Point-of-care ultrasound can improve efficacy and safety of pediatric procedures performed in the emergency department. This article reviews ultrasound guidance for the following pediatric emergency medicine procedures: soft tissue (abscess incision and drainage, foreign body identification and removal, and peritonsillar abscess drainage), musculoskeletal and neurologic (hip arthrocentesis, peripheral nerve blocks, and lumbar puncture), vascular access (peripheral intravenous access and central line placement), and critical care (endotracheal tube placement, pericardiocentesis, thoracentesis, and paracentesis). By incorporating ultrasound, emergency physicians caring for pediatric patients have the potential to enhance their procedural scope, confidence, safety, and success.
Point-of-care ultrasound (POCUS) is an efficient and effective adjunct to pediatric emergency procedures.
POCUS has the potential to improve the care of pediatric patients by expanding the procedural scope of emergency physicians and increasing procedural success.
Ultrasound guidance can make infrequently performed, life-saving pediatric procedures safer and with fewer complications.
Pediatric patients pose unique opportunities and challenges with regard to ultrasound-guided procedures in comparison with adults.
Point-of-care ultrasound (POCUS) has an established role in emergency medicine and critical care to improve the success and safety of procedures. Recent advances in the literature address gaps in knowledge related to procedural applications of ultrasound in pediatric emergency medicine. Pediatric emergency physicians and trainees now are expected to achieve competency in basic ultrasound-guided procedures. This review summarizes the evidence and approach for the utilization of ultrasound to guide both common and uncommon procedures in ill and injured pediatric emergency department (PED) patients.
Soft tissue procedures
Abscess Incision and Drainage
Although diagnosis and management of skin and soft tissue infections (SSTIs) traditionally have been guided clinically, clinical examination is imprecise and unreliable in differentiating cellulitis from drainable abscess among pediatric patients. , Two systematic reviews that included both adult and pediatric studies have shown that POCUS has high sensitivity (94.6%–97%) and moderate specificity (83%–85.4%) for diagnosis of abscess , and commonly leads to an appropriate change in management, helping to avoid unnecessary procedures. Accuracy of POCUS, however, may be lower in children than adults (sensitivity 89.9% and specificity 79.9%). In 2 recent prospective observational studies of pediatric patients undergoing SSTI evaluation in the emergency department (ED), use of POCUS was associated with decreased incision and drainage (I&D) and a change in management in 23% of cases, without resulting in any difference in clinical outcomes compared with those managed without POCUS. , One retrospective PED cohort showed a significantly decreased length of stay (73 minutes overall and 89 minutes in those discharged from ED) among patients who underwent POCUS compared with those who received radiology-performed ultrasound. In a recent randomized controlled trial (RCT) among adult ED patients requiring abscess I&D, those who underwent POCUS before I&D (and during/after at clinician discretion) were significantly less likely to experience treatment failure and need repeat I&D than those who underwent physical examination alone (3.7% vs 17%, respectively).
A high-frequency linear transducer is used to evaluate soft tissue for abscess, which appears as an anechoic or hypoechoic fluid collection that may have surrounding cobblestoning (fluid tracking between lobules of subcutaneous fat) due to cellulitis ( Fig. 1 ) and/or hyperemia on color Doppler (rather than internal vascularity as in a lymph node or other vascular structure). Compression can reveal ”pustalsis,” or “squish sign,” secondary to movement of purulent material ( Fig. 2 ). , The area of interest should be examined in 2 planes, making sure to evaluate the full extent and depth of the infection, to identify the largest pocket of fluid and any surrounding structures, and to estimate the abscess volume. In 1 retrospective study, fluid collection deeper than 0.4 cm from the skin was predictive of treatment failure (area under the curve 0.83; sensitivity 85%; and specificity 68%). Both static preprocedural site marking and dynamic real-time ultrasound-guided techniques can be utilized. With the dynamic approach, the scalpel or needle can be visualized directly and the success of the procedure can be evaluated. Ultrasound-guided needle aspiration, however, is less likely to result in clinical resolution compared with I&D.
