• Enterovirus;
  • fever without source;
  • infant;
  • invasive bacterial infection;
  • PCR


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Transparency Declaration
  8. References

Clin Microbiol Infect 2012; 18: 856–861


Infants under 3 months of age with fever without source (FWS) generally undergo a full, invasive septic evaluation to exclude invasive bacterial infection (IBI). Enterovirus (EV) infections are mostly banal and self-limiting and show a high prevalence rate at this age. We aimed to investigate the prevalence of IBI in EV-infected and uninfected infants under 3 months of age with FWS. This was a prospective observational cohort study of infants aged <90 days who were admitted because of FWS. As per protocol, blood and urine analysis and culture were obtained in all cases, and RNA EV from blood and/or cerebrospinal fluid samples was determined by real-time PCR. Three hundred and eighty-one previously healthy infants with FWS were included. EV infection was diagnosed in 64 children (16.8%; 95% confidence interval, 13.2–20.9%) and showed an uneventful evolution in all cases. Laboratory markers of infection were consistently lower in EV-infected patients; only one case of IBI (1.6%) was observed in an EV-infected patient as compared with 25.2% in EV-negative infants (p <0.001). Intravenous antibiotic use and length of stay were no different in EV-infected and uninfected patients. In our study, febrile infants (<90 days) diagnosed with EV infection showed a low risk of IBI when compared with uninfected patients. The systematic investigation of EV infection in young infants with FWS may allow a more conservative approach to the management of these patients. Further studies on this diagnostic approach are needed.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Transparency Declaration
  8. References

Enteroviruses (EV) are single-strained RNA viruses that belong to the Picornaviridae family, a large virus family composed of more than 70 serotypes. These viruses remain a major cause of febrile illness in children. Most of these infections are subclinical but they can also cause severe disease [1,2]. The highest attack rates occur in children aged <1 year. The reported prevalence of EV infection in patients <12 months with a febrile event ranges from 15% to 50% [1,3,4].

Fever without source (FWS) is one of the most frequent reasons for consultation in paediatric emergencies, especially in infants younger than 3 months. These patients have very unspecific clinical signs, fever commonly being the only clinical manifestation of the disease. About 8.5% of infants aged under 90 days with fever will develop a potentially severe invasive bacterial infection (IBI) [5,6]. Considering these data, and according to published guidelines [7], these patients usually undergo aggressive diagnostic and therapeutic procedures in the emergency room. EV infections may mimic bacterial disease and often lead to unnecessary diagnostic testing and antibiotic use, until a bacterial infection has been ruled out. The ability to diagnose viral infections has improved substantially in recent years, and because most febrile infants are presumed to have viral infections, reaching a specific diagnosis could contribute significantly to their management. In fact, it is known that the diagnosis of a viral infection reduces the risk of suffering from IBI [6,8]. Similarly, we recently reported a lower risk of positive blood or urine cultures among infants with a positive rapid influenza test [9], as previously described in febrile infants with respiratory syncytial virus (RSV) infection [10,11]. In this study, we aimed to determine the prevalence rate of EV infection among febrile infants aged <90 days in our setting, by including an EV real-time polymerase chain reaction (RT-PCR) technique in the routine evaluation of these patients in the Emergency Room. We also aimed to study how often EV and IBI co-infections occur in these patients.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Transparency Declaration
  8. References

Patients and laboratory testing

During a 2-year period (August 2006–2008) prospective data collection of patients <90 days of age admitted because of FWS (a rectal temperature ≥38°C) was carried out in a monographic referral paediatric centre, located in Barcelona (Spain), that attends to about 110 000 patients per year in its Emergency Department. The study was reviewed and approved by the local ethics committee and informed consent was obtained from parents or legal guardians at inclusion.

