See Related:
Authors' Articles
 

Return to
Table of Contents

Letter

In Reply

Letter

In Reply

Letter

In Reply

Letter

In Reply

Letter

In Reply

Letter

In Reply

The evaluation of young febrile infants continues to be controversial. At the center of this debate is the utility of the laboratory evaluation, particularly in the presence of an apparent viral source of fever. In their retrospective review, Purcell and Fergie document the low risk of concurrent SBIs.¹ These data add to a growing body of evidence on this topic.^2-4 Although RSV may cause fever,³ an occasional infant with RSV may have a concurrent SBI, typically, a urinary tract infection.^{1, 3, 4}

The more pertinent question, however, is how low a risk of SBI is too low to investigate with laboratory tests? Answering this question requires sophisticated and creative analyses, including cost-effectiveness analyses, an example of which was recently performed on older febrile infants at risk of occult bacteremia.⁵ The article by Purcell and Fergie fails to address this question. Furthermore, the study population was generated by reviewing the medical records of patients based on International Classification of Diseases, Ninth Revision (ICD-9) discharge diagnoses of RSV bronchiolitis or pneumonia. This could have led to a bias because infants who came to the emergency department with fever and wheezing may have been coded as "fever" or "rule-out sepsis." For those found to have urinary tract infections, bacteremia, or meningitis, the likely discharge diagnosis codes would reflect those bacterial infections, not RSV bronchiolitis or pneumonia. This enrollment bias could have been avoided if the authors had also reviewed the medical records of febrile infants and children hospitalized with SBIs during the same period to ensure that none had concurrent RSV infections. In addition, this retrospective data set was gathered for a different objective than that of the study and was missing information about patient clinical appearance and laboratory information. Finally, the strong conclusions that "Performing all of these cultures . . . in infants and children with typical signs and symptoms of RSV bronchiolitis and a positive RSV rapid antigen test, even in the presence of fever, is unnecessary . . . " are overstated given the limitations in the data analyzed by the authors.

The utility and effectiveness of laboratory evaluation of the febrile infant with RSV infection can most accurately be assessed by a prospective study of febrile infants with RSV infections and/or bronchiolitis, with uniform gathering of clinical and laboratory information. Once accurate information about the prevalence of concurrent SBIs is obtained, cost-effectiveness analyses can then be performed to provide the best estimate of the cost-effectiveness of different evaluation strategies.

 
Nathan Kuppermann, MD, MPH
Departments of Pediatrics and Internal Medicine
University of California–Davis School of Medicine
2315 Stockton Blvd
PSSB, Suite 2100
Sacramento, CA 95817
e-mail: nkuppermann@ucdavis.edu
 
 

1. Purcell K, Fergie J. Concurrent serious bacterial infections in 2396 infants and children hospitalized with respiratory syncytial virus lower respiratory tract infections. Arch Pediatr Adolesc Med. 2002;156:322-324. ABSTRACT  |  FULL TEXT  |  PDF  |  MEDLINE

2. Liebelt EL, Qi K, Harvey K. Diagnostic testing for serious bacterial infections in infants aged 90 days or younger with bronchiolitis. Arch Pediatr Adolesc Med. 1999;153:525-530. ABSTRACT  |  FULL TEXT  |  PDF  |  MEDLINE

3. Kuppermann N, Bank DE, Walton EA, Senac MO, McCaslin I. Risks for bacteremia and urinary tract infections in young febrile children with bronchiolitis. Arch Pediatr Adolesc Med. 1997;151:1207-1214. MEDLINE

4. Antonow JA, Hansen K, McKinstry CA, Byington CL. Sepsis evaluations in hospitalized infants with bronchiolitis. Pediatr Infect Dis J. 1998;17:231-236. MEDLINE

5. Lee GM, Fleisher GR, Harper MB. Management of febrile children in the age of the conjugate pneumococcal vaccine: a cost-effectiveness analysis. Pediatrics. 2001;108:835-844.
 

 

We thank Drs McGregor and Tung as well as Dr Kuppermann for their interest and comments on our recently published article.¹ At the center of the controversy is the first sentence of our concluding statement:

Performing all of these cultures (full sepsis/meningitis workups) on admission in infants and children with typical signs and symptoms of RSV bronchiolitis and a positive RSV rapid antigen test, even in the presence of fever, is unnecessary and adds to the cost, discomfort, and stress of the hospitalization.

