Laboratory Risk Indicator for Necrotizing Fasciitis Score and Patient Outcomes.

CONTEXT: The Laboratory Risk Indicator for Necrotizing Fasciitis (LRINEC) score helps to diagnose necrotizing soft-tissue infection (NSTI). The LRINEC score has been reported to be associated with poor prognosis, although few studies have evaluated this association. AIMS: We aimed to describe the characteristics of NSTI and assess whether the LRINEC score was associated with mortality and amputation. SETTINGS AND
DESIGN: We conducted a retrospective observational study from January 2007 to May 2018, in a Japanese tertiary care hospital. SUBJECTS AND METHODS: Patients with NSTI were identified through our hospital database using the discharge diagnosis. We extracted data on patient characteristics, laboratory examinations, microbiological information, treatment, and in-hospital mortality. STATISTICAL ANALYSIS USED: We estimated the odds ratios (ORs) and associated 95% confidence intervals (CIs) for in-hospital mortality using logistic regression models.
RESULTS: We identified 58 patients. The median LRINEC score was 8 (interquartile range [IQR]: 6-9). Forty-four patients (75.9%) scored 6 or more. The eight patients with amputations had a median score of 6 (IQR: 4.5-7.5) versus 8 (IQR: 7-9) for patients who underwent debridement (P = 0.091). Survivors and nonsurvivors had median scores of 8 (IQR: 6-9) and 6 (IQR: 5-8), respectively (P = 0.148). The OR for mortality in patients with liver cirrhosis was 10.5 (95% CI: 1.00-110.36; P = 0.050).
CONCLUSIONS: There was no association between the LRINEC score and patients' outcomes: mortality and amputation. Further studies are warranted to evaluate the utility of the LRINEC score and factors associated with poor prognosis in patients with NSTI. Copyright:

INTRODUCTION

Necrotizing fasciitis and other necrotizing soft-tissue infections (NSTIs) are life-threatening, medical, and surgical emergencies. NSTI rapidly progress, involving the fascia and subcutaneous tissues. The reported mortality rate ranged between 20% and 40% despite radical treatment.[123] Early diagnosis and aggressive surgical treatment can reduce the mortality and morbidity.[4]

The symptoms that help identify NSTI include fever, hypotension, subcutaneous hemorrhage, hemorrhagic bullae, necrosis, and purpura. However, these symptoms are not specific to NSTI, thereby making it difficult for physicians to distinguish NSTI from other soft-tissue infections. Therefore, it is often time-consuming to attain correct diagnoses.[12567]

In 2004, Wong et al. recommended the Laboratory Risk Indicator for Necrotizing Fasciitis (LRINEC score), based on routine laboratory investigations, that can help to distinguish necrotizing fasciitis from other soft-tissue infections. They reported that a score ≥6 had a sensitivity of 92.9% and a specificity of 91.6% for indicating necrotizing fasciitis.[5]

Another previous study suggested that patients with an LRINEC score ≥6 had a higher risk of mortality and amputation than those with a score <6.[8] However, few studies, especially those in Japan, have evaluated the association of the LRINEC score with mortality and the necessity of amputation. Moreover, in Japan, a limited number of studies have reported on the characteristics of NSTI.

This study therefore aimed to describe the characteristics of NSTI and to assess the association between the LRINEC score and mortality and the necessity of amputation in a Japanese large tertiary care hospital.

SUBJECTS AND METHODS

We conducted a retrospective observational study involving patients with NSTI admitted to Kurashiki Central Hospital from January 2007 to May 2018. Kurashiki Central Hospital is a 1166-bedded tertiary care hospital located in the Western part of Japan and covers approximately 800,000 people. Approximately five patients with NSTIs were annually referred and admitted to the hospital. Patients with NSTI were identified through our hospital database using the discharge diagnosis of necrotizing fasciitis and gas gangrene; these cases were registered with ICD-10 (International Classification of Diseases, Tenth Revision) codes A48.0 and M72.6. In this study, diagnoses were considered eligible if made based on one of the following factors: (1) histopathological examinations of tissues, (2) intraoperative findings, and (3) clinical grounds by the physicians.