Our protocol for these patients (after clinical interview and complete physical examination) includes a complete blood count, blood bacterial culture, urinalysis and culture. In all newborns (up to 28 days), and in all infants of any age with malaise or laboratory alteration (white blood count (WBC) ≥15 000/mm3 or ≤5000/mm3 or elevated reactive C protein (CRP) ≥20 mg/L, or elevated procalcitonin (PCT) ≥0.5 ng/mL) [14,15], a lumbar puncture and cerebrospinal fluid (CSF) culture are also performed. All patients are admitted and an intravenous empirical antibiotic is implemented in most cases, except in patients who appear healthy without laboratory alterations.

In addition, for this particular study, EV RNA was determined by means of real-time PCR from blood and/or CSF. RNA extraction was performed using NucliSENS® EasyMAG (BioMérieux, Marcy l'Etoile, France). This method is based on extraction using magnetic spheres. RNA extracts were amplified with the MutaPLATE® Enterovirus real time RT-PCR (TaqMan) kit (Laboratories Immundiagnostik, Bensheim, Germany) according to the manufacturer’s instructions. This assay is a qualitative screening assay containing specific primers, TaqMan probes and additional material for the detection of the 5′-untranslated region (5′-UTR) of the EV genome. This procedure has been shown to be a fast (it can be carried out in 5 h), standardizable, sensitive, specific and reliable method for EV RNA detection [12,13]. It was performed every 48–72 h depending on the number of samples to analyse. The cost of the technique was c. 140 US$ per sample.

Exclusion criteria were as follows: (i) patients with previous chronic disease(s), (ii) premature infants (gestational age <37 weeks at birth), and (iii) patients for whom blood or CSF samples were not available for EV PCR determination (because they were transferred from other centers, or there was insufficient sampling or PCR inhibition).

Enterovirus infection was diagnosed upon the detection of EV RNA in blood, CSF or both. Enteroviral meningitis was defined either by positive EV RNA in CSF (with or without pleocytosis), or by positive blood EV RNA plus CSF pleocytosis (with or without EV RNA positivity in CSF).

The results of all bacterial cultures were obtained from computerized microbiology records. Invasive bacterial infection was defined as follows. (i) Bacteraemia: the presence of a bacterial pathogen in the blood culture; Staphylococcus epidermidis and Streptococcus viridans were considered as contaminating microorganisms. (ii) Urinary tract infection: >10 000 colony forming units/ml in a sterile urine sample. (iii) Bacterial meningitis: defined as the growth of a microorganism in CSF. (iv) Bacterial pneumonia: defined as a suggestive image on chest X-ray, whether blood or pleural positive cultures were associated or not. (v) Bacterial enteritis: diagnosed with the growth of a microorganism in stool culture. (vi) Osteoarticular infection: revealed by suggestive image tests (gammagraphy or magnetic resonance) and/or by a positive culture of the sample obtained from the affected zone. (vii) Infection of soft tissues (cellulites or similar): diagnosed by clinical examination, with or without positive blood or pus cultures.

Data analysis

Data were analysed with the statistical package SPSS v15.0 for Windows (SPSS Inc., Chicago, IL, USA). Continuous variables were described with mean and standard deviation, or median and interquartile range. Discrete variables were described with frequencies and percentage, with a 95% confidence interval. Comparison of means was carried out using the Student t-test for independent samples, while frequencies were analysed by contingency tables and χ2.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Transparency Declaration
  8. References

During the 24-month study period, a total of 381 (179 females, 47%; mean age: 35 days, standard deviation (SD): 21 days) patients with FWS under 3 months of age, fulfilling inclusion criteria, were admitted to our centre and included in the study (Fig. 1). At admission, the mean (SD) of the fever peak was 38.4°C (0.2), with a median of 6 h (p25–75: 3–12 h) evolution. The most common complaints were rhinorrhea (27%), irritability (23%), refusal of food (21%) and gastrointestinal symptoms (19%).


Figure 1.  Algorithm summarizing the patients included in the study according to final results regarding EV infection. EV, enterovirus; CSF, cerebrospinal fluid.

Download figure to PowerPoint

Discharge diagnoses of all patients are summarized in Table 1.

