“Talk Dirty to Me”: Bedside Testing for Necrotizing Fasciitis

Necrotizing fasciitis (NF) is a rare, but serious infection involving the subcutaneous tissue and fascia of the body. The culprit organisms are gram-positive, gram-negative, aerobic, and/or anaerobic bacteria, and sometimes even fungi, NF may become fulminant, leading to extensive destruction of tissue, systemic toxicity, limb loss, and even death (mortality rate ranging from 20% to 80%).[2] Although laboratory testing, scores, and imaging may aid in the diagnosis, NF is a clinical diagnosis. Furthermore, since surgery is the definitive treatment for NF, blood testing and imaging should not delay surgical consultation. Below is a well-summarized table of all the clinical signs of NF.

Table 1: Clinical Features Suggestive of Necrotizing Soft Tissue Infections[3]

Laboratory testing is not specific, and different scores that include vital signs and laboratory values have tried to distinguish NF from other soft tissue infections. 

The most common score is the LRINEC score. In the original retrospective study by Wong et al., the LRINEC score utilized WBC, CRP, Hgb, Na, Cr, and glucose to risk stratify 89 patients who initially did not meet criteria for severe soft tissue infection (clinical impression of severe infection based on documentation, the use of parenteral antibiotics for ≥48 hours, and abscesses [when present] needing surgical debridement), but were later diagnosed with NF. A score of 8 and above is considered high risk, 6 and above is considered intermediate risk, and 5 and below is considered low risk for NF. In the study, a LRINEC score of 6 and above had a sensitivity of 89% (95% CI, 81.7–95.3), a specificity of 97% (95% CI, 93.7–98.7), a positive predictive value of 92% (95% CI, 84.3–96.0) and a negative predictive value of 96% (95% CI, 92.6–97.9). Of note, 7% to 10% of patients with NF were stratified by this score as “Low Risk”.[4]

Though the LRINEC score may be somewhat useful in ruling in NF, it is not reliable in ruling out NF, especially in light of more recent validation studies, including a prospective study by Hsiao et al., and a systematic review and meta-analysis by Fernando et al. These studies found the LRINEC score to have even lower sensitivities than in the original study, with LRINEC scores of 6 and above having sensitivities of 43% to 68% and specificities of 60% to 85%.[5,6,7] 

Attempts have been made to propose other scoring systems, such as a score by Wall et al. that ruled in NF if patients had a WBC count > 15.4 and/or Na <135 mmol/L, or the NAS score proposed by Harasawa et al. that utilized MAP, CRP, Hgb, Cr, and glucose in risk stratifying patients with NF. The Wall et al. study, unfortunately, was a small sample size retrospective study conducted at a single center, thus limiting the utility of the scoring system. Additionally, a later retrospective comparative study of the usefulness of this score showed a poor sensitivity of 24% and a positive predictive value of 23%.[7,8]

The NAS score (seen in table 2) in the paper by Harasawa et al. does not include WBC count or Na in its score. Rather it uses vital signs, MAP, and other lab values, to risk stratify 104 patients with soft tissue infections for having NF. In the NAS score, a score of 6 and above is considered high risk and a score of 5 and below is considered low risk for NF. In the study, a NAS score of 6 had a sensitivity of 87.5%, specificity of 91.3%, positive predictive value of 75.0%, and negative predictive value of 96.1%. This study also compared NAS to the LRINEC score, with LRINEC having a sensitivity of 87.5%, specificity of 80%, positive predictive value of 56.8%, and negative predictive value of 95.5%. Though comparable to the LRINEC score, it is important to note that the NAS score does not have any prospective validation studies as of yet.[9]

Table 2: NAS Score

Ultimately, NF is a clinical diagnosis. No scoring system can definitively rule in or out NF. A higher LRINEC score may prompt a sense of urgency in your surgery colleagues to evaluate your patient before completion of imaging, especially if initial clinical signs of NF are equivocal.

No imaging can definitively rule out NF, and different modalities of imaging offer various sensitivities in detecting NF.  

