Reference: Chest 2016 Jul;150(1):131
Level 1 [likely reliable] evidence
Pneumonia is the leading cause of death in children worldwide, with over 150 million cases diagnosed in children under the age of 5 every year (Bull World Health Organ. 2008;86(5):408). The gold standard for diagnosis is chest x-ray demonstrating lung infiltrates (Prim Care Respir J. 2010;19(3):237–241), as physical exam alone has been shown to be unreliable (Ann Emerg Med. 2005;46:465-467). While the radiation dose associated with chest x-ray is generally considered safe in children (Pediatr Radiol. 2009;39 (Suppl 1):S4), chest ultrasound offers an alternative method for diagnosis with no radiation exposure at all. Compared to x-ray, chest ultrasound is less expensive, more portable, and improves inter-observer standardization for methodology and interpretation. Regarding imaging in children, the Federal Drug Administration states that “if there is a medical need for a particular imaging procedure and other exams using no or less radiation are unsuitable, then the benefits exceed the risks, and radiation risk considerations should not influence the physician’s decision to perform the study or the patient’s decision to have the procedure” (US FDA Center for Devices and Radiological Health, “Pediatric X-ray Imaging”, 2017). Could ultrasound reduce the use of chest x-ray in children without increasing the number of missed cases of pneumonia?
A recent randomized controlled trial compared the feasibility and safety of lung ultrasound and chest x-ray for the diagnosis of pneumonia in children. One hundred ninety-one patients aged < 21 years (median age 3 years) with suspected pneumonia were randomized to either chest x-ray (CXR) (control) or lung ultrasound (LUS) followed by CXR if there was clinical uncertainty or by request. Patients were screened by convenience sampling and all staff, patients, and guardians were aware of group assignments. The sonologists were attending physicians and fellows in the pediatric emergency department (ED) with varying levels of experience who each underwent a pre-study training session on LUS. The LUS findings were documented prior to CXR to maintain blinding to the CXR results, and radiologists were blinded to the LUS results. All patients in the control arm underwent CXR followed by LUS. The primary outcome was the rate of CXR reduction. Secondary outcomes included missed cases of pneumonia, defined as cases diagnosed by a healthcare provider at a repeat ED visit or other health care visit.
Of the 103 patients assigned to receive LUS, 63 subsequently underwent CXR. Comparing LUS vs. CXR, LUS reduced use of CXR by 39% (61% vs 100% respectively, 95% CI 30.0%-48.9%, NNT 3). This implies that for every three children who receive LUS, one will not require a subsequent x-ray. Ultimately, pneumonia was diagnosed by CXR in 14 of the patients in the LUS arm and 12 patients in the control arm. When stratified between novice and experienced sonologist, the reduction in CXR was 30.0% (95% CI 23.5%-36.5%) and 60.6% (95% CI, 47.0%-74.1%), respectively (no p value reported). When stratifying by age, CXR use was reduced in 47.9% of patients aged less than two years, and in 30.9% of patients aged over two years (no p value reported). Although several secondary outcomes were evaluated and showed no significant differences between groups (including missed cases of pneumonia and reduction of antibiotic use), the study was likely underpowered to detect statistically significant differences in these outcomes.
In this trial, the use of ultrasound in children suspected to have pneumonia reduced the use of x-ray, particularly in children under the age of two years. It appears that the rate of missed cases did not increase, even when LUS was performed by a less experienced technician. Compared to x-ray, lung ultrasound could provide a safer, lower-cost, and more readily available method of reliably diagnosing pneumonia worldwide.
For more information, see the Community-acquired pneumonia in children topic in Dynamed Plus. DynaMed users, see the Community-acquired pneumonia in children topic in Dynamed Classic.
Sarah Dalrymple, MD is a third year Family Medicine resident at the University of Virginia. Her specific interests include obstetrics, women’s health, and teaching. She lives in Charlottesville with her husband, a neurology resident, and their wonderful dog. She’ll be graduating in a few short months and is looking forward to further training in academic family medicine.
Faculty contributions by Katharine C. DeGeorge, MD, MS.