Table 1.   Discharge diagnosis
Diagnosis n = 381EV− (n = 317)EV+ (n = 64)
  1. a Streptococcus agalactiae (n = 6) and Escherichia coli, one patient.

  2. bRotavirus (n = 4), Campylobacter jejunii and Salmonella enteritidis (two cases each).

  3. cInfluenza type A (n = 10) and B (n = 2).

  4. dRespiratory syncitial virus detected in six cases.

Invasive bacterial infection
 Urinary tract infection65641
 Acute gastroenteritisb440
 Cutaneous cellulitis110
Non-invasive bacterial infections: viral or presumable viral infections
 Febrile syndrome without source17414925
 Aseptic meningitis39732
 Acute gastroenteritis27243
 Upper respiratory tract infection24222
 Acute otitis media15150

Blood and urine cultures were performed in all patients. Lumbar puncture was performed in 255 (74.8%) patients. Thirty (7.8%) positive blood cultures were observed; almost half of them (14 out of 30) were considered as contaminated. The remaining 16 blood cultures tested positive for: Streptococcus agalactiae (= 6), Escherichia coli (n = 9) and Proteus mirabilis (n = 1). All gram-negative bacteraemias were secondary to urinary tract infections.

Urine cultures were positive in 65 cases: E. coli in 57 cases, Klebsiella pneumoniae in three cases, and K. oxytoca, Enterobacter cloacae, P. mirabilis, Citrobacter koseri and S. agalactiae in one case each.

Meningitis was diagnosed in 39 cases; all CSF bacterial cultures tested negative. Enteroviral meningitis was diagnosed in 32 out of 39 infants with CSF pleocytosis; in seven cases, no germ was identified.

Blood EV PCR was performed in all but 83 patients, either because of insufficient sampling or PCR inhibition; cerebrospinal fluid EV PCR was performed in 237 infants. Enterovirus infection was diagnosed in 64 patients (56.3% male; prevalence rate, 16.8%; 95% confidence interval (CI), 13.2–20.9%; Fig. 2), 44 (68.8%) of them during March to August (Fig. 2). Enterovirus was considered as the aetiology of FWS in 62 patients: FWS (n = 25), aseptic meningitis (n = 32), acute gastroenteritis (n = 3) and upper respiratory tract infection (n = 2). Co-infection occurred in two patients: a 27-day-old neonate with influenza A and EV co-infection, who was finally diagnosed with flu; and a 37-day-old infant with a urinary tract infection caused by S. agalactiae and an EV co-infection. Three infants, none of them EV-infected, required admission to the intensive care unit: a newborn aged 9 days with hypoxaemia secondary to RSV bronchiolitis that developed after admission; a 47-day-old infant because of urinary sepsis; and an infant aged 2 months, because of aseptic meningitis and secondary apnoeas. No patient died.


Figure 2.  Monthly distribution of EV cases.

Download figure to PowerPoint

Comparison of patients with positive EV PCR and negative EV PCR

Age or clinical symptoms at presentation did not differ between EV-positive and EV-negative patients, except for exanthema (more common among EV-positive patients) and rhinorrhea (less common among EV-positive infants). Laboratory markers of infection were consistently lower in patients infected with EV (Table 2). It is important to note that only two infants with EV infection (finally diagnosed with sterile enteritis and FWS, respectively) showed PCT values higher than 2 ng/mL. The number of patients receiving empirical antibiotics at admission, the length of antibiotic therapy and duration of hospital stay did not differ between EV-positive and EV-negative patients.

