Considerable technological advances, good safety profile, and ease of use have converged to support the use of ultrasound (US) as an important adjunct in the evaluation of superficial soft tissue infections (SSTI) in general and the differential diagnosis of cellulitis and abscess in particular. However, its use in neonatal populations has not been described. Pediatric studies report clinical examination is not always a reliable method of distinguishing cellulitis from abscess. Two (2) case reports are presented to supplement the growing body of published data that describe US imaging of SSTIs.In both cases, the US exam included the affected area as well as contralateral or adjacent normal skin for comparison. Case 1 describes a preterm infant boy who required placement of a peripheral intravenous (PIV) line and subsequently developed warm, painful, erythematous, and indurated skin in the area of the insertion. Point-of-care US (POC-US) was used to diagnose cellulitis, which initiated treatment with intravenous antibiotics. Case 2 involves a 7-year-old boy with multiple comorbidities who developed a PIV extravasation-related injury that subsequently progressed to cellulitis, likely secondary to wound infection with methicillin-resistant Staphylococcus aureus. Both patients healed completely and without any complications. Because treatment of cellulitis is different from that of abscess, it is important to obtain real-time data supportive of an accurate diagnosis. In these cases, POC-US confirmed the clinical diagnosis of cellulitis and ruled out the presence of an abscess.

Cellulitis and abscess are commonly recognized superficial soft tissue infections (SSTI).¹ Clinically, these 2 entities may appear similar: the skin may be warm, indurated, erythematous, and painful. Typical fluctuance specific for abscess formation is absent in almost half the patients.² The reliability of a physical exam may be poor, especially for children <4 years of age.³ Additionally, in neonatal and pediatric patients, pain at the affected site may limit adequate physical exam. In cases that are suspicious for abscess, blind needle aspiration is suggested.¹ This approach may lead to subjecting the patient to an unnecessary invasive procedure. Therefore, making an accurate and objective diagnosis is important because management of cellulitis requires antibiotics only while an abscess requires incision and drainage together with antibiotics.¹ Coincidentally, these 2 conditions frequently overlap. In fact, abscess may represent a continuation of untreated cellulitis. In previously healthy individuals, white blood cell count and C-reactive protein may rise, but critically ill neonates and children may have previously elevated parameters to complicate management.⁴ Ultrasound (US) can be an asset in providing information regarding the progression of inflammation and direction of treatment.

Normal skin and soft tissue US appearance. The 2 most superficial layers of the skin, the epidermis and dermis, appear hyperechoic on an US exam. Differentiation of the epidermis from the dermis may be difficult unless the lowest possible US depth is used. Below the epidermodermal layer is subcutaneous tissue. It is less echogenic than the more superficial layers and contains septae, subcutaneous fat, blood vessels, nerves, and lymphatic tissue. Beneath the subcutaneous tissue is fascia, which appears as a hyperechoic strip. In the longitudinal US view, muscles have specific hypoechoic echotexture with hyperechoic striations that represent well-organized muscle fibers. In the transverse view, muscle fibers appear as hyperechoic dots as the US plain cuts them crossways (see Figure 1a and 1b).⁵ 

Sonographic appearance of SSTIs.

Cellulitis. US images of cellulitis range from detecting simple increases in echogenicity and thickening of subcutaneous tissues to advanced stages of skin compromise represented by subcutaneous edema and disarray or separation of the tissue planes. Later disease stages are characterized by accumulation of the inflammatory edema that creates hyperechogenic conglomerates of fat lobules visible in the subcutaneous tissue (“cobblestone” appearance).^4-6

Abscess. On US, abscess is usually round in appearance with an irregular border. Classically, it looks like an anechoic fluid collection that represents advanced liquefaction and pus formation. Septations (ie, division of the affected area into different parts) may be present and are easy to identify. Tissue debris inside an abscess may occasionally appear hyperechogenic. On the “compression test” (light compression of the skin with an US probe), purulent material will swirl, giving rise to the “squish or swirl sign.” Rarely, in the initial stage of development an abscess may appear hyperechogenic. Color flow Doppler is recommended to confirm the abscess is not a blood vessel. At the same time, the area around the abscess will show hyperemic changes.⁶ It is essential to understand that the progression from cellulitis to frank abscess formation is a dynamic process necessitating serial US exams. Furthermore, the fact that US is becoming more affordable, quick, and nonionizing makes it an ideal imaging modality in the diagnosis and follow-up of SSTIs as compared to computerized tomography (CT) scan or magnetic resonance imaging (MRI).^4,7

