Office-Based Treatment and Outcomes for Febrile Infants With Clinically Diagnosed Bronchiolitis
Luginbuhl LM, Newman TB, Pantell RH, Finch SA, Wasserman RC
Pediatrics. 2008;122:947-954
Summary
The authors note that there is conflicting evidence in regard to whether febrile infants with bronchiolitis are at lower risk for serious bacterial infection (SBI) and can therefore be treated differently from febrile infants with no obvious focus of infection. Much of the data on SBI rates in febrile infants with bronchiolitis has been obtained in acute-care settings (mostly in emergency departments), so it is also unclear whether the outcomes of those studies reflect outcomes that would be seen in office practices. Office practitioners are generally less likely to obtain diagnostic testing in febrile infants, both with and without bronchiolitis.
This study was completed to determine whether the risk for SBI in patients in office practices is different from that of patients in acute-care settings. The authors evaluated 3 outcomes: frequency of sepsis evaluation in febrile infants with bronchiolitis; frequency of empirical antibiotic use; and frequency of SBI in febrile patients with bronchiolitis.
This study was conducted by Pediatric Research in Office Settings (PROS) practices (219 sites). Infants 3 months of age or younger were enrolled from 1995 to 1998. All had temperatures of at least 38° C. The infants were treated according to each practitioner's usual practice, but patient-enrollment instruments collected standard demographic and clinical information on the patients. The final study cohort included 3066 infants.
It is worth noting that the practices made treatment decisions before most laboratory testing results were available. Seven percent of patients (n = 218) were diagnosed with bronchiolitis. In comparing febrile infants with bronchiolitis vs those without, infants with bronchiolitis were younger (by 1.2 weeks), were more likely to be enrolled in Medicaid, and were more likely to have ill family members. The practitioners were less likely to conduct sepsis evaluations (14% vs 28%) and were less likely to obtain urine or cerebrospinal fluid culture, independently. Children with bronchiolitis were more likely to be admitted to the hospital on the day of initial presentation (47% vs 34%), were more likely to have chest radiographs performed and with positive findings, and were more likely to have abnormal oxygen saturation levels.
There were 3 predictors of patients receiving a complete sepsis work-up: younger age (odds ratio: 7.2 for children 1-30 days old), higher temperature (odds ratio: 6.0 for children with temperature ≥ 39° C), and the fact that respiratory syncytial virus testing was completed, regardless of result. Among infants with bronchiolitis, an ill appearance, age younger than 30 days, and general distress were associated with higher use of parenteral antibiotics. Of the 218 infants with clinically diagnosed bronchiolitis, 125 had cultures and none of these had SBI. The upper limits of the confidence intervals for different SBIs in infants with bronchiolitis were approximately 2.5%, except for urinary tract infection, which was 4.2%. Overall, practitioners were half as likely to perform full sepsis evaluations on febrile infants with bronchiolitis compared with febrile infants without.
The authors concluded that SBIs are rare in infants with bronchiolitis treated in office settings and that the more limited diagnostic and testing approach conducted by office practitioners appears to be appropriate.
Viewpoint
The authors point out that their study cannot be directly compared to the bulk of studies on febrile infants, for several reasons: First, the patients in this study were a different patient cohort, comprising those found in office practices rather than febrile infants presenting to acute-care settings. Second, most of the acute-care studies required that all patients receive full sepsis evaluations in order to calculate rates of SBI. Not all infants were evaluated for SBI in this study, but all were required to have follow-up. Therefore, this study replicated one of the options for managing febrile infants: no testing but subsequent close follow-up. At least in regard to patients with bronchiolitis, such an approach appears to be appropriate.
Abstract