Correspondence
Outpatient Management of Fever in Selected Infants
N Engl J Med 1994; 330:938-940March 31, 1994
- Article
To the Editor:
The screening criteria to identify young infants at low risk of serious bacterial infection as originally proposed by Dagan et al.1,2 and subsequently modified and used by others3,4 do not include a chest film or a lumbar puncture. The savings achieved with the approach described by Baker et al. (Nov. 11 issue)5 would be substantially greater if these two diagnostic tests could be dispensed with.
What was the usefulness of the lumbar puncture in the infants who otherwise were at low risk of serious bacterial infection? Did any infant who did not appear to be seriously ill (Infant Observation Score, ≤ 10) and who had a white-cell count below 15,000 per cubic millimeter, a ratio of band forms to neutrophils below 0.2, and a negative microscopical examination of the urine for bacteria or fewer than 10 white cells per high-power field have bacterial meningitis? One of the 285 infants not treated with antibiotics had aseptic meningitis. Did this child meet the other low-risk criteria? In view of the lack of specific therapy and the generally favorable outcome of infants with aseptic meningitis, one could argue that hospitalization is not necessary and that a lumbar puncture is therefore not necessary in febrile infants who appear well6.
The authors should also comment on the need for a chest film, which is not always part of a sepsis workup. They report no cases of bacterial pneumonia among the 747 infants. Did any child who met the modified screening criteria and who had no signs or symptoms of pneumonia have a chest film with an infiltrate suggestive of pneumonia? None had a positive blood culture, so even if an infiltrate was present, it was more likely than not of viral origin.
Finally, during the third year of the study, an additional laboratory criterion was added for entry to the groups receiving no antibiotic treatment. The total number of subjects who met the modified screening criteria is not reported. This information would be of value to other investigators and should be reported so that the 95 percent confidence interval for these modified criteria can be calculated and these data can be combined with similar data for a meta-analysis.
6 ReferencesLarry J. Baraff, M.D.
UCLA School of Medicine, Los Angeles, CA 90024-17441
Dagan R ,Powell KR ,Hall CB ,Menegus MA . Identification of infants unlikely to have serious bacterial infection although hospitalized for suspected sepsis. J Pediatr 1985;107:855-860
CrossRef | Web of Science | Medline2
Dagan R ,Sofer S ,Phillip M ,Shachak E . Ambulatory care of febrile infants younger than 2 months of age classified as being at low risk for having serious bacterial infections. J Pediatr 1988;112:355-360
CrossRef | Web of Science | Medline3
McCarthy CA ,Powell KR ,Jaskiewicz JA , et al. Outpatient management of selected infants younger than two months of age evaluated for possible sepsis. Pediatr Infect Dis J 1990;9:385-389
CrossRef | Web of Science | Medline4
Powell KR . Evaluation and management of febrile infants younger than 60 days of age. Pediatr Infect Dis J 1990;9:153-157
CrossRef | Web of Science | Medline5
Baker MD ,Bell LM ,Avner JR . Outpatient management without antibiotics of fever in selected infants. N Engl J Med 1993;329:1437-1441
Full Text | Web of Science | Medline6
Rorabaugh ML ,Berlin LE ,Heldrich F , et al. Aseptic meningitis in infants younger than 2 years of age: acute illness and neurologic complications. Pediatrics 1993;92:206-211
Web of Science | Medline
To the Editor:
We were disappointed by the lack of use of routine viral diagnostic methods in the study by Baker and his colleagues and by the lack of comment on their absence in the editorial by McCarthy (Nov. 11 issue)1. No fewer than 450 of the children (60.2 percent) were recorded as having had a “viral syndrome,” but there was no mention of confirmatory laboratory tests except in a footnote to 1, which stated that respiratory syncytial virus was identified in 30 of the 34 infants with bronchiolitis.