Foreign Body Identification and Removal
Failure to diagnose retained foreign bodies (FBs) is a leading cause of malpractice claims against emergency physicians, and ultrasound has an important role in the evaluation and management of foreign material in wounds and soft tissue. Although plain radiography has high sensitivity for radiopaque FBs, such as metal and glass, it is less useful for radiolucent materials like wood and plastic. In cadaveric and experimental soft tissue FB models, ultrasound has demonstrated better detection of wooden FBs than CT, , with particularly impressive sensitivity for small wooden FBs (as small as 2.5–4 mm). , Emergency physicians and trainees similarly were able to detect 29 of 30 FBs in an experimental model, with sensitivity and specificity of 96.7% and 70%, respectively. Studies in the clinical setting have been more limited. In a prospective sample of 105 patients in a PED, in which 12 of 131 wounds ultimately were found to have FBs, POCUS had diagnostic accuracy similar to radiography (66.7% sensitivity and 96.6% specificity) and, when used together parallel with radiograph, had a combined sensitivity of 83.3%. In another study of 123 patients with suspected soft tissue FBs, sonographer-performed ultrasound demonstrated sensitivity and specificity of 94.5% and 53.8%, respectively, for FB at surgery. Although there have been no prospective studies, case reports and case series have described ultrasound-guided removal of soft tissue FBs in the ED setting.
A high-frequency linear transducer provides the high image resolution required to visualize subcutaneous FBs ( Fig. 3 ). A transparent medical dressing should be placed over the probe if scanning over an open wound. A water bath can be utilized to acoustically enhance the resolution of more superficial FBs, by placing the patient’s hand or foot in a water basin and introducing the tip of the probe into the water approximately 1 cm above the skin. The FB typically appears hyperechoic and may demonstrate acoustic shadowing or ring-down artifact, depending on its composition. A surrounding, hypoechoic halo may be seen due to edema, abscess formation, or granulation. Ultrasound-guided FB removal often involves preprocedural site marking, which can be drawn in both axes of the FB. 29 Real-time ultrasound can provide in-plane guidance of a needle or scalpel as it is inserted toward and contacts the FB. The puncture or incision can be extended to allow direct visualization of or access to the FB. Lastly, forceps then can be inserted along the tract or incision, with or without ultrasound guidance, in order to gently grasp and remove the FB. ,
Peritonsillar Abscess Drainage
Peritonsillar abscess (PTA) is a deep space neck infection seen most commonly in older school-aged children, adolescents, and young adults, which can be challenging to distinguish accurately from peritonsillar cellulitis; clinical impression alone is unreliable, with sensitivity and specificity of 78% and 50%, respectively, in 1 study. As such, ultrasound can play an important role in lieu of computed tomography (CT) in the diagnosis of PTA prior to and procedural guidance during needle aspiration. One prospective RCT among 28 adult ED patients that compared intraoral POCUS confirmation and localization of PTA prior to needle aspiration versus clinical diagnosis and landmark-based aspiration found that POCUS provided higher diagnostic accuracy (100% vs 64%, respectively), higher rate of successful aspiration (100% vs 50%, respectively), lower subspecialty consultation (7% vs 50%, respectively), and lower CT rate (0% vs 50%, respectively). Transcutaneous cervical ultrasound can be utilized in children or patients with trismus who may not tolerate the intraoral approach, with sensitivity of 80% to 91% and a specificity of 80% to 93%. One recent retrospective cohort study of 179 children presenting to an ED with suspected PTA evaluated clinical outcomes and financial impact of a protocol incorporating radiology-performed transcervical ultrasound compared with traditional diagnosis by examination and/or CT. Only one-third of patients in the ultrasound group were found to have abscess on ultrasound, and this group was significantly less likely to be exposed to CT radiation (odds ratio [OR] 0.34), to undergo a procedure (OR 0.38), and to have a length of stay longer than 23 hours (OR 0.35) without experiencing any difference in treatment failure requiring readmission within 30 days (8% vs 14%, respectively). A low-cost and realistic PTA training model successfully increased emergency medicine resident comfort with this procedure.
