 |
  |
|
|
|
 |
COVER ARTICLE/CME Urgent Care 2(2):20, 2007 Fever. In most urgent care settings, it’s number one on the top 10 list of patient presentations in young children. A careful history and complete physical examination will often reveal the cause of a child’s fever, which is usually a viral infection. But for patients who have a fever without a source (FWS), deciding which supplemental evaluations to use can be vexing and controversial. Your main objectives? To determine if a serious bacterial infection (SBI) is present and whether the evaluation can be completed in your facility or requires transfer to an emergency department or tertiary pediatric center. VACCINES AND SBI DECLINE Streptococcus pneumoniae is the most common organism causing invasive bacterial disease outside the neonatal period. Before the present vaccine became available, it was the cause of most positive blood cultures obtained from young febrile children in emergency departments. The pneumococcal conjugate vaccine includes coverage for the seven serotypes responsible for most invasive pneumococcal disease. Since routine vaccination began, studies have shown dramatic declines in invasive pneumococcal infections and even greater reductions in vaccine-covered serotypes. How this will change our approach to diagnosing and treating the febrile child is still uncertain. As the incidence of SBI continues to decline, current management strategies for some age groups will stop being cost effective. Many practitioners are advocating a less aggressive approach in evaluating the immunized febrile child and are adopting such practices. However, others are urging caution. They argue that not all serotypes are covered by the vaccine, and some children won’t have protective antibodies because they were incompletely immunized or had an inadequate immune response to the vaccine. This two-part article presents practical guidelines for evaluating febrile patients up to age 36 months, including a review of recent developments in diagnostic approaches and management. Infants up to three months old are discussed in this first article. DEFINITION OF FEVER A smaller study in 1993 measured the rectal temperatures of 691 healthy infants under three months old in a well-baby clinic. If fever is defined as being two standard deviations above the mean, the study suggests that fever in neonates is 100.4°F, fever in infants one to two months old is 100.5°F, and fever in infants two to three months old is 100.7°F. Temperature may be measured in several ways. Chances are, you measure temperatures in children in your clinic with digital axillary or tympanic membrane thermometers. However, readings using these methods have been found to be inconsistent. Rectal temperature remains the standard, especially in younger infants. Obtaining an accurate temperature measurement is crucial and will determine if an infant meets the threshold of most fever evaluation guidelines. TACTILE FEVER AND FEVER DOCUMENTED AT HOME Neonates and young infants who present with no fever in the office but have a history of tactile or documented fever at home pose an additional clinical challenge. Although we don’t want to miss an SBI, we also don’t want to subject a child to an invasive workup unnecessarily. Several studies have shown that a parent’s ability to assess fever by palpation is generally good. But there are pitfalls to this method. For example, excessive bundling can raise skin temperature (but not rectal temperature), so an infant may feel feverish to the touch when his temperature is actually normal. If a neonate or young infant has a history of tactile fever but his physical examination is normal, consider taking several temperatures over one to two hours. Laboratory studies are generally not indicated if repeat temperature measurements and clinical evaluation remain normal. You can discharge the infant with short-term monitoring of rectal temperature at home and close clinical follow-up. A young infant with a history of fever by rectal temperature at home needs to be evaluated carefully, even if the patient is afebrile on presentation. One study of 292 young infants admitted for fever revealed that, of those with a history of elevated rectal temperature at home but not in the emergency department, only eight out of 40 (20%) had a fever in the hospital, but four out of 40 (10%) had an SBI. A large office-based study of infants younger than three months old reported that, out of 835 who were febrile at home but not in the office, only six (0.7%) had bacteremia or bacterial meningitis, or both. However, in this study, these patients accounted for nearly 10% of all febrile infants identified with these two serious bacterial infections. So generally, a home temperature assessment of fever should be treated the same as one obtained in the clinic setting. TAKING A HISTORY A thorough history and physical examination are invaluable in assessing the febrile pediatric patient. Although the differential diagnosis of fever includes many noninfectious causes, the etiology in young children is almost always infectious. Focus your evaluation on identifying whether an SBI is present. Although viral infections can cause serious consequences, bacterial infections are usually associated with worse outcomes. When taking a history, be sure to identify patients who may be at higher risk for SBI. This includes children who have chronic illnesses (cystic fibrosis, diabetes, or congenital heart anomalies), a history of prematurity, or indwelling medical devices (venous access catheters or ventriculoperitoneal shunts). Children who are immunocompromised from prolonged steroid use or diseases such as leukemia, sickle cell disease, or HIV infection are at increased risk as well. These patients warrant closer scrutiny and often need further workup. Also be alert for febrile patients with a history of recent antibiotic use. One retrospective case series of children younger than two years old with bacterial meningitis found that 83 of 258 (32%) had taken oral antibiotics for more than