A variety of respiratory viruses (respiratory syncytial virus; influenza A or B; parainfluenza types 1, 2, 3, and 4; adenovirus; and measles) can be reliably identified by immunofluorescence assay2 or immunoassay3 within two to three hours. Antibiotics can then be withheld safely, investigations limited, and an earlier discharge considered. The diagnosis of a “viral syndrome,” on the other hand, is one of exclusion that does not bolster the clinician's confidence in caring for the patient and encourages the defensive use of antibiotics, further investigations, and longer hospital stays. Some years ago, Stokes et al.4 confirmed the variety of viruses involved in febrile convulsions and highlighted the value of a specific diagnosis.
The diagnosis of a viral illness does not come cheap, and a full assessment may cost up to $120. Nevertheless, the potential savings are even greater. Baker et al. reported a savings of up to $3,100 per patient by omitting antibiotic therapy ($600) and using outpatient management ($2,500). The proportion of patients cared for in this manner could have been higher had viral diagnostic tests been used in all cases.
4 ReferencesC.R. Madeley, M.D.
J.S.M. Peiris, M.B., B.S., M.D.
Royal Victoria Infirmary, Newcastle upon Tyne NE1 4LP, United KingdomA.W. Craft, M.D., B.S.
University of Newcastle upon Tyne, Newcastle upon Tyne NE2 4HH, United Kingdom1
McCarthy PL . Infants with fever. N Engl J Med 1993;329:1493-1494
Full Text | Web of Science | Medline2
Gardner PS, McQuillin J. Rapid virus diagnosis: application of immunofluorescence. 2nd ed. London: Butterworth, 1980.
3
Halonen P, Obert G, Hierholzer JC. Direct detection of viral antigens in respiratory infections by immunoassays: a four year experience and new developments. In: de la Maza LM, Peterson EM, eds. Medical virology IV. Hillsdale, N.J.: Lawrence Erlbaum, 1985.
4
Stokes MJ ,Downham MA ,Webb JK ,McQuillin J ,Gardner PS . Viruses and febrile convulsions. Arch Dis Child 1977;52:129-133
CrossRef | Web of Science | Medline
To the Editor:
We are not convinced that the protocol of Baker et al. can be easily or safely implemented in the average community hospital emergency department. A significant percentage of pediatric outpatient care in this country is provided by nonpediatricians,1 some of whom may be unfamiliar with the often subtle symptoms and signs of serious illness in infants. As Baker emphasizes, assessment of a febrile infant by an experienced pediatrician is absolutely crucial. Realistically, however, outside of the university setting or urban centers, immediate access to a pediatrician is not always possible.
The failure to diagnose meningitis is a common reason for malpractice suits. We are concerned about the doctors who do not possess the clinical skills required to perform or interpret a sepsis workup and who cannot consult a pediatrician about the case in a timely fashion, and who may thus mistakenly discharge some very sick babies. They may find comfort in this study, which contends that only a small percentage of febrile infants have bacteremia2.
Although lauding Baker and colleagues for their thorough and clinically relevant study, we must concur with McCarthy in favoring a “circumspect approach” to the care of infants with fever3. Irrespective of the apparently low rate of bacteremia in this population, we believe that the ostensible risk-benefit ratio continues to support hospitalization for these children or, when appropriate, aggressive treatment on an outpatient basis, given the catastrophic repercussions for infant, family, and physician of a “missed” diagnosis. In all candor, the cumulative cost of all the ceftriaxone administered to all the infants in this age group during any given year could conceivably be less than a single award to a plaintiff in a suit brought for brain damage resulting from the failure to treat meningitis.
4 ReferencesRobert F. Perry, M.D.
East Carolina University School of Medicine, Greenville, NC 27834Andrew P. Garlisi, M.D.
St. Anthony Hospital, Michigan City, IN 463601
The field of pediatrics. In: Behrman RE, ed. Nelson textbook of pediatrics. 14th ed. Philadelphia: W.B. Saunders, 1992:1-6.