The transcutaneous approach utilizes a high-frequency linear transducer placed below and parallel to the mandible and directed cephalad toward the peritonsillar region on both the unaffected and unaffected sides. , Movement of the tongue allows for identification of the air-mucosal interface below the tongue and the pharyngeal tonsil posteriorly ( Fig. 4 A). A PTA appears as a hypoechoic or heterogeneous structure lateral and deep to the tonsil on the affected side (see Fig. 4 B) and may demonstrate surrounding hyperemia on color Doppler. The intraoral approach typically utilizes an endocavitary transducer with overlying sterile glove or probe cover after topical application of lidocaine spray in order to confirm, localize, and measure the depth of the abscess as well as the carotid artery. , Real-time needle visualization can be performed by attaching a biopsy adaptor to the transducer, and a case series recently described use of small, pencil-shaped burr hole and hockey stick transducers. Incision, deloculation, and drainage may be attempted if needle aspiration is unsuccessful.
Musculoskeletal and neurologic procedures
Septic arthritis of the hip is a not-to-miss and joint-threatening diagnosis in infants and children with limp, hip pain, fever, and/or refusal to bear weight. POCUS can allow for rapid diagnosis and, if clinically indicated, needle aspiration of a hip joint effusion. Joint fluid cell count above 25,000 to 50,000 [WBCs]/mm 3 is concerning for septic arthritis and is an important variable to help differentiate from other diagnoses, such as transient synovitis, Lyme arthritis, juvenile idiopathic arthritis, and reactive arthritis due to an adjacent osteomyelitis. Although standard of care for definitive management of hip septic arthritis is open anterior arthrotomy, sequential aspiration has been described. One PED has demonstrated that pediatric emergency medicine physicians and fellows , can diagnose hip effusions accurately with POCUS through a small prospective sample and larger retrospective cohort of pediatric patients (sensitivity 80%–85% and specificity 98%). Test accuracy improved with increasing provider experience (2.4% increased odds of accurate scan per additional study performed) and when provider confidence in the POCUS result was high (sensitivity 90% and specificity 100%). POCUS-guided hip aspiration performed in the ED by emergency physicians has been described in both pediatric and adult case series. , In 1 case series, 17 children with septic arthritis underwent POCUS diagnosis of hip effusion by emergency physicians and landmark-based needle aspiration in the ED by orthopedics prior to admission for intravenous antibiotics; 59% of these patients required no additional hip aspirations and all were treated successfully without arthrotomy. A recent retrospective study comparing ED-performed with radiology-performed hip ultrasound or arthrocentesis in adult patients found that those who received ED-performed studies had significantly faster time to ultrasound (68 minutes vs 208 minutes, respectively) and to arthrocentesis (211 minutes vs 602 minutes, respectively). A 3-dimensional printed model for ultrasound-guided hip arthrocentesis recently has been described.
To evaluate for hip effusion, a high-frequency linear or curvilinear transducer is placed on the anterior hip along the axis of the femoral neck with the probe marker rotated toward the umbilicus. The concave anterior synovial recess of the femoral neck is identified distal to the femoral head and physis and the capsular synovial thickness is measured from the cortical surface of the bone and the posterior surface of the iliopsoas. , A measurement greater than 5 mm or more than 2-mm difference with the asymptomatic hip meets criteria for hip joint effusion in children. An anechoic or hypoechoic hip effusion typically has a biconvex lens appearance, with the anterior synovial layer pushed outward into a convex shape. As with other pediatric joints, care should be taken to not mistake the hypoechoic cartilage of the femoral head within the hip joint for an effusion within the joint capsule anterior to the femoral neck. Although hip aspirations may be performed successfully after preprocedural site marking, smaller anatomy and closer proximity of the neurovascular bundle makes real-time guidance preferable in pediatric patients. First, with the transducer in transverse orientation parallel to the inguinal crease, the locations of the femoral neck as well as the femoral vessels on color Doppler are identified and marked. The probe then is rotated 90° along the long axis of the femoral neck, the anterior synovial recess is centered, and a final mark is made intersecting perpendicularly with the first, localizing the target for arthrocentesis. Ensuring appropriate anxiolysis or sedation, local anesthesia, and sterile technique, an 18-gauge to 20-gauge spinal needle is visualized in-plane along the axis of femoral neck as the tip enters the joint and joint fluid is aspirated ( Fig. 5 ). Because the relative position of the vessels may change with any rotation of the hip, their position should be reconfirmed if the patient is moved after sedation.