Table 2.   Clinical data and test results and management for the two groups
ClinicalEV positive (n = 64)EV negative (n = 317)p
 Fever (mean °C)38.538.4NS
 Evolution (mean hours)10.9410.93NS
Symptoms (%)
 Food refusal15.923.9NS
 Rhinorrhoea 8.630.8<0.001
 Gastrointestinal symptoms1920NS
 Exanthema 6.91.7<0.05
Laboratory tests
 White blood cell count/mm3 (mean)10 29613 038<0.001
 Total neutrophil count/mm3 (mean)42096300<0.001
 Band neutrophil count/mm3 (mean)242487<0.005
 CRP (mg/l) (mean)12.825.7<0.001
 PCT (ng/dL) (mean)0.581.80<0.01
 White CSF cells/mm3 (mean)246.2560.72<0.05
Management and evolution
 Antibiotic treatment75.4%73.4%NS
 Median length of antibiotic treatment (days)2.42.9NS
 Median length of stay (days)3.74.1NS

Prevalence of IBI

As previously mentioned, only one case of IBI was diagnosed, in one of the 64 patients diagnosed with EV infection (a 37-day-old infant with a urinary tract infection caused by S. agalactiae). This represents a 1.6% (95% CI: 0.04–8.4%) prevalence rate for IBI among EV-positive patients, significantly lower than the prevalence of IBI in EV-negative patients (80 out of 317 cases; 25.2%; 95% CI, 20.4–30.0%; p <0.001).


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Transparency Declaration
  8. References

The majority of febrile children who present at primary care and emergency departments are younger than 3 years of age. Twenty per cent of them have FWS, occult bacteraemia occurring in 3% of cases. The percentage of bacterial infections increases in infants aged <3 months, by up to 10% [5,7,16]. The evaluation and management of these infants are controversial because of the non-specificity of clinical symptoms and signs [17]. In these infants, the most prevalent causes of fever are viral infections, which are associated with a reduction in the risk of IBI [6]. Among them, EV remains a major cause of febrile illness [1,17]. However, recent guidelines on the management of infants with FWS do not include a systematic detection of viral infections in these patients [18].

To our knowledge, this is the first prospective study to investigate the role of EV infection in febrile infants under 90 days of age in our geographical area. Ahmed et al. [19] determined that PCR assay of CSF is useful for the rapid and reliable diagnosis of EV meningitis in these young infants, while Ritticher et al. [17] identified EV as the cause of FWS in 20% of young infants by means of EV PCR in blood and CSF samples. Both studies were performed in the United States. In our experience, EV infection was identified as the cause of fever in 64 of the 381 studied subjects; this prevalence rate (16.8%) is similar to that described in the literature [3,6,17,20,21]. Of note, most EV infections were observed from March to August, earlier in time than the classical EV season, which has usually been described as in the summer and fall months [17,22]. Patients diagnosed with EV were mostly newborns (56.3%), as Ritticher [17] described in her series. Twenty per cent of EV-infected patients were younger than 14 days and more than 37% were younger than 28 days. Rotbart [23] also described a high rate of EV infection (47%) in patients under 30 days of age. In our series, only three patients were 7 days old or less at the time of diagnosis of EV infection; most severe cases of EV-related illnesses have been described in these very young infants [2,24–26]. The rate of EV infection among neonates might have been overestimated in our study by the fact that lumbar puncture was systematically performed in all patients under 28 days, but not in those patients aged over 1 month.

In our cohort, 84.2% of cases of meningitis were caused by EV and, among all EV-positive patients, half of them were diagnosed with meningitis, as previously reported [18,19]. It is important to note that 41.6% of EV meningitis did not show pleocytosis, a finding that has been associated with young ages (<90 days) by other authors [17,22].

The IBI prevalence in EV-infected infants has been described as being close to 7%, mostly consisting of urinary tract infections [6,17,18,27]. We only detected one patient affected with a urinary tract infection (that had already been diagnosed and treated in the emergency room) and EV co-infection. Despite this, nearly 75% of EV-infected infants received intravenous antibiotic treatment for at least 72 h, pending the results of bacterial cultures, and length of stay was not different between EV-infected and uninfected children. The fact that EV PCR was performed in our laboratory only every 48–72 h, should not be forgotten. Similarly, Rotbart [23] reported that 94% of their cohort of EV-infected infants received at least one dose of parenteral antibiotic and more than 80% of them were admitted to hospital.

