Published online December 1, 2006
PEDIATRICS
Vol. 118
No. 6
December 2006, pp.
2606-2607
(doi:10.1542/peds.2006-2847)
Febrile Confusion: Do the Hyperpyrexia Study Conclusions Fit?: In Reply.
Barbara W. Trautner, MD
Department of Internal Medicine,
Section of Infectious Diseases
Charles G. Macias, MD, MPH
A. Chantal Caviness, MD, MPH
Department of Pediatrics,
Section of Emergency Medicine,
Baylor College of Medicine,
Houston, TX 77030
We appreciate the interest in our study investigating the risk of serious bacterial illness (SBI) in children with hyperpyrexia and the relationship of the clinical presentation in defining those patients at risk.1 Gunasekera et al question the conclusion that "children with hyperpyrexia are at equally high risk for serious bacterial infection and for viral illness" on the basis of a concern that 62 of 103 children may have been ignored or distributed equally into the bacterial and viral groups. They contend that these children are more likely to have clinical courses consistent with viral illness. It is important to note that our intention was to evaluate the risk of SBI using hard end points (specifically, positive cultures). Therefore, our statistical analysis took into account only those patients for whom cultures were positive. No sensitivity analysis was offered for those patients for whom cultures were negative. We agree that cultures are often negative in children with viral illnesses, but we sought to avoid assumptions of classification of those children without culture-proven disease. Our conclusion can be restated to reflect that children with hyperpyrexia are at equal risk for culture-positive bacterial or viral illness. Nonetheless, we agree that the crucial message that antibiotics should be considered for children with hyperpyrexia remains.
The authors of this letter highlight 2 interesting points regarding the role of pneumonia and the role of diarrhea in children with hyperpyrexia. The first is that pneumonia should be included in the category of SBI, thus increasing the risk of SBI from 19.4% to 35.0%. Again, our concern for adhering to hard end points (positive cultures) led to the analysis we provided. As we mentioned in our article, the presence of a lobar infiltrate on chest radiograph does not exclude the presence of viral infection.2,3
Nonetheless, we believe that the inclusion of pneumonia into the category of children with SBI is really most pertinent for those children without underlying risk factors for bacterial illness in which SBI may otherwise go untreated. Of the 84 subjects without an underlying predisposition to bacterial illness, 13 had SBI excluding pneumonia. Only 37 of these had a chest radiograph, and 13 were consistent with lobar pneumonia. Of these 13, 2 were bacteremic, for a total of 24 subjects with SBI (Table 1).
Table 1 demonstrates 2 separate analyses for SBI with and without pneumonia. It should be noted that the odds ratios (ORs) for predictors of SBI are not significantly different in children with and without pneumonia given the ranges of their confidence intervals (CIs).
In regards to the role of diarrhea, Gunasekera et al offer an alternate prevalence calculation for the subgroup and address the role of dysentery. In our cohort, however, of the patients with diarrhea, 3 had concurrent bacteremia and 3 had concurrent urinary tract infection. Therefore, the alternate prevalence estimates offered by Gunasekera et al would apply only if the diarrhea had existed in isolation. With regard to the role of dysentery, if one removes the 1 case of dysentery from the analysis, the OR for SBI (and its corresponding 95% CI) remains significant, indicating that diarrhea is not merely predictive of dysentery. However, we agree that there was potential for an ascertainment bias regarding the association of diarrhea with SBI, because the treating physician decided which subjects required stool bacterial cultures. It is interesting to note that a recent publication studied children who presented with diarrhea to the emergency department in Seattle, Washington, during the same time frame as our study.4 The authors found that physician judgment was accurate at identifying patients for whom bacterial culture would yield positive results (sensitivity: 75%; specificity: 65%).
In our original analysis of all 103 patients with hyperpyrexia (see Table 5 in ref 1), we demonstrated that a preexisting condition significantly increased the risk of SBI (OR: 3.19; 95% CI: 1.06–9.61). Excluding the 19 subjects with underlying conditions left 84 previously healthy subjects. Of these, 13 (15.5%) had an SBI (excluding pneumonia), and 19 (22.6%) had a proven viral illness. However, analysis of this smaller subgroup is severely limited by the small sample size and should be interpreted with caution.
Our study raises questions as it answers others. Although it is the largest series of children with hyperpyrexia studied in recent times, it is still not large enough to permit analysis of subgroups. On the basis of our findings, we believe that children with hyperpyrexia are at high risk for serious bacterial infection and for viral illness and that no aspect of the clinical presentation reliably distinguishes between the 2 entities. Clinical judgment of the treating physician remains paramount in determining the need for antibiotics.
REFERENCES
- Trautner BW, Caviness AC, Gerlacher GR, Demmler G, Macias CG. Prospective evaluation of the risk of serious bacterial infection in children who present to the emergency department with hyperpyrexia (temperature of 106°F or higher). Pediatrics. 2006;118:34–40[Abstract/Free Full Text]
- Michelow I, Olsen K, Lozano J, et al. Epidemiology and clinical characteristics of community-acquired pneumonia in hospitalized children. Pediatrics. 2004;113:701–707[Abstract/Free Full Text]
- Tsolia M, Psarras S, Bossios A, et al. Etiology of community-acquired pneumonia in hospitalized school age children: evidence for high prevalence of viral infections. Clin Infect Dis. 2004;39:681
–686[CrossRef][ISI][Medline]
- Klein EJ, Boster DR, Stapp JR, et al. Diarrhea etiology in a children's hospital emergency department: a prospective cohort study. Clin Infect Dis. 2006;43:807–813[CrossRef][ISI][Medline]
PEDIATRICS (ISSN 1098-4275). ©2006 by the American Academy of Pediatrics