Peripheral Nerve Blocks
Regional anesthesia, including peripheral nerve blocks (PNBs), long has played an important role in perioperative pain management and classically has been performed via anatomic landmarks. One RCT in the PED setting demonstrated that axillary nerve block after midazolam anxiolysis is a feasible alternative to ketamine sedation for children ages 8 and older undergoing forearm fracture reduction. Another PED-based RCT showed that fascia iliaca compartment block performed by pediatric emergency medicine attendings and fellows for children ages 15 months to 18 years with femur fracture can result in lower pain scores, longer duration of analgesia, and higher staff satisfaction in comparison with traditional analgesia. Although these RCTs used landmark approaches, ultrasound guidance increases pediatric PNB success and quality and has helped broaden its utilization in emergency medicine. , PNBs have an outstanding safety profile in children under general anesthesia, and, by allowing for real-time visualization of the target nerve, surrounding blood vessels, needle tip, and local anesthetic administration, ultrasound can decrease the risk of nerve injury and local anesthetic systemic toxicity (LAST). In a preimplementation and postimplementation retrospective cohort study of children with femur fractures in a PED, ultrasound-guided femoral nerve block was associated with a 3-times longer duration of initial analgesia (6 h vs 2 h), lower total morphine dose, and fewer nursing interventions in comparison with systemic analgesia alone. One child who mistakenly was given twice the dose of local anesthetic experienced a brief seizure more than 6 hours following the procedure, whereas the remainder of the adverse events across both groups were related to intravenous narcotic medication, including an episode of hypoxia. A prospective descriptive study supported the efficacy, safety, and efficiency of forearm nerve blocks in the ED for pediatric hand injuries. Case reports and case series further exemplify the broadening potential of PNBs in the PED, including a femoral nerve block in an infant, fifth phalanx fracture reduction, popliteal sciatic nerve block for ankle laceration, and posterior tibial nerve block for foot FB removal. No pediatric ultrasound-guided brachial plexus blocks have been described in the ED setting. Educators have made use of online and simulation-based curricula and homemade meat-based nerve block models to teach ultrasound-guided PNBs.
Peripheral nerves have a round or triangular honeycomb-like appearance in cross-section on ultrasound, with multiple hypoechoic nerve fascicles separated by echogenic interfascicular perineurium, and are best visualized with a high-frequency linear transducer. They commonly traverse adjacent to vasculature and within fascial planes between muscle groups ( Fig. 6 ). Care should be taken to confirm the appropriate weight-based and maximum dose of local anesthetic and to ensure an adequate degree of patient anxiolysis or sedation. With the target nerve typically in the transverse axis, both in-plane and out-of-plane needle insertion can be utilized, although the in-plane approach allows for the best visualization of the needle tip and local anesthetic spread. General strategies for avoiding nerve injury, vascular injury, and LAST include targeting the nerve-encompassing fascial plane, hydrodissection with normal saline prior to local anesthetic injection, intermittent aspiration for blood, sequential injection of small volume aliquots, and cardiac monitoring.
