaga A, et al. Reduced incidence of invasive pneumococcal
disease after introduction of the 13-valent conjugate vaccine
in Navarre, Spain, 2001e2003. Vaccine 2014;32:2553e62.
38. Tam P-YI, Madoff LC, Coombes B, Pelton SI. Invasive pneumo-
coccal disease after implementation of 13-valent conjugate
vaccine. Pediatrics 2014;134:210e7.
39. Harboe ZB, Dalby T, Weinberger DM, Benfield T, Mølbak K,
Slotved HC, et al. Impact of 13-valent pneumococcal conjugate
vaccination in invasive pneumococcal disease incidence and
mortality. Clin Infect Dis 2014;59:1066e73.
40. Moore CE, Paul J, Foster D, Mahar SA, Griffiths D, Knox K, et al.
Reduction of invasive pneumococcal disease 3 years after the
introduction of the 13-valent conjugate vaccine in the Oxford-
shire region of England. J Infect Dis 2014;210:1001e12.
41. Lepoutre A, Varon E, Georges S, Dorléans F, Janoir C,
Gutmann L, et al. Impact of the pneumococcal conjugate vac-
cines on invasive pneumococcal disease in France, 2001e2012.
Vaccine 2015;33:359e66.
42. Camacho-Badilla K, Falleiros-Arlant LH, Brea J, Avila-Aguero ML.
Challenges in the surveillance of invasive pneumococcal disease
in the postvaccination era. J Pediatr 2015;4:91e3.
43. Moore MR, Link-Gelles R, Schaffner W, Lynfield R, Lexau C,
Bennett NM, et al. Effect of use of 13-valent pneumococcal
conjugate vaccine in children on invasive pneumococcal dis-
ease in children and adults in the USA: analysis of multisite,
population-based surveillance. Lancet 2015;15:301e9.
44. Ben-Shimol S, Greenberg D, Givon-Laevi N, Schlesinger Y,
Somekh E, Aviner S, et al. Early impact of sequential introduc-
tion of 7-valent and 13-valent pneumococcal conjugate vac-
cine on IPD in Israeli children <5 years: an active prospective
national surveillance. Vaccine 2014;32:3452e9.
45. Farnham AC, Zimmerman CM, Papadouka V, Konty KJ,
Zucker JR, Nattanmai GV. Invasive pneumococcal disease
following the introduction of 13-valent conjugate vaccine in
children in New York City from 2007 to 2012. JAMA Pediatr
2015;169:646e52.
46. Kaplan SL, Barson WJ, Lin PL, Romero JR, Bradley JS, Tan TQ,
et al. Early trends for invasive pneumococcal infections in chil-
dren after the introduction of the 13-valent pneumococcal
conjugate vaccine. Pediatr Infect Dis J 2013;32:203e7.
47. Yildirim I, Shea KM, Little BA, Silverio AL, Pelton SI. Vaccina-
tion, underlying comorbidities, and risk of invasive pneumo-
coccal disease. Pediatrics 2015;135:495e503.
48. Colvin JM, Muenzer JT, Jaffe DM, Smason A, Deych E,
Shannon WD, et al. Detection of viruses in young children
with fever without an apparent source. Pediatrics 2012;130:
e1455e62.
49. Roberts KB, Subcommittee on Urinary Tract Infection, Steering
Committee on Quality Improvement and Management.
American Academy of Pediatrics. Urinary tract infection: clin-
ical practice guidelines for the diagnosis and management of
the initial UTI in febrile infants and children 2 to 24 months.
Pediatrics 2011;128:595e610.