The following information was extracted from the medical charts: (1) patient characteristics (age; sex; and medical history, including diabetes mellitus, liver cirrhosis, chronic kidney diseases, hemodialysis, and alcoholism, use of immunomodulatory drugs (e.g., systemic corticosteroids), cancer, and infection sites); (2) laboratory examinations (white blood cell counts and levels of hemoglobin, sodium, creatinine, glucose, and C-reactive protein); (3) microbiological information when cultures were submitted; (4) treatment (surgical debridement, amputation, conservative therapy, and palliative care); and (5) in-hospital mortality and the length of hospital stay. The LRINEC scores on admission were calculated for each patient using the data obtained from the medical charts.

To analyze the patient characteristics and outcomes, we used the Mann–Whitney test for continuous variables and Fisher's exact test for categorical variables. We performed univariate logistic regression analysis to estimate odds ratios (ORs) and their 95% confidence intervals (CIs) for in-hospital mortality. Receiver operating characteristic (ROC) curves were calculated to assess the LRINEC score for predicting outcomes (amputation and in-hospital mortality). Analyses were performed using Stata version 15.1 (Stata, College Station, TX, USA). A two-tailed P < 0.05 was considered to be statistically significant. This study was performed in accordance with the Declaration of Helsinki and approved by the Kurashiki Central Hospital Institutional Ethics Committee (No. 2184, July 23, 2018). The need for informed consent from the participants was waived.

RESULTS

We identified a total of 120 with relevant ICD-10 codes.

Following the removal of duplicates, we reviewed the medical records of 85 patients. We eventually included a total of 58 patients with NSTI, including 34 males and 24 females, with a median age of 66.5 years (interquartile range [IQR]: 57–75). There were 36 (62.1%) cases of diabetes mellitus, 4 (6.9%) liver cirrhosis, 10 (17.2%) chronic kidney disease, 4 alcoholism (6.9%), 8 (13.8%) malignancy, 3 (5.2%) on hemodialysis, and 8 (13.8%), who used immunomodulators. The median LRINEC score was 8 (IQR: 6–9). Regarding the sensitivity of the LRINEC score, 44 patients (75.9%) scored 6 or more, which was the threshold for the diagnosis of NSTI. Thirty patients (51.7%) scored 8 or more on admission to the hospital. The most common source of infection was found in lower limbs (27 patients, 46.6%), followed by perianal areas (14 patients, 24.1%) and torso (11 patients, 19.0%) [Table 1].

Table 1

Characteristics of the study participants

	Total (n=58)	LRINEC score <6 (n=14)	LRINEC score ≥6 (n=44)	P
Sex, male, n (%)	34 (58.6)	11 (78.6)	23 (52.3)	0.121
Age (year), median (IQR)	66.5 (57-75)	68 (57-80)	65 (57.5-73.5)	0.429
Comorbidities, n (%)
Diabetes mellitus	36 (62.1)	8 (57.1)	28 (63.6)	0.756
Hypertension	23 (39.7)	8 (57.1)	15 (34.1)	0.209
Liver cirrhosis	4 (6.9)	3 (21.4)	1 (2.3)	0.040
Chronic kidney disease	10 (17.2)	1 (7.1)	9 (20.5)	0.424
Hemodialysis	3 (5.2)	1 (7.1)	2 (4.6)	1.00
Malignancy	8 (13.8)	2 (14.3)	6 (13.6)	1.00
Alcoholism	4 (6.9)	2 (14.3)	2 (4.6)	0.243
PAD	13 (22.4)	2 (14.3)	11 (25)	0.489
Immunomodulators	8 (13.8)	0 (0)	8 (18.2)	0.179
SBP <90 mmHg, n (%)	23 (39.7)	4 (28.6)	19 (43.2)	0.369
Sites of infections, n (%)
Upper limb	6 (10.3)	1 (7.1)	5 (11.4)
Lower limb	27 (46.6)	6 (42.9)	21 (47.7)
Perineum	14 (24.1)	4 (28.6)	10 (22.7)
Trunk	11 (19)	3 (21.4)	8 (18.2)	0.968
Mortality, n (%)	15 (25.9)	5 (35.7)	10 (22.7)	0.484
Length of stay, days, median (IQR)	28 (17-60)	23.5 (16-46)	37.5 (17-65)	0.206
Length of stay in survival group, days, median (IQR)	46 (23-65)	24 (24-57)	48.5 (23-75)	0.339
Underwent surgical intervention, n (%)	51 (87.9)	12 (85.7)	39 (88.6)	1.00
Amputation, n (%)*	8 (24.2)	2 (28.6)	6 (23.1)	1.00

