Bedside Ultrasound in the Diagnosis of Complex Hand Infections

A Case Series

Brett A. Marvel, MD; Gavin R. Budhram, MD


J Emerg Med. 2015;48(1):63-68. 

In This Article


Infections occur within the hand more frequently than in other parts of the body and are most prominent in patient populations exhibiting certain risk factors: peripheral vascular disease, diabetes mellitus, and end-stage renal failure.[3] Staphylococcus and streptococcus are common offending pathogens, but given the more unique routes of entry in the hand, a broader infectious spectrum must be considered. For example, bites involving human saliva have been noted to have a median of four concomitant pathogens, often including Gram-negative rods and anaerobes. Similarly, dog bites are usually polymicrobial, often containing Capnocytophaga canimorsus, a Gram-negative rod that can cause overwhelming sepsis. P. multocida is a highly virulent Gram-negative rod commonly complicating both dog and cat bites, however, due to the puncture-like nature of cat bites, these wounds often develop into deep-space infections.[4]

The anatomy of the hand is complex, containing deep spaces and tendon sheaths that can harbor infection. The deep palmar spaces of the hand include the midpalmar space, the hypothenar space, and the thenar space (Figure 6). Three additional superficial spaces on the dorsum of the hand are the dorsal subcutaneous space, the dorsal subaponeurotic space, and the interdigital web space. Although not technically deep spaces owing to their lack of well-defined anatomic borders, infections in the superficial spaces present similarly to deep space infections.[5] In the normal hand, all of these spaces are potential rather than actual spaces.

Figure 6.

Normal anatomy of the hand showing thenar and midpalmar potential spaces. This figure was published in: Marx J, Hockberger R, Walls R. Rosen's Emergency Medicine Concepts and Clinical Practice, 7th edn., Hand, page 522, Maryland Heights, MO: Mosby Elsevier; 2009.

On the flexor surface of the hand, the flexor tendons are surrounded by synovial sheaths. The tendon sheath of the thumb is continuous with the radial bursa of the palm, and the small finger sheath is continuous with the ulnar palmar bursa. The ulnar bursa surrounds the superficial and deep flexor tendons, and the radial bursa surrounds the flexor pollicis longus. These two bursae communicate in 80% of people. In most instances the tendon sheaths of the index, middle, and ring finger do not communicate. On the extensor surface, tendons are briefly covered by synovial sheaths to reduce friction as they pass over the dorsum of the wrist, but are uncovered as they pass through the dorsal subaponeurotic space.[5]

When assessing patients with hand infections, the EP must be able to distinguish between relatively benign superficial infections and possibly devastating deep space or tendon sheath infections. Traditionally, this is accomplished only by physical examination, with the examiner attempting to elicit induration, fluctuance, and the clinical signs of suppurative flexor tenosynovitis: fusiform swelling, pain with passive extension, pain with palpation along the tendon sheath, and a finger held in slight flexion. The distinction between cellulitis, abscess, and flexor tenosynovitis, however, can still be difficult by physical examination alone, and little support for Kanavel's clinical signs are found in the literature because they were originally published over 100 years ago.[6] A delay in diagnosis, or worse, a misdiagnosis, is known to result in increased morbidity to the patient.[7] For example, amputation rates may approach 17% in misdiagnosed cases of pyogenic flexor tenosynovitis.[3]

