A 2-month-old immunized, previously well infant has a fever for 1 day. He has no associated symptoms. Clinically there is no obvious focus of infection and the child appears well.
Which ONE of the following is CORRECT?
Answer: B: Tepid sponging is not recommended for children with fever according to the 2007 National Institute for Health and Clinical Excellence (NICE) guidelines for managing feverish children.
The risk of serious bacterial illness (SBI) is higher in young infants <3 months of age. Additionally, the clinical clues that are often used to detect serious illness are not reliable in this age group. Although UTI is the most common SBI identified, 1–3% of febrile infants have bacteremia and/or bacterial meningitis. While ill-looking children 1–3 months of age require a full septic work-up, empiric antibiotics and admission, there is some debate about the management and disposition of children in the age group of 1–3 months who appear well. Various strategies (Rochester, Philadelphia and Boston criteria) have been tested to identify a set of low-risk criteria based on clinical and laboratory findings. If these criteria are met, it may allow for less aggressive treatment, withhold empirical antibiotic therapy, or allow management as an outpatient. Unlike the other studies, the Rochester criteria did not include spinal fluid analysis as a routine part of their low-risk criteria, based the attainment of urine cultures upon the results of urinalyses, and included infants younger than 1 month. An extensive discussion of each criteria is beyond the scope of this book. There are, however, no conclusive data to support omission of LP from routine evaluation of fever in this age group. While awaiting further definitive evidence, it is the author’s opinion that the safest approach for most emergency clinicians is to investigate these children fully, including LP and coverage with empiric antibiotics.
PCV7 provides protection against S. pneumoniae but is ineffective against GBS. Furthermore, PCV7 is probably not as effective in this age group, as only the first dose of vaccination would have been given. PCV7 is usually given at 2, 4 and 6 months. The omission of LP in the setting of positive urine in infants <3 months is controversial. The difference in management of a UTI versus meningitis in an infant <3 months is usually in the length of treatment with intravenous antibiotics, with meningitis needing a prolonged length of treatment – 14 days in most cases. The traditional theory states that if a UTI is found during septic work-up, it is a reflection of bacteremic seeding to the urine rather than ascending urine infection, and therefore the infant needs to have CSF culture to exclude this bacteremic seeding to the CSF. There is little evidence for this postulation. Small retrospective studies have shown that healthy, nontoxic-appearing infants with evidence of UTI have bacteremia rates of 6–10%. These studies indicate the risk for serious complications, such as meningitis, is low, and repeatedly show that CSF is negative for pathogens in these patients with UTI diagnoses. These studies conclude that it would appear to be safe, less invasive, and more cost-effective to administer intravenous antibiotics and monitor these patients and to perform LPs only in patients with positive findings on BC or in those whose urine cultures yield a pathogen likely to be associated with meningitis (namely GBS). The author’s conclusion is that this area needs more study to prospectively prove that this is a safe and valid approach. Anecdotally, it is the practice of many paediatric emergency clinicians to omit LP in the setting of a positive UTI, commence empiric antibiotic cover and admit to the ward under a paediatrician who can observe the clinical course and LP at a later stage if concerns of meningitis are raised. This is the practice of the author, in well-appearing non-septic infants 2–3 months age with confirmed UTI on a valid clean catch, catheter or suprapubic urine specimen.
The safest approach once again depends on the clinical experience of the ED clinician in this area; this entails a full septic screen for all under 3 months including LP.
References:
A 3-month-old previously well child has fever for 2 days without a source. The child appears well and has no evidence of sepsis or toxicity. Urine screening is negative, WCC is 8/mm3 , neutrophil count 4/mm3 , CSF is normal, chest X-ray (CXR) is normal and blood culture (BC) is pending.
Answer: D: Current evidence supports the use of screening for serious bacterial infection with full septic work-up including LP, and if all indices are normal, patients can be discharged with or without antibiotics. Routine follow-up care in the ED in 12 hours is mandatory.
Although most of the emphasis in the management of fever focuses on the detection of underlying SBI, it must not be forgotten that certain viruses, in particular HSV, may cause high morbidity and mortality. Congenital HSV infection should be suspected in full-term infants younger than 4 weeks and in premature infants (<32 weeks’ gestation) younger than 8 weeks who have any of the following symptoms: history of HSV lesions in the mother in the third trimester; skin lesions suspicious for HSV on the infant, ill-appearing; seizure associated with the current illness; abnormal liver function test (more than 100 for the AST/ALT); and CSF pleocytosis (bloody, un-interpretable CSF should be considered case by case). It is important to remember that most (60–80%) of the mothers of HSV-positive babies have no known history of HSV infection and a high index of suspicion should be maintained. Acyclovir (60 mg/kg per day, given in divided doses via intravenous infusion) should be empirically administered to all children with suspected congenital HSV infection as mentioned above.