Examination of pediatric skin requires a 10-MHz or higher linear probe. An acoustic water bath or a copious “stand-off” gel pad is used to minimize the compression of the skin layers. The authors warm the gel before the exam, especially in neonates residing in a temperature-controlled environment. If infection is suspected, an US probe cover is necessary.

The purpose of this case report is to briefly overview the use of US in pediatric patients as well as to demonstrate comparable utility in the sensitive neonatal population.

For these cases, the authors used a Zonare Z.One PRO device (ZONARE Medical Systems, Inc, Mountain View, CA) with 14 and 20 MHz linear probes set to the “superficial” or “skin” preset. The US exam included the affected area as well as contralateral or adjacent normal skin for comparison. The probe was held perpendicular to the skin. Still images and video clips were recorded in 2 orthogonal planes. Images were securely saved to the machine’s database. Consent to share the results was obtained from the patients’ parents.

Case 1. The patient was a baby boy of 29-weeks’ gestation born via emergent cesarean delivery due to maternal preeclampsia and pulmonary edema. The mother was 39 years old with a history of gestational hypertension treated with labetalol. All prenatal labs were negative with unknown Group B streptococcus status. The mother received betamethasone, magnesium, and antibiotics before delivery. Apgar scores were 8 and 9. Shortly after delivery, the baby developed mild respiratory distress that was managed with continuous positive airway pressure (CPAP). His chest x-ray was consistent with mild respiratory distress syndrome. Ampicillin and gentamicin were started after the blood culture was drawn. An umbilical venous catheter was placed on day 1 and total parenteral nutrition was initiated. Antibiotics were stopped after 2 days due to normal labs, stable clinical status, and negative blood culture. The baby remained stable on nasal CPAP with nasogastric feeds advancing as expected. The central line was removed on day 7 when the baby achieved an enteral feeding of 100 cc/kg/day. A peripheral intravenous line (PIV) was placed on day 8 for 4 more days for administration of additional IV fluid.

On day 14, baby’s physical exam showed warm, painful, red, and indurated right forearm skin around the area where the PIV was placed (see Figure 2). Blood samples were obtained for complete blood count and culture, and the infant was started on vancomycin and amikacin. The erythematous, swollen area was marked and POC-US was performed on the affected forearm. US showed edematous skin, tissue plane disarray, and hyperechogenicity. No abscess was appreciated (see Figure 3). Further, the white blood cell count was abnormal with significant left shift. Early cellulitis was diagnosed. The baby was continued on antibiotics for 7 days until complete resolution of infection. At the end of his antibiotic therapy, he had a follow-up POC-US exam that showed complete resolution of all signs of inflammation.

Case 2. A 7-year-old boy with global developmental delay, seizure disorder, central apnea, and severe asthma presented to the emergency room after having a febrile seizure and worsening severe respiratory distress that required increases in bilevel positive airway pressure (BiPAP) settings, respiratory medications, magnesium sulfate, and systemic steroids. His extensive medical history included recurrent respiratory infections, multiple intubations, subglottic stenosis, multiple medications, and BiPAP at night. His respiratory viral panel was positive for rhinovirus/enterovirus and coronavirus. He was admitted to the pediatric intensive care unit (PICU), where he was intubated and placed on high-frequency oscillatory ventilation. Due to the possibility of superimposed pneumonia, he was placed on broad antibiotic coverage for 10 days. During his initial PICU stay, he developed hypotension that was treated with multiple fluid boluses and vasopressors. A peripheral arterial line was placed in his right radial artery to facilitate blood pressure monitoring. A PIV was placed in his right wrist vein. On day 7, the PIV was removed due to a suspected grade 4 extravasation, followed by removal of the arterial line. Overlying skin and soft tissue presented with marked blistering, swelling, and erythema. Plastic and vascular surgery evaluated the lesion and recommended elevation and local antimicrobial ointment. Mupirocin was started. A US Doppler scan of the affected area did not show signs of deep venous thrombosis. Over the next week, a few of the blisters ruptured, leaving a partial-thickness wound with irregular edges, periwound erythema, and skin slough covering 40% of the wound bed. Wound care physicians changed management to topical medical-grade Manuka honey (Leptospermum honey), covered by a silicone dressing. Wound healing proceeded as expected.