Pediatric lumbar puncture (LP) is a common diagnostic and therapeutic procedure that emergency medicine physicians and pediatricians are expected to master. Unsuccessful and traumatic LP rates are high, however, particularly among infants and obese patients. As demonstrated in a recent meta-analysis of 6 adult and 6 pediatric studies, preprocedural site marking with ultrasound has the potential to increase success rates, decrease traumatic LPs, shorten time to successful LP, decrease the number of needle passes, and decrease pain associated with the procedure. Among the pediatric subgroup, ultrasound-assisted LP was associated with a fewer traumatic LPs (OR 0.25), fewer needle passes (adjusted mean difference −0.47), and a trend toward higher success rates (OR 2.55; 95% CI, 0.99–6.52) in comparison with landmark-based LP. Three RCTs have compared preprocedural site marking by ultrasound with the traditional landmark approach among newborns and young infants undergoing LP in the PED. Success rates were 75% v 44% (cerebrospinal fluid [CSF] obtained within 3 attempts, <1000 red blood cells [RBCs]/mm 3 ); 95% v 68% (adequate CSF with three 0.5-mL samples, <10,000 RBCs/mm 3 P = .023) and 75% v 65% (CSF obtained on any attempt, <1000 RBCs/mm 3 P = .25) in the intervention and control groups of each of these studies, respectively. The insignificant differences found in the latter study may be a result of both a higher baseline success rate and a much wider variation in the experience of the study sonographers (16 pediatric emergency medicine attendings/fellows rather than 3 dedicated study investigators). Similar RCTs among older children have shown mixed results.
Depending on patient body habitus and whether or not spinal landmarks are palpable, there are 2 primary approaches to ultrasound-assisted LP: midline and paramedian. , In the midline approach, a linear (infants/children) or curvilinear (adolescents/adults) transducer is first oriented transversely to identify and mark midline through the recognition of hyperechoic and shadowing spinous processes. The transducer then can be oriented sagittally, beginning at the level of the sacrum and sliding cephalad, to identify the desired interspinous spaces and to estimate the required needle insertion depth ( Fig. 7 ). In newborns and infants, ultrasound also can be utilized to identify and mark the conus medullaris, determine the appropriate angle of needle entry, and diagnosis epidural hematoma after unsuccessful LP. In patients with nonpalpable bony landmarks, the spinous processes can be difficult to identify by ultrasound, and the paramedian approach can be useful. The curvilinear probe first is oriented with probe marker cephalad, sliding just lateral to midline in order to visualize the articular processes as hyperechoic and shadowing humps ( Fig. 8 ). Anatomically, the articular processes align with the interspinous space, whereas the respective spinous (midline) and transverse (most lateral) processes are cephalad relative to the space. After sliding caudal to identify the sacrum, the L5/S1, L4/L5, and L3/L4 articular processes are identified, and the latter 2 are marked. Lastly, midline is marked using a combination of anatomic (eg, gluteal cleft) and ultrasound landmarks. For both approaches to ultrasound-assisted landmark identification, the LP subsequently is performed at the intersection of the midline and interspinous markings, although paramedian needle entry through the interlaminar space and real-time guidance also are possible.