GBS disease is unlikely at this age but can occur up to 90 days after birth. There are two clinical types:
Vaginal or rectal colonization occurs in up to approximately 30% of pregnant women and is the usual source for GBS transmission to newborn infants. In the absence of maternal chemoprophylaxis, approximately 50% of infants born to colonized women acquire GBS colonization, and 1–2% of these infants develop invasive disease.
The IT ratio (immature:total neutrophil) has been in use in neonatal nurseries for two decades as a part of neonatal risk stratification to assess for sepsis. The Philadelphia criteria group was able to improve their sensitivity and negative predictive value (NPV) to 100% for risk stratifying infant 2–3 months into a low-risk group by including the band: total neutrophil count to their screening technique.
A 2-year-old fully immunised boy has a 1-day history of fever to 40°C. He appears clinically well with no clear focus of infection.
Which ONE of the following is the BEST answer?
Answer: C: The current approach to fever with no focus in a child in the age category of 3–36 months who is well appearing, is to perform appropriate urine screening and careful observation in the ED, with a follow-up visit arranged in 12–24 hours either within the ED or with the child’s GP. The approach to these children is dramatically different now after the introduction of PCV7 in the late 1990s. In this ‘post PCV7’ era, the rate of invasive pneumococcal bacteremia and subsequently SBI is <1%, making empiric screening for SBI and testing with WCC and BC cost-ineffective. Furthermore, the incidence of E. coli, Neisseria and Salmonella are all now increased within this <1% occult bacteremia category, and WCC screening is not sensitive for the detection of bacteria other than streptococcus. A WCC that is negative (i.e. <15,000) will therefore have a significant false negative if Neisseria or Salmonella are the offending organisms causing bacteremia. Clinical observation of the child in the ED may be more sensitive.
There is still a definite correlation between high fever >39°C and bacterial, specifically pneumococcus infection. However, the rate of this infection is so low that it makes it statistically more likely that fever >39°C is going to be due to viral causes.
Which ONE of the following is TRUE in relation to measuring the body temperature in a child?
Answer: D: The most accurate method to measure temperature is via a rectal thermometer, particularly in high-risk groups such as infants 0–3 months of age, as axillary, oral or tympanic thermometers are unreliable in this age group. The rectal route should not be used in patients who are potentially immunocompromised due to the risk of mucosal damage, bacteremia or transmission of infection. However, parental acceptance and ease of use in the ED settings may require that axillary and tympanic methods be used instead of this gold standard. The author’s approach to this dilemma is tympanic thermometers for older children, and axillary checks for infants. Electronic or infrared versions are equally accurate.
On average, axillary temperature measurement using an electronic thermometer underestimates body temperature by at least 0.5°C and in some children may underestimate by as much as 2°C. In neonates the axillary route appears to be more accurate, with a difference from rectal temperature of around 0.5°C and a sensitivity of 98%. Tympanic measurement differs on average from body temperature by 0.3°C, with a sensitivity to detect fever ranging from 51% to 97%. Some studies reported that tympanic measurements are difficult or inaccurate in infants under the age of 3 months due to the different anatomy – a smaller, shorter canal where the infrared beam may not access the tympanic membrane accurately.
Regarding the diagnostic approach to children with fever of 39°C without an obvious focus and who are ‘well appearing’ and ‘non-toxic’, which ONE of the following is the BEST answer?
Answer: D: Ceftriaxone is contraindicated in neonates due to concerns about it displacing bilirubin from protein binding sites and inducing kernicterus, as well as the risk of calcium precipitation in neonates with intravenous infusion. Cefotaxime is recommended instead.
A 3-year-old unimmunized child is assumed to have a susceptibility to pneumococcus and haemophilus b (HIB), which predates the era of HIB and PCV7 vaccination. For this reason it is prudent to perform screening for bacteremia even if the child is well appearing. If the screening WCC is negative, the child can be discharged with defined follow-up arranged in 12 hours in the ED, for clinical review as well as BC results.
In children over 3 months of age, C-reactive protein, interleukin and procalcitonin have repeatedly been shown to be of limited value to an ED clinician attempting to risk stratify children who appear well and have fever without source.