Other studies in EV-infected infants have shown a reduction in the number of days of hospitalization because of early obtaining of PCR results in the first 24 h after admission. Nigrovic and Chiang [28] demonstrated that when the prevalence of EV meningitis is higher than 6%, systematic EV determination by PCR is cost-effective as it prevents unnecessary antibiotic therapy and hospital admissions. King et al. [29], in a retrospective review of 478 patients aged <90 days in whom CSF EV PCR was determined, also concluded that confirmed EV meningitis would potentially reduce the length of stay and the duration of antibiotic treatment, as has been suggested by other authors as well [17,20,30].

Our study has several limitations, including its observational design, the fact that both CSF and blood EV PCR determination were not performed in all patients, and a potential under-diagnosis bias in infants with afebrile EV infection. Finally, other viral co-infections were not systematically investigated and probably were missed in some cases.

In summary, our results suggest that febrile infants (<90 days) diagnosed with EV infection show a low risk of IBI when compared with uninfected patients. The systematic investigation of EV infection in young infants with FWS in the emergency department during March to August, when the prevalence of EV infection is higher in our geographical area, may allow a more conservative approach to the management of these patients, and ultimately, decrease the need for antibiotic use and the length of hospitalization. Until randomized clinical trials on this issue are available, current recommendations on the management of young infants with FWS should be strictly followed.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Transparency Declaration
  8. References
  • 1
    Nigrovic LE. What’s new with enteroviral infections? Curr Opin Pediatr 2001; 13: 8094.
  • 2
    Jordan I, Esteva C, Esteban E, Noguera A, García JJ, Muñoz C. Severe enterovirus disease in febrile neonates. Enferm Infecc Microbiol Clin 2009; 27: 399402.
  • 3
    McCarthy PL, Klig JE, Kennedy WP, Kahn JS. Fever without apparent source on clinical examination. Lower respiratory infections in children and Enterovirus infections. Curr Opin Pediatr 2000; 12: 7795.
  • 4
    Khertsuriani N, LaMonte-Fowlkes A, Oberste MS, Pallansch MA; Centers for Disease Control and Prevention. Enterovirus Surveillance – United States, 1970–2005. Surveillance Summaries, 2006. MMWR Surveill Summ 2006; 55: 120.
  • 5
    Baraff LJ, Oslund SA, Schruger DK, Stephen ML. Probability of bacterial infections in febrile infants less than three months of age: a meta-analysis. Pediatr Infect Dis J 1992; 11: 257264.
  • 6
    Byington CL, Enríquez R, Hoff C et al. Serious bacterial infections in febrile infants 1 to 90 days old with and without viral infections. Pediatrics 2004; 113: 11621166.
  • 7
    Baraff LJ. Management of fever without source in infants and children. Ann Emerg Med 2000; 36: 602614.
  • 8
    Greenes DS, Harper MB. Low risk of bacteremia in febrile children with recognizable viral syndromes. Pediatr Infect Dis J 1999; 18: 258261.
  • 9
    Mintegi S, García García JJ, Benito J et al. Rapid influenza test in young febrile infants for the identification of low-risk patients. Pediatr Infect Dis J 2009; 28: 10261028.
  • 10
    Levine DA, Platt SL, Dayan PS et al. (Multicenter RSV-SBI Study Group of the Pediatric Emergency Medicine Collaborative Research Committee of the American Academy of Pediatrics) Risk of serious bacterial infection in young febrile infants with respiratory syncytial virus infections. Pediatrics 2004; 113: 17281734.
  • 11