*Affected region were extremities. SD: Standard deviation, IQR: Interquartile range, SBP: Systolic blood pressure, PAD: Peripheral artery disease, LRINEC: Laboratory Risk Indicator for Necrotizing Fasciitis

Fifty-five patients had positive cultures. Twenty-two patients had monoculture (37.9%), while four of them had potentially polymicrobial flora because they developed Fournier gangrene, or the causative agent was Enterococcus. Polymicrobial flora was found in 32 patients, while only one patient had a positive monoculture of a fungus (Candida tropicalis). Staphylococcus aureus was the most common pathogen observed, with detection seen in 18 patients (31.0%).

Fifteen patients died (25.9%) during hospitalization (10 patients with a LRINEC score ≥6 and 5 with a LRINEC score <6; P = 0.484). A total of 51 patients underwent at least one surgery, and 43 patients underwent surgical debridement alone. Amputations were performed in 8 (27.6%) out of 29 patients with extremity surgery. Two patients with surgical debridement and six with extremity surgery died. Conservative therapy and/or palliative care was instituted for 7 patients (12.1%), all of whom died. The eight patients who underwent amputations had a median LRINEC score of 6 (IQR: 4.5–7.5), as opposed to 8 (IQR: 7-9) among 21 patients undergoing only debridement (P = 0.091). The area under the ROC curve for amputation was 0.330 (95% CI: 0.101–0.559) [Figure 1]. Survivors and nonsurvivors had a median LRINEC score of 8 (IQR: 6–9) and 6 (IQR: 5–8), respectively (P = 0.148). The area under the ROC curve for in-hospital mortality was 0.354 (95% CI: 0.206–0.501) [Figure 2].

Figure 1

Receiver operating characteristic curve for Laboratory Risk Indicator for Necrotizing Fasciitis score for predicting amputation. The area under the curve was 0.330

Figure 2

Receiver operating characteristic curve for Laboratory Risk Indicator for Necrotizing Fasciitis score for predicting in-hospital mortality. The area under the curve was 0.354

Patients were also classified by the risk estimated with the LRINEC score. We examined the association between LRINEC score and infection sites, comorbidities and mortality. In the univariate analysis, the OR for mortality in patients with liver cirrhosis was 10.5 (95% CI: 1.00–110.36; P = 0.050); however, this association was not statistically significant [Table 2].

Table 2

Odds ratios for mortality

	Crude model, OR (95% CI)	P
LRINEC score ≥6	0.53 (0.14-1.94)	0.338
SBP ≤90 mmHg	3.11 (0.92-10.46)	0.067
Age (year)	1.03 (0.98-1.08)	0.278
Sex, male	1.08 (0.33-3.58)	0.900
Sites of infections
Upper limb	Reference	-
Lower limb	0.84 (0.13-5.56)	0.858
Perineum	0.33 (0.03-3.2)	0.341
Trunk	0.75 (0.09-6.47)	0.794
Diabetes mellitus	0.61 (0.19-2.02)	0.420
Hypertension	0.46 (0.13-1.67)	0.238
Immunomodulators	0.95 (0.17-5.3)	0.952
Liver cirrhosis	10.5 (1.00-110.36)	0.050
Malignancy	0.37 (0.04-3.26)	0.369
Hemodialysis	1.46 (0.12-17.41)	0.763
Alcoholism	3.15 (0.40-24.67)	0.274
PAD	0.45 (0.09-2.3)	0.336
Chronic kidney disease	0.67 (0.13-3.59)	0.643