Peripheral Intravenous Access
Peripheral intravenous (PIV) catheter placement is the most common approach to vascular access in the ED, and infants, toddlers, and children with a history of prematurity have an increased risk of difficult intravenous access. Although routine use of ultrasound to guide PIV placement is not indicated for the pediatric patient, it can be helpful for those whose veins cannot be visualized or palpated as well as those who have had failed prior attempts. , Two RCTs have compared ultrasound guidance with conventional placement in the PED among patients who had 1 or more failed prior attempts. , In the first, preprocedural vein identification and marking by a physician ultrasonographer showed no benefit for successful venous cannulation by the nurse. Conversely, in the second, a dual-operator approach to real-time ultrasound guidance was associated with decreased time to success, fewer attempts, and fewer needle redirections, although the study may not have been powered to detect a significant improvement in overall success (80% vs 64%, respectively; P = .208). A recent prospective, nonrandomized, precomparative/postcomparative study in a PED actually found a significantly increased adjusted odds of failure using dual-operator ultrasound guidance compared with conventional technique; however, the ultrasound group was significantly more likely to have comorbidities and higher difficult intravenous access scores. The investigators highlighted the challenge of ultrasound guidance in uncooperative pediatric patients and the limitation of using short 24-gauge catheters. In contrast, among infants and toddlers less than 3 years old who had received inhaled general anesthesia, those randomized to ultrasound-guided PIV placement by a single operator had higher first-time success rate, decreased time, and decreased punctures compared with blind approach to access veins that were neither visible nor palpable. Two recent training program evaluations demonstrated acquisition of competency with single-operator ultrasound-guided PIV placement among pediatric nurses. ,
The short axis out-of-plane is the approach utilized most commonly for real-time ultrasound-guided PIV placement in children and infants. A high-frequency linear transducer is used to select the target vessel and distinguish from arteries with compression and color Doppler, noting trajectory, depth, diameter, bifurcations, and valves. The cephalic vein in the forearm, larger diameter, and higher uninterrupted length on longitudinal view are independent predictors of successful PIV placement. , After application of antiseptic and sterile gel, the target vein is centered in cross-section, the needle is placed directly adjacent and perpendicular to the long axis of the probe and introduced into the skin at a 30º to 45º angle, and the hyperechoic needle tip is visualized on the screen. Upon tenting of the vein and blood return, the angle of entry is decreased and the needle advanced slightly, before threading the catheter. Dynamic needle tip visualization, involving an iterative process of sliding the probe proximally until the needle tip disappears and then advancing needle until the tip reappears, can ensure that the needle remains in the center of the vessel and increase success. Utilization of a guide wire as well as longer catheters to increase dwell time also can be considered.
Central Venous Catheter Placement
The Institute of Medicine, evidence-based consensus guidelines, , and a recent pediatric systematic review recommend the use of ultrasound to improve the safety and success of central venous catheter (CVC) placement, a lifesaving, albeit infrequent, , procedure used in the resuscitation of critically ill or injured children in the ED. Two meta-analyses, , which included 8 RCTs and 12 RCTs, respectively, support the routine use of ultrasound-guided CVC placement in pediatric patients in order to decrease rate of cannulation failure, , number of attempts, , and risk of arterial puncture. Only 1 retrospective study has compared femoral and internal jugular CVC placement with and without ultrasound assistance in the PED setting, finding higher odds of success in the ultrasound group (OR 13.1; 95% CI, 2.9–59.4), which remained significant even after adjusting for physician and patient factors. This same pediatric tertiary care ED also demonstrated significant increase in ultrasound use for CVC placement after implementing an emergency ultrasound program as well as improved competency after implementing an ultrasound-guided CVC placement educational intervention. Prospective observational studies in the pediatric critical care setting, including a preultrasound/postultrasound guidance protocol evaluation and a propensity score matched cohort study, found fewer complications, , faster procedure time, increased first attempt success, and decreased puncture attempts with ultrasound guidance compared with anatomic landmark approaches. Randomized and quasi-randomized studies in the pediatric cardiology, anesthesia, and surgery literature also support the use of ultrasound for CVC placement.
The femoral vein and right internal jugular vein are the most common sites for pediatric ultrasound-guided CVC placement. Using a high-frequency linear transducer, the vessels should be scanned in long-axis and short-axis to visualize anatomy and identify adjacent structures, including any overlap between the femoral vein and artery. Although static, preprocedural site marking is beneficial in comparison to landmark techniques for femoral vein cannulation, dynamic real-time ultrasound guidance generally is recommended. Using a sterile probe cover and appropriate sterile technique, a single-operator, out-of-plane short-axis approach with dynamic needle tip visualization should be utilized to guide insertion of the needle. During the procedure, the probe can be rotated to the long axis to visualize the catheter tip and guide wire and to evaluate for posterior wall penetration ( Fig. 9 ). There also may be utility to confirming catheter tip placement and evaluating for pneumothorax after completion of the procedure.