One (1) week into the treatment, the wound area became increasingly painful, erythematous, warm, and indurated. A clindamycin IV was started and changed to vancomycin IV as per infection disease service recommendation. To ensure no underlying abscess was present, POC-US was performed. The US exam included 3 areas around the wound: the transitional area between normal skin and erythema edge, an area directly over the erythema, and an area at the edge of the wound (see Figure 4). The transitional area showed minimal tissue swelling (see Figure 5). The area over the erythema showed the typical US cobblestone appearance of cellulitis with marked dermal swelling (see Figure 6). Finally, the area at the edge of the wound showed cellulitic changes as well as some deeper tissue planes destruction, with coagulated blood noticeably present (see Figure 7). During the following week, the patient had daily wound exams and POC-US evaluations of the wound area. Medical grade honey was used to help with the wound healing. The patient was discharged home after completing 7 days of IV vancomycin. His wound healed completely 2 weeks after the discharge.

Several prospective studies^8-10 performed in more than 300 adults have shown POC-US (compared to physical exam alone) improves accuracy in diagnosing SSTIs. Sensitivity and specificity ranged from 65% to 100% and 89% to 100%, respectively, which is consistently more accurate than the physical exam. Further, in a prospective study⁸ in adults, POC-US not only showed better sensitivity and specificity, but it also changed the initial management decision based solely on the physical exam alone. This change occurred in 17% to 56% of the patients, subsequently preventing initiation of unnecessary procedures (ie, incision and drainage in cases of cellulitis only) and/or additional imaging.

In the pediatric (but not neonatal) population, sonographic detection of SSTIs has been studied in a large number of patients.^11,12 Sensitivity and specificity ranged from 91% to 96% and 30% to 83%, respectively.^2,11-13 Change of the initial management strategy based on the physical exam alone occurred in 14% to 27% of cases after US was performed; this fact is very important because this population is very sensitive to procedural pain.^2,11 A recently published meta-analysis/systematic review¹⁴ of adult and pediatric studies that examined POC-US ability to diagnose skin abscess as a primary outcome found sensitivity 96.2%, specificity 82.9%, positive likelihood ratio 5.63, and negative likelihood ratio 0.05.

US is quick and poses no radiation risk to patients. It has been suggested as an initial imaging modality in SSTI evaluation.¹⁵ A retrospective study¹⁶ of 612 adult patients evaluated with both US and CT found US is more sensitive but CT is more specific for diagnosing SSTIs. As shown in a prospective, observational study⁹ performed among 40 adult patients, a novice sonographer with only short, focused training can reach high sensitivity and specificity in differentiating cellulitis from abscess. The same conclusion was reached in a study of 107 pediatric patients that examined interrater reliability between expert sonologist and novice ER physicians who had only 1 day of training in US use.¹⁷

Wider availability of POC-US may facilitate diagnosis and guide intervention and follow-up of SSTIs in both adult and pediatric patients. Its use in neonatal critical care may be potentially extended to examination of PIV extravasations, wound management, central line thrombophlebitis evaluation, and other clinical applications. At present, to the best of the authors’ knowledge, POC-US has not been used in any of these cases. Conversely, implementation of POC-US is limited by the need for trained personnel, wider availability of structured training programs, and credentialing processes.

Dr. Kurepa is an attending neonatologist and Director of point-of-care ultrasound; Ms. Galiczewski and Ms. Civale are neonatal nurse practitioners; and Dr. Vitaliya Boyar is an attending neonatologist and Director of neonatal wound services, Cohen Children’s Medical Center, New Hyde Park, NY.

Please address correspondence to: Dalibor Kurepa, MD, Steven and Alexandra Cohen Children’s Medical Center, 269-01 76th Avenue, New Hyde Park, NY 11040; email: dkurepa@northwell.edu.