Differentiating Lower Extremity Necrotizing Soft Tissue Infection from Severe Cellulitis by Laboratory Parameters and Relevant History Points

Machine Learning Approach to Classify Vibrio vulnificus Necrotizing Fasciitis, Non-Vibrio Necrotizing Fasciitis and Cellulitis

Background

Recent advancements in artificial intelligence have led to increased adoption of machine learning in disease identification, particularly for challenging diagnoses like necrotizing fasciitis and Vibrio vulnificus infections. This shift is driven by the technology's efficiency, objectivity, and accuracy, offering potential solutions to longstanding diagnostic hurdles in clinical practice.

Methods

This investigation incorporated 180 inpatients suffering from soft tissue infections. The participants were categorized into groups: cellulitis, non-Vibrio necrotizing fasciitis (NF), or V. Vulnificus NF. To predict the three relevant outcomes, we employed Light Gradient Boosting Machine (LightGBM) and 5-fold cross-validation methodologies for the development of a multi-class categorization model. Moreover, we applied the SHapley Additive exPlanations (SHAP) methodology to decipher the model's predictions.

Results

The multi-classification model possesses substantial predictive capacity, with a weighted-average AUC of 0.86, sensitivity of 87.2%, specificity of 74.5%, NPV of 81.6%, and PPV of 85.4%. The model's calibration was assessed using the Brier score, yielding a weighted mean of 0.084. This low value demonstrates a strong correlation between predicted probabilities and actual outcomes, indicating high predictive accuracy and reliability in the model's forecasts.

Conclusions

We effectively developed a multiclassification model aimed at forecasting the occurrence of cellulitis, non-Vibrio NF, or V. Vulnificus NF in patients suffering from soft tissue infection, and we further described the model's predictions using the SHAP algorithm.

Biochemical analysis of soft tissue infectious fluids and its diagnostic value in necrotizing soft tