Chest radiography in children with acute lower respiratory infection.
Clinical question How useful are chest radiographs in the management of children with acute lower respiratory infection?
Background Acute lower respiratory infection in children is a common problem, but the diagnostic role of routine chest radiography remains unclear. This randomized trial assessed whether routine chest radiography improves the outcomes of children suspected to have pneumonia.
Population studied This study at a South African teaching hospital enrolled 522 children aged 2 months to 59 months meeting the World Health Organization criteria for pneumonia: cough, tachypnea (defined as a respiratory rate of [is greater than or equal to] 50 breaths/min in children aged 2 to 11 months, and [is greater than or equal to] 40 breaths/min in children aged 12 months or more) but drinking well and without retractions, cyanosis, change of consciousness, or stridor. Exclusion criteria included cough lasting longer than 14 days, focal wheezes, possible tuberculosis, or a clinician's assessment that radiography was mandatory. Only 76% of the patients had telephones, suggesting some differences between this population and that of the United States. Little additional information was provided about the population and the usual medical care. The results are probably generalizable to US family physicians, but more details about the patients (race, population) and clinical management (antibiotic choice, role of blood culture, and use of pulse oximetry) would be valuable.
Study design and validity Eligible children were randomly allocated to either the intervention group (chest radiography) or the control group (standard care only) after identification by an experienced intake nurse. Telephone follow-up was acceptable (78%) and chart follow-up was excellent (100%). Outcome assessment was performed blindly and reliability of chart abstraction was high. Analysis was by intention to treat; confounding was assessed by appropriate statistical methods.
In general, this was a well-designed trial. The possible selection bias of using patients with a greater risk of adverse outcomes is a strength of the study, making radiography more likely to be beneficial. The most significant weakness of the study may be its relatively low power (90% to detect a 50% difference between treatment. groups in duration of symptoms, corresponding to 3 to 4 sick days). Symptoms severe enough to keep a child out of school for an additional 3 days would certainly be an outcome of significance to parents.
Outcomes measured The primary outcome measured was duration of symptoms assessed by twice weekly telephone interviews. Although this approach is still relatively unusual in clinical studies, it is one of the best available. Secondary outcomes measured included final diagnoses, use of antibiotics, hospitalizations, and number of follow-up appointments. Cost, patient satisfaction, and parents' days away from work were not addressed.
Results Of the 581 children enrolled, 59 were excluded; the most common reasons for exclusion were severity of illness (5), not ill enough (7), protocol violations (4), and unilateral wheeze (3). Study groups were similar at baseline. The duration of symptoms of patients receiving radiography was not different from that of the control group (median recovery time of 7 days in both groups). Control for confounding variables such as age, duration of symptoms before presentation, respiratory rate, and clinician's assessment of need for radiography did not change results. There were no deaths in the study. Radiographed children were more likely to be given the diagnosis of pneumonia (14.4% vs 8.4%, P =.03) and given antibiotics (60% vs 52%, P =.05, NNT = 12.5), and there was a trend toward more hospital admissions (4.7% vs 2.3%) and more follow-up appointments (13.5% vs 8.6%).
Recommendations for clinical practice This study provides good evidence that routine chest radiography in children with acute lower respiratory infection does not improve outcomes. In fact, the results show that radiography leads to more antibiotic use and possibly more hospitalizations and office visits, conceivably resulting from an increased identification of children with clinically insignificant pneumonias. Clinicians should understand that these results do not apply to the clinically toxic child.
The methodology of this study--random allocation of a diagnostic test and active surveillance of patients--is a great advance over prior work. Future studies should have increased power and should address the role of radiography in settings where blood culture and pulse oximetry is available and routinely performed.
Angela Hager, MD Warren P. Newton, MD, MPH University of North Carolina at Chapel Hill E-mail: Anewton355@aol.com
|Printer friendly Cite/link Email Feedback|
|Author:||Hager, Angela; Newton, Warren P.|
|Publication:||Journal of Family Practice|
|Date:||May 1, 1998|
|Previous Article:||Appropriate follow-up of patients with colon cancer.|
|Next Article:||The patient's differential diagnosis.|