Drs McGregor and Tung express concern because of their case report. Although they imply that the infant had S pneumoniae meningitis from the time of admission, we cannot be completely sure that the infant did not get infected shortly after admission. We also caution against their implication that not diagnosing meningitis on day 1 of symptoms is associated with a bad outcome, or that the outcome is going to be worse than if meningitis is diagnosed at once and antibiotic therapy is initiated earlier. Although these concepts may seem intuitive, they are not correct.^{2, 3} Additionally, their patient had only minimal rhinorrhea and no reported wheezing and shortness of breath, a state that is not typical of RSV bronchiolitis.

Concurrent SBIs can occur with RSV bronchiolitis, but the situation that Drs McGregor and Tung report is rare based on our review of the literature. We believe that the concern expressed by these physicians helps to explain the widespread use of antibiotics in infants and children with RSV and other viral infections. However, this concern needs to be put into the proper context. Respiratory syncytial virus bronchiolitis is an illness that it is estimated to cause more than 100 000 hospitalizations per year in the United States, and there are only a few case reports of concurrent bacterial sepsis or meningitis, the most potentially serious concurrent bacterial infections. Our intention was to show the low risk of concurrent SBIs and hopefully help curtail the overuse of antibiotics in infants and children with RSV bronchiolitis. As we discussed in our article, most of the serious concurrent bacterial infections reported are urinary tract infections, or to be more precise, urine cultures positive for organisms. This distinction is important, as the number of positive urine cultures reported in ours and in other studies can be explained by the usual incidence of asymptomatic bacteriuria.⁴ The same issue of concurrent urinary tract infections was found in a study of febrile infants aged 1 to 60 days with aseptic meningitis, in which the authors found that 0.7% of these patients also had urine cultures positive for organisms.⁵

Drs McGregor and Tung point to our exclusion of sepsis/meningitis work-ups done after the first day of hospitalization. We wanted to focus only on the practical question of concurrent SBIs present at the time of admission. Clearly, nosocomial infections occur; this was studied in a prospective study by Hall et al.⁶ Drs McGregor and Tung also bring up the issue of cost data. We did not do a cost analysis; the numbers we mentioned are from the study by Antonow et al.⁷ Of course, there will be cost savings from a reduction in the use of diagnostic tests and antibiotics and possibly from a shortened hospitalization if physicians are less concerned about concurrent SBIs in infants and children with typical signs and symptoms and confirmed RSV infection.

Dr Kuppermann questions the validity of our conclusion based on our study design. We admit that our retrospective study has many limitations as stated in our article. However, we believe that our recommendation is consistent with our findings. Also, our findings are in agreement with his prospective study of febrile children aged 24 months or younger with bronchiolitis⁸ and the retrospective study by Liebelt et al⁹ of infants younger than 90 days with bronchiolitis. Of the 805 infants younger than 90 days and the 285 younger than 6 weeks in our study, we found no cases of bacterial sepsis or meningitis. This again is in agreement with the study by Liebelt et al.⁶ Dr Kuppermann suggested the possibility of missing some children with SBIs because of miscoding. We searched the medical records using the International Classification of Diseases, Ninth Edition codes for RSV bronchiolitis (466.11) and RSV pneumonia (480.1). These diagnoses could have been the primary discharge diagnosis or one of many secondary discharge diagnoses. Therefore, even if a child had a primary discharge diagnosis of bacterial sepsis or meningitis, the patient's medical record would still have been identified because of a secondary diagnosis of RSV infection. In relation to the ruled out sepsis diagnosis, we are always reminded by our hospital that this is not a diagnosis and that the coders will have to find an appropriate diagnosis after discussing the case with the attending physician.

Finally, while Drs Kuppermann, McGregor, and Tung take objection with the first sentence of our concluding statement, we would like to emphasize the second sentence: "However, laboratory testing for bacterial infections should be considered in severely ill-appearing infants and children with atypical signs and symptoms or clinical courses due to the small but real possibility of concurrent serious bacterial infections."