two days before the meningitis was identified. Unidentified meningitis patients who had been placed on oral antibiotics were less likely to exhibit fever and altered mental status. They also had a longer duration of other symptoms before diagnosis, with more frequent vomiting and infections of the ears, nose, and throat. So consider the possibility of partially treated meningitis when evaluating febrile infants taking antibiotics. Another important consideration is the height of the fever, because the incidence of SBI increases with higher temperatures, especially in young infants. In fact, 38% of infants younger than three months who have temperatures of 104°F or higher have an SBI. The duration of the fever, the method of temperature assessment, and antipyretic use should all be noted. However, the duration of the fever won’t help identify young children with occult bacteremia, and whether or not a patient responds to antipyretics won’t help differentiate bacterial from viral infections. During the history, ask caretakers if the child has other signs and symptoms associated with SBI, such as coughing, vomiting, or ear pulling. In addition, ask about the child’s immunization status, day-care attendance, and exposure to specific infectious agents. Day-care center attendance and recurrent ear infections have been identified as risk factors for invasive pneumococcal infections in children younger than two years. PHYSICAL EXAMINATION POINTERS Initially, your most critical task is to assess the overall appearance of the child. The sicker or more toxic a febrile infant appears, the more the likelihood of an SBI. Regardless of age, the toxic-appearing young child requires an aggressive workup, empirical antibiotic administration, and hospitalization. Inexperienced clinicians may not feel confident assessing a child’s appearance. A thorough knowledge of the parameters of normal, age-specific behavior and activity helps in assessing ill infants and identifying subtle deficits. Although for experienced clinicians this process is partly intuitive, it’s also based on specific aspects of an infant’s appearance and behavior. The Yale Observation Scale (see table below) can supplement your subjective impressions of a febrile child. This tool identifies six observations that are significant and independent predictors of serious disease: quality of cry, reaction to parental stimulation, state of alertness, color, hydration status, and response to social overtures. To obtain a total score, add the scores for all six items. A study of this scale found that only 2.7% of patients with scores of 10 or lower had a serious illness, but 92.3% with scores of 16 or higher had a serious infection. Although this scale won’t identify all infants with an SBI, it’s still a helpful clinical tool.
The physical examination may reveal a recognizable viral or bacterial infection. Viral problems, such as stomatitis, croup, chickenpox, and influenza A, have lower associated rates of bacteremia and other SBIs. Further diagnostic workup often isn’t required when such infections are identified. If you suspect a focal bacterial infection, such as pneumonia, meningitis, or osteomyelitis, evaluate the patient to identify the specific infection and start directed antibiotic therapy. The exception to this rule is the treatment of neonates and infants under three months old, who often require a more aggressive diagnostic evaluation. Because they have the same incidence of bacteremia, febrile children with otitis media should undergo the same diagnostic testing as children with fever only. Always consider further evaluation when focal findings don’t adequately explain the severity of the patient’s symptoms. APPROACH TO THE FEBRILE NEONATE Febrile neonates (up to 28 days old) are at particularly high risk for bacterial infections. An immature immunologic system makes them vulnerable to more virulent bacteria, such as group B Streptococci, Escherichia coli, and Listeria monocytogenes, as well as viral infections such as herpes simplex. S. pneumoniae infections are infrequent in this age group but, when present, they have a high mortality rate (14%). Most febrile neonates seen in pediatric emergency departments have viral infections, but approximately 12% have an SBI, usually a urinary tract infection (UTI) or occult bacteremia. Being able to identify neonates unlikely to have a serious infection will decrease costs, family stress, and iatrogenic complications that can result from hospitalization. Several studies have applied the three commonly known risk-stratification strategies (known as the Rochester, Philadelphia, and Boston criteria) to the neonatal age group (see table below). These criteria use the history, physical examination, and laboratory test results to categorize patients as low or high risk for SBI and identify those who need less aggressive treatment. Although helpful in other age groups, these criteria have been found insufficient in determining which neonates are truly low risk for SBI.
The routine workup for febrile neonates must be more aggressive than in other age groups because of the greater probability of an SBI and our inability to predict which neonates have serious infections. This age group is still at risk for SBI even when other viral infections, such as respiratory syncytial virus, have been identified. All febrile neonates should have a blood culture drawn, a lumbar puncture for cerebrospinal fluid culture and studies, and urine obtained by catheterization for urinalysis and culture (see algorithm below). The higher contamination rate of bagged urine specimens makes them inadequate in young infants. A urine culture is essential because rapid urine testing won’t detect all UTIs. Chest films should be taken when respiratory symptoms are present. Stool leukocytes and culture are indicated in neonates with diarrhea. A white blood cell (WBC) count is usually obtained, but it shouldn’t influence the type of workup you perform because it doesn’t adequately differentiate SBI from other causes of fever in neonates.