    Titus MO, Wright SW. Prevalence of serious bacterial infections in febrile infants with respiratory syncytial virus infection. Pediatrics 2003; 112: 282284.
  • 12
    Rogers SL, Henderson E, Nix WA. Comparative evaluation of Taqman real-time PCR and semi-nested VP1 PCR for detection of enteroviruses in clinical specimens. J Clin Virol 2010; 49: 7374.
  • 13
    Verstrepen WA, Kuhn S, Kockx MM, Van De Vyvere ME, Mertens AH. Rapid detection of enterovirus RNA in cerebrospinal fluid specimens with a novel single-tube real-time reverse transcription-PCR assay. J Clin Microbiol 2001; 39: 40934096.
  • 14
    Fernández López A, Luaces Cubells C, Valls Tolosa C et al. Procalcitonin in the early diagnosis of invasive bacterial infection in febrile infants. An Esp Pediatr 2001; 55: 321328.
  • 15
    Fernández Lopez A, Luaces Cubells C, García García JJ, Fernández Pou J; Spanish Society of Pediatric Emergencies. Procalcitonin in pediatric emergency departments for the early diagnosis of invasive bacterial infections in febrile infants: results of a multicenter study and utility of a rapid qualitative test for this marker. Pediatr Infect Dis J 2003; 22: 895903.
  • 16
    Gajdos V, L’Hélias LF, Mollet-Boudjemline A, Perreaux F, Trioche P, Labrune P. Factors predicting serious bacterial infections in febrile infants less than three months old: multivariate analysis. Arch Pediatr 2005; 12: 397403.
  • 17
    Rittichier KR, Bryan PA, Basset KE et al. Diagnosis and outcomes of enterovirus infections in young infants. Pediatr Infect Dis J 2005; 24: 546550.
  • 18
    Baraff LJ. Management of in infants and young children with fever without source. Pediatr Ann 2008; 37: 673679.
  • 19
    Ahmed A, Brito F, Goto C et al. Clinical utility of the polymerase chain reaction for diagnosis of enteroviral. J Pediatr 1997; 131: 393397.
  • 20
    Byington CL, Taggart EW, Carrol KC, Hillyard DR. A polymerase chain reaction-based epidemiologic investigation of the incidence of nonpolio enteroviral infections in febrile and afebrile infants 90 days and younger. Pediatrics 1999; 103: e27.
  • 21
    Sawyer MH. Enterovirus infections: diagnosis and treatment. Curr Opin Pediatr 2001; 13: 6569.
  • 22
    Seiden JA, Zorc JJ, Hodinka RL, Shah S. Lack of cerebrospinal fluid pleocytosis in young infants with enterovirus infections of the central nervous system. Pediatr Emerg Care 2010; 26: 7781.
  • 23
    Rotbart HA, Mc Cracken GH, Whitley RJ et al. Clinical significance of enteroviruses in serious summer febrile illnesses of children. Pediatr Infect Dis J 1999; 18: 869874.
  • 24
    Lin TY, Kao HT, Hsieh SH et al. Neonatal enterovirus infections: emphasis on risk factors of severe and fatal infections. Pediatr Infect Dis J 2003; 22: 889894.
  • 25
    Wu JM, Wang JN, Tsai YC et al. Cardiopulmonary manifestations of fulminant enterovirus 71 infection. Pediatrics 2002; 109: e26.
  • 26
    Tebruegge M, Curtis N. Enterovirus infections in neonates. Semin Fetal Neonatal Med 2009; 14: 222227.
  • 27
    Finkelstein Y, Mosseri R, Garty BZ. Concomitant aseptic meningitis and bacterial urinary tract infection in young febrile infants. Pediatr Infect Dis J 2001; 20: 630632.
  • 28
    Nigrovic LE, Chiang VW. Cost analysis of enteroviral polymerase chain reaction in infants with fever and cerebrospinal fluid pleocytosis. Arch Pediatr Adolesc Med 2000; 154: 817821.
  • 29
    King RL, Lorch SA, Cohen DM, Hodinka RL, Cohn KA, Shah SS. Routine cerebrospinal fluid enterovirus polymerase chain reaction testing reduces hospitalization and antibiotic use for infants 90 days of age or younger. Pediatrics 2007; 120: 489495.
  • 30
    Ramers C, Billman G, Hartin M, Ho S, Sawer MH. Impact of a diagnostic cerebrospinal fluid enterovirus polymerase chain reaction test on patient management. JAMA 2000; 283: 26802685.