CI: Confidence interval, OR: Odds ratio; SBP: Systolic blood pressure, PAD: Peripheral artery disease, LRINEC: Laboratory Risk Indicator for Necrotizing Fasciitis

DISCUSSION

In the present study, we found no association between the LRINEC score with either the mortality or the necessity of amputation. However, the analysis of the association was performed on only a small number of cases. Regarding the sensitivity of the LRINEC score, only 44 (75.9%) patients with NSTI scored ≥6 on admission, which suggests that the sensitivity is lower for the early diagnosis of NSTI.

Our analysis suggested that there was no association between the LRINEC score and patient prognosis. However, Tilkorn et al. similarly reported no significant difference in the LRINEC score of survivors and nonsurvivors.[9] The association between the LRINEC score and the treatment outcomes varied between these studies, which may be derived from multiple factors (e.g., differences in the distribution of each patient's predisposing factors or etiologies in the studies). Ray-Zack et al. reported a grading system called the American Association for the Surgery of Trauma Emergency General Surgery grade (the AAST-EGS grade) for skin and soft-tissue infection. This system demonstrated a more accurate association of the LRINEC score with clinical outcomes.[4] This score includes anatomic severity and clinical features, which may contribute to improving the accuracy. Knowledge of underlying disease severity, comorbidities, and acute physiology rather than laboratory data may be more crucial for treatment outcomes, given the clinical discrepancy between the course of the disease and its manifestation.[4]

For patients with NSTI, early recognition and prompt aggressive debridement of all necrotic tissues has been shown to improve the survival.[35101112] However, the paucity of cutaneous findings early in the course of the disease makes early diagnosis difficult. Only 14.6% of the patients were diagnosed or suspected as having necrotizing fasciitis at the time of admission to the hospital.[5] Regarding the usefulness of the LRINEC score for the diagnosis of NSTI, a recent meta-analysis reported that the sensitivity and specificity of the LRINEC score were 68.2% and 84.8%, respectively.[13] These figures were lower than those initially reported by Wong et al. (i.e., 92.9% and 91.6%, respectively). A score with a cutoff of 6 points would miss over 30% of cases in the meta-analysis[13] and 24% in our cohort. Furthermore, given the low incidence rate for NSTI,[14] the positive predictive value could be even lower. Many previous studies examining the LRINEC score were retrospective in design, and not all patients in the non-disease groups were suspected to have NSTI. Therefore, the LRINEC score should be validated prospectively in patients with suspected NSTI, regardless of its likeness in the initial assessment.

Diabetes mellitus was the most common comorbidity in our patients. Univariate logistic regression analysis suggested that liver cirrhosis was narrowly associated with in-hospital mortality. Diabetes mellitus, liver cirrhosis, and other immunocompromised conditions were known as predisposing factors.[15] Patients with liver cirrhosis were susceptible to infections and had excessive activation of pro-inflammatory cytokines. These factors might be associated with the worse treatment outcome.[16]

This study had some limitations that should be addressed. First, we could only perform a univariate analysis because of the relatively small number of patients who died during hospitalization. Second, as a single-center study, the generalizability of our findings may be limited. However, because our hospital covers most emergency patients and referrals of critically ill patients from Kurashiki, a typical medium-sized city in Japan, we believe that our study participants were a representative sample of NSTI patients in Japan.

CONCLUSIONS

This study did not find any association between the LRINEC score and treatment outcomes, mortality and amputation rates. However, liver cirrhosis might be associated with mortality. Further studies are warranted to evaluate the utility of the LRINEC score and factors associated with poor prognosis in patients with NSTI.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.