All febrile neonates should be hospitalized and receive intravenous antibiotics, even when laboratory screening tests are normal. Treatment should include IV gentamicin (2.5 mg/kg) or a third-generation cephalosporin such as cefotaxime (50 or 100 mg/kg if meningitis is suspected). Ceftriaxone isn’t recommended for young neonates because it may cause unconjugated hyperbilirubinemia. In addition, ampicillin (50 or 100 mg/kg if meningitis is suspected) is used empirically for coverage of L. monocytogenes. Neonatal herpes simplex virus (HSV) infection also may have significant morbidity and mortality but usually isn’t associated with fever. High-dose IV acyclovir (20 mg/kg) improves outcomes in infected neonates and should be considered for febrile neonates with risk factors for HSV. These risk factors include primary maternal infection (especially in newborns delivered vaginally), fetal scalp electrode use, seizures, cerebrospinal fluid pleocytosis, prolonged rupture of the membranes, and skin, eye, or mouth lesions. managing the febrile young infant In the past, a fairly aggressive approach was used in evaluating febrile patients one to three months old, including a full sepsis workup, hospitalization, and usually intravenous antibiotics. This was understandable given the efficacy of antibiotics in decreasing the morbidity of H. influenzae infections and other SBIs. Also, the clinical information obtained on physical examination of the febrile young infant, as with the neonate, is still quite limited. Although inadequate for neonates, the Rochester, Boston, and Philadelphia criteria referred to earlier have successfully differentiated febrile young infants at low risk for SBI from those at high risk in multiple studies. These risk-stratification strategies help identify infants one to three months old who can be managed less aggressively. Laboratory testing is required in virtually all febrile infants aged one to three months because the physical examination will miss a substantial number of SBIs. Pneumococcal infections are a minority in this age group, so the PCV7 vaccine is unlikely to change the incidence of SBI or the need for laboratory evaluation. Diagnostic testing for febrile young infants is similar to neonatal testing. Tests include a WBC count with differential, catheterized urinalysis and urine culture, blood culture, stool leukocytes and culture if diarrhea is present, and chest films if respiratory symptoms are present. A lumbar puncture probably isn’t required on all patients but must be strongly considered. Although bacterial meningitis is relatively rare in infants younger than three months (4.1 per 1000 febrile infants), the WBC count and clinical examination aren’t reliable in ruling out meningitis in this age group. Test results will help you categorize the febrile young infant as high or low risk for SBI and determine subsequent treatment and disposition. A WBC count of 15,000 or higher or 5000 or lower is considered abnormal, as is a band-to-neutrophil ratio of 0.2 or higher. Abnormal urine findings include a positive leukocyte esterase or nitrite on dipstick testing, 5 WBC/high-power field (hpf) or higher on microscopy, or bacteria seen on a Gram-stained, uncentrifuged urine sample. Stool samples should have less than 5 WBC/hpf. Chest films shouldn’t show evidence of pneumonia. Cerebrospinal fluid analysis should reveal less than 8 WBC/hpf and no organisms on Gram stain. Any of these abnormalities places the infant in the high-risk category for SBI. A documented viral infection doesn’t eliminate the likelihood of SBI, but it does seem to lower the risk. For example, young infants with documented viral infections categorized as high risk under the Rochester criteria had an SBI rate of 4.2% compared to 12.3% in those with no identifiable source of infection. Urinary tract infections consistently have been identified as being the most common SBI associated with known viral infections in young infants. This underscores the importance of obtaining urine studies on all febrile infants in this age group. Hospitalization and antibiotic therapy are recommended for any febrile young infant who appears ill or has an abnormal laboratory result (see algorithm below). The usual treatment is IV or IM ceftriaxone (50 or 100 mg/kg if meningitis is suspected). Also consider other antibiotics because of the increasing incidence of pneumococcal antibiotic resistance and the possible presence of organisms not sensitive to third-generation cephalosporins, such as enterococci, Listeria, and gram-positive cocci. Vancomycin in particular is most frequently added to the regimen, especially in infants who appear sick.
Outpatient management is a reasonable alternative for well-appearing, full-term infants in this age group who have normal test results. Young infants with UTIs who aren’t ill also may be candidates for home treatment. However, before discharge, make sure that the family can easily return for reevaluation if the infant gets worse, that reliable follow-up can occur within 24 hours, and that both the primary health care provider and the family agree with the plan. Withholding antibiotics is acceptable in low-risk, febrile young infants. Other treatment options are a ceftriaxone dose before discharge or oral amoxicillin until cultures are negative. Keep in mind that these patients shouldn’t receive antibiotics if they haven’t had a spinal tap because this will make follow-up evaluation for meningitis much more difficult in an infant who remains febrile. Next month: Evaluation of the 3- to 36-month-old child.
To go to the test page where you can earn your credits online, click here. CONTINUING MEDICAL EDUCATION CME Mission Statement The American Academy of Urgent Care Medicine (AAUCM) is committed to bringing change in health care provider behavior through innovative educational programs that improve patient care. CME Needs
Target Audience
Accreditation Credit Designation Disclosure Policy Evidence-Based Content Statement
march 2007 CME Learning Objectives: |
||||||||||
| Copyright ©2006-2007 Quadrant
HealthCom Inc., Parsippany, NJ, USA. All rights reserved. |
){kind=small}
){kind=small}
){kind=small}
){kind=small}
){kind=link}