The management of febrile infants 7–60 days old is highly variable.10 Even though multiple studies show that CXR is not useful for evaluation of febrile infants with no respiratory symptoms, CXR rates are still between 30-60% this high rate might be explained by fear of missing an occult bacterial pneumonia.1,4,5,6,11,12 In total 65% of the infants had a CXR which is similar to previous studies.1,4,5,6 CXR rates did not differ in infants who had respiratory symptoms and sick contacts, and who had no respiratory symptoms and no sick contacts, 69% and 62%, respectively. All infants with no respiratory symptoms and no sick contact had a negative CXR which was similar to the results of the study by Bramson. They examined 361 febrile infants <3 months of age with no pulmonary symptoms, none had a positive CXR.4 Another study by Crain found that only 2 out of 148 infants with no respiratory symptoms had a positive CXR.11 Sensitivity of having no respiratory symptoms to predict a negative CXR was 93%.11 One study checked physician variations among pediatric emergency physicians and found that CXR rates were between 30-60% regardless of reported sick contacts which, in combination with respiratory symptoms, may be helpful to guide decision making.6
Despite the low probability of a positive finding, many physicians order the chest x-ray in fear of missing an occult pneumonia.13 Rates of occult bacterial pneumonia in this population is about 1-3%.12 One study showed the incidence of pneumonia in children <12 months of age with high fever (>39) without a source and WBC count greater than 20x109/L was 7.8%. However, this study was limited to the selected children who had CXR and WBC ordered.14 In 2003, ACEP released an editorial mentioning that CXR should only be performed in infants who present with respiratory symptoms, abnormal pulse oximetry, hyperpyrexia, or marked leukocytosis.15 Decreasing the rate of CXR in infants with no respiratory symptoms was one of the reasons our institution participated in Project REVISE.
There are no clear guidelines indicating when to perform RSV/flu testing in this population. Studies show that most febrile infants who were diagnosed with flu/RSV had either respiratory symptoms and/or sick contacts.8,9,16,17,18 Identifying influenza and RSV may reduce antibiotic use and decrease unnecessary diagnostic testing.2
Benito-Fernandez found that about 40% of all febrile infants <3 months of age who were tested for flu during the flu season were positive. 75% had a sick contact and 46% had mild respiratory symptoms.9 Another study showed that 34% of flu positive infants had no respiratory symptoms and the majority had history of a sick contact. Flu positivity rates were much higher during the peak season (up to 50%). One study also found decrease in antibiotic use in flu positive infants.16
Smitherman evaluated the rate of SBI in flu positive infants 0-36 months of age and found that the odds of any SBI (excluding pneumonia) in the flu-negative group were 86% less than in those in the flu-positive group (OR: 0.14; 95% CI (Confidence Interval): 0.04 – 0.46).17
Febrile infants < 60 days of age with RSV infections were at significantly lower risk of SBI compared to the infants with negative RSV (rate of SBIs 7% vs 12.5%) but risk of UTI still remained significant.8 The classic clinical presentation of bronchiolitis usually starts with URI and progresses to the lower respiratory tract over several days. Fever can be present in about a third of infants with bronchiolitis.18
A systematic review was done to study the incidence of apnea in hospitalized infants with RSV bronchiolitis. When seriously ill patients were excluded, the incidence of apnea was between 1.2-4.3%.19 Another study found that the frequency of apnea in infants <12 months of age with bronchiolitis due to RSV and other viruses was 5.6% and one third of them presented with apnea as the first manifestation of bronchiolitis.20 Infants with respiratory symptoms were at greater risk for apnea but the risk of apnea with no respiratory symptoms was not clear.19
To our knowledge, there are no studies that look at the costs of CXR and RSV/flu testing in well-appearing febrile infants <60 days of age. In a study done by Ziegler in 2010, they found the cost savings of not doing a CXR in adult trauma patients was approximately $103 resulting in an overall cost savings of $30,592.21
Amand calculated the healthcare resource use and cost of RSV infants across multiple age groups. They found that higher annual costs in the RSV infants compared to the matched controls across all age groups; ranging from $7,535 to $40,405.22 None of the well-appearing infants < 60 days of age with no sick contacts and respiratory symptoms tested positive for RSV. Although, RSV causes significant cost to healthcare, cost may be decreased by not testing this population for RSV/flu.
One of the inherent limitations of this study was the retrospective nature of the data since we relied on history and physical examination findings documented in the chart. Identification of the cohort was limited to infants for whom a CBC was performed in the PED or pediatric floor. Although it is standard of care to obtain a CBC in febrile infants we may have missed patients if they did not have the standard work-up. One of the other limitations was that our rate of CXR was 62% at the start of this REVISE study and we were actively trying to reduce the number of CXRs during the study period, which might have affected the rate of CXRs. Also, the sample size was not large enough to ensure power and the wide 95% CI around the estimates due to small sample size to conclude there is no necessity of CXR and RSV/flu testing in febrile infants with no respiratory symptoms and no sick contacts.