2
Caspe WB ,Chamudes O ,Louie B . The evaluation and treatment of the febrile infant. Pediatr Infect Dis 1983;2:131-135
CrossRef | Medline3
McCarthy PL . Infants with fever. N Engl J Med 1993;329:1493-1494
Full Text | Web of Science | Medline4
Baskin MN ,O'Rourke EJ ,Fleisher GR . Outpatient treatment of febrile infants 28 to 89 days of age with intramuscular administration of ceftriaxone. J Pediatr 1992;120:22-27
CrossRef | Web of Science | Medline
Author/Editor Response
The authors reply:
To the Editor: We appreciate Dr. Baraff's interest in our study and the opportunity to clarify the issues he raises. Although a substantial savings would result from the omission of lumbar punctures from sepsis workups in febrile infants, we are convinced that this practice would be unsafe. Since the conclusion of our original study, we have encountered one 29-day-old infant with Haemophilus influenzae type b meningitis who appeared well (Infant Observation Score, 6) and had a white-cell count below 15,000 per cubic millimeter, a ratio of band forms to neutrophils below 0.20, and a normal urinalysis. The only test identifying meningitis in this child was the lumbar puncture.
Although it is true that aseptic meningitis eventually developed in 1 of the 285 infants not treated with antibiotics, this disease was not detectable at the time of the initial screening, when the results of the lumbar puncture were normal. As we indicated, this child appeared more ill on follow-up examination, and a second lumbar puncture indicated meningitis. We are convinced that neither the results of the initial spinal fluid analysis nor the findings on a physical examination alone can be used reliably to distinguish bacterial from aseptic meningitis. Therefore, we continue to include both in all evaluations of febrile infants.
We also contend that a chest film is a necessary part of the sepsis workup for infants. In all 28 infants with pneumonia in the study, discrete infiltrates were apparent on the chest film. Five of these infants met our modified screening criteria. Of these, three had respiratory rates of 40 or less and had no clinical signs of pneumonia other than fever. These data support our continued inclusion of chest films in the evaluation of febrile infants.
Finally, as we discussed in our paper, our modified screening criteria (which included a band-to-neutrophil ratio of less than 0.20) were 100 percent effective in identifying infants with serious bacterial illness. In the 78 months since use of this protocol began, there have been no screening failures. We have confidence in our approach to the febrile infant and continue to monitor its success.
We concur with Drs. Perry and Garlisi that an experienced pediatrician's assessment is an important part of the evaluation of a febrile infant. The physician who is unfamiliar with the subtle symptoms and signs of serious illness in infants should not consider outpatient management as an option with or without the use of antibiotics. It should be emphasized that the use of intramuscular ceftriaxone as an outpatient therapy depends on the outcome of the clinical assessment of the febrile infant and on the selection of a well-appearing low-risk group.1 One other option to consider is admitting infants who fulfull the criteria for a low risk of serious illness but not treating them with antibiotics. This option might be chosen by physicians who feel uncomfortable with their ability to interpret the clinical appearance of these infants.
We appreciate the comments by Madeley et al. Although we support the concept of identifying the cause of disease in febrile infants whenever possible, we do not think that routine testing for common respiratory viruses is necessary at this time in febrile infants who are not hospitalized. Although during certain times of the year rapid viral testing is used for infection-control purposes in hospitalized infants, it is expensive and does not rule in or rule out a virus as the only cause of fever. In general, the sensitivity of immunofluorescence assays for respiratory viruses is 75 to 85 percent (excluding respiratory syncytial viruses). Furthermore, these tests require approximately three hours to complete, and the results often would not change the management plan. Our method of evaluating febrile infants addresses the important issue of the identification of possible bacterial disease. In the future, as the sensitivity of diagnostic methods improves and antiviral therapies become available, this strategy will need to be rethought.
1 ReferencesM. Douglas Baker, M.D.
Louis M. Bell, M.D.
Jeffrey Avner, M.D.
Children's Hospital of Philadelphia, Philadelphia, PA 191041
Baskin MN ,O'Rourke EJ ,Fleisher GR . Outpatient treatment of febrile infants 28 to 89 days of age with intramuscular administration of ceftriaxone. J Pediatr 1992;120:22-27
CrossRef | Web of Science | Medline