X-Ray:

- Adds little value in the diagnosis of NF as early NF may be mistaken for cellulitis on radiographs, with the common finding of soft tissue thickening and increased opacity.[10]  

- Although the late classical finding of dissecting gas along fascial planes in the absence of trauma may be a specific sign for NF, it is only seen in 25% to 55.0% of patients.[10]  

- X-rays may be more sensitive than physical exams in the detection of soft-tissue gas, with radiographic findings being present before detectable clinical crepitus, but their utility for detecting deep fascial gas is limited.[10,11] 

CT:

- CT imaging without contrast will show asymmetric fascial thickening and fat stranding and is more sensitive than X-rays in aiding the diagnosis of NF with a sensitivity of 80%.[10] CT imaging with contrast will show fascial involvement and lack of enhancement of the fascia and has a sensitivity of >92%.[12] The most specific finding, gas within fluid collections along subfascial planes, is better visualized on CT than X-ray, but is present in less than 55% of cases.[13,14]  

MR:

- MR is the most sensitive modality with a sensitivity of >93%.[13,14] When MR detects a large, low-signal foci of gas, there is a specificity of 100%, though sensitivity in these cases are around 43%.[13] Additionally, the absence of deep fascial abnormality on MR was found to have a high negative predictive value.[15] 

- Although MR imaging has a higher sensitivity compared to CT, it is important to consider the benefits of CT over MR, including speed, cost, ease of use, and greater accessibility for patients with metal implants. 

Point-of-Care Ultrasound:

The main advantage of POCUS is that it can be done at bedside and aid in the diagnosis of NF without the need to wait for lab results and CT/MR imaging. Additionally, POCUS was found to have a sensitivity of 88.2% (95% CI, 63.6–98.5), specificity of 93.3% (95% CI, 81.7–98.6), positive predictive value of 83.3% (95% CI, 58.6–96.4), negative predictive value of 95.4% (95% CI, 84.5–99.4), and accuracy of 91.9% (95% CI, 82.2–97.3) in regards to the diagnosis of NF.[16] Viewing the suspected area of involvement with a linear probe, common POCUS findings characteristic of NF are subcutaneous tissue thickening, fascial thickening, fascial fluid, subcutaneous air, and abnormal muscle architecture.[17,18,19] Subcutaneous tissue thickening is referred to as “cobblestoning” and is characterized by thickening and increased echogenicity of the hypodermal tissue, which is crossed by thin ribbons of hypoechoic tissue that dissociate the fat lobules;[14] subcutaneous air can be seen as multiple echogenic foci with posterior “dirty acoustic shadowing” (as opposed to clean, uniform shadowing)[10] as shown in figure 1.

Figure 1: Cobblestoning and Multiple Echogenic Foci with Posterior Dirty Acoustic Shadowing[20] 

Table 3: The sensitivity, specificity, and accuracy of diagnosis of NF by depth of fluid accumulation on ultrasonography.

Figure 2: On the left, arrows indicate anechoic (black) fluid tracking along the echogenic (bright white) deep fascial plane. On the right, arrowheads delineate the thickened, irregular fascia with anechoic fluid (black) tracking along the fascial plane.[19]

Figure 3: The “plus symbols” delineate the superficial aspect of the deep fascial plane and deep
aspect of the distorted edematous muscle with the arrows indicating a thin layer of
anechoic (black) fluid tracking along the echogenic (bright white) deep fascial plane.[19]

Another bedside test you can utilize is the “finger test”, which has a sensitivity of 100% (95% CI, 69.2–100), a specificity of 80% (95% CI, 59.3–93.2), a positive predictive value of 66.7% (95% CI, 38.4–88.2), a negative predictive value of 100% (95% CI, 83.2–100), and an overall accuracy of 85.7% (95% CI, 69.7–95.2).[21] To perform the “finger test”, first infiltrate the area of suspected involvement with local anesthesia, make a 2-cm incision down to the deep fascia, then gently probe the area with your index finger.[22,23] The presence of characteristic “gray-dishwater discharge”, lack of bleeding on tissue surgical dissection, and minimal tissue resistance to blunt finger dissection are all positive findings correlated with NF.[22,23,24]

Figure 4: Left: Physician performing the “finger test”. Right: Diagram of the space between the subcutaneous tissue and fascia during the “finger test.”[21,25] 

Figure 5: Left: Green arrow pointing to “dirty-dishwater discharge” [Image obtained intra-operatively from our case]. Right: stock photo of “dirty dishwater” for comparison.

The patient was taken to the OR four times for debridement, partial closure, re-exploration, and wound vac placement. Blood cultures grew Group A strep and Methicillin Resistant Staph epidermidis, and abscess cultures grew Peptostreptococcus asaccharolyticus; the antibiotic regimen was tailored to treat these specific microbes. On hospital day 12, the patient was transferred to another hospital for plastic surgery skin grafting and was ultimately discharged to subacute rehab after a 23-day hospitalization.