Ultrasound is an excellent imaging modality for superficial infections. It can also often spare the patient the radiation exposure associated with traditional imaging methods while lessening the diagnostic financial burden. Ultrasound is known to be more sensitive than computed tomography (CT) for the detection of small cutaneous abscesses. In one recent study by Gaspari et al., ultrasound had a 96.7% sensitivity for abscess vs. 76.7% for CT.[8] The authors attribute this difference to the ability of ultrasound to provide greater sub-millimeter detail combined with the ability for dynamic imaging.[8] Tayal et al. demonstrated marked superiority of bedside ultrasound over physical examination in detecting or ruling out clinically significant abscesses, showing that ultrasound correctly changed management in 71/126 cases, as demonstrated by incisional findings or follow-up.[9] Two studies also document the utility of ultrasound in diagnosing septic flexor tenosynovitis. Jeffrey et al. correctly identified 5 of 6 patients with surgically proven bacterial flexor tenosynovitis, missing only one with a very small amount of fluid in the tendon sheath.[10] Schecter et al. also correctly identified 11 of 12 patients needing surgical drainage based on anechoic fluid in the tendon sheath.[11]

The above cases demonstrate the potential value of bedside ultrasound in distinguishing different types of hand infections with similar physical examination findings. In the first case, ultrasound correctly identified uncomplicated cellulitis even though the patient complained of exquisite pain with passive range of motion, possibly saving her from an unnecessary surgery. In case 2, an abscess was found in the subaponeuritic space even though the physical examination seemed most consistent with cellulitis. Ultrasound was used in the third case to correctly identify suppurative flexor tenosynovitis when the physical examination was interpreted differently by the EP and the surgeon. In the final case, flexor tenosynovitis was suspected based on the clinical examination, but ultrasound also identified the concomitant joint space infection and allowed the surgeon to incise and drain both spaces.

Ultrasound imaging of the hand is best accomplished with a high-frequency linear array transducer. Skin is typically seen as a thin, hyperechoic superficial layer. Subcutaneous tissue is hypoechoic, and skeletal muscle is easily identified by its characteristic hypoechoic texture with echogenic internal striations that appear feather-like in a long axis and speckled in a short axis. In a long axis, tendons are hyperechoic fibrillar structures that move when the corresponding joints are ranged. In a short axis, tendons are hyperechoic and typically round, ovoid, or flat. Bones are hyperechoic and cast a black shadow (Figure 7).

Figure 7.

Normal transverse view of the hand showing subcutaneous tissue (yellow asterisk), tendons (white arrow), muscle (white asterisk), and bone (yellow arrow).

Image quality of the hand is generally enhanced by using a water bath technique. First described in the medical literature over 30 years ago, water bath technique is accomplished by immersing the patient's hand in a large basin filled with warm water.[12] A linear array transducer is then placed into the bath and aimed at the desired imaging surface, but does not touch the skin (Figure 8). The use of water bath in hand imaging has some particular benefits. The hand may be positioned several centimeters from the tip of the transducer, which improves imaging because the focal zone is usually in this area rather than directly adjacent to the transducer. Second, placing the transducer directly onto the skin is usually painful because it necessitates conforming the hand to the contours of the transducer. By using a water bath as a conducting medium, the entire surface area of the hand may be imaged with minimal patient movement.

Figure 8.

Water bath ultrasound. The hand is immersed in a bath of warm water. The transducer is aligned over the finger but does not contact the skin.

Cellulitis displays a characteristic "cobblestone" appearance on ultrasound: thickened and abnormally hyperechoic skin and subcutaneous tissue with areas of hypoechoic edema that traverse it in a reticular pattern. A similar pattern is also seen in edema from noninfectious causes such as severe venous insufficiency. Abscess cavities are most often spherical or elliptical, with the liquefied contents demonstrating a range of echogenicity, and may also contain debris, septae, or gas. They are almost universally surrounded by a rim of hyperechoic soft tissue and display posterior acoustic enhancement. In suppurative flexor tenosynovitis, the ultrasound examination demonstrates a hypoechoic fluid collection within the tendon sheath and thickening of the tendon, as compared to the contralateral side.[10] Gas-producing organisms can create unique challenges to image acquisition. Gas particles are not emendable to sound wave propagation and therefore, inhibit visualization of deeper structures. If the images are easily obtained in non-infected tissues and then have a "dirty" appearance over the area of infection, it would seem prudent to proceed to CT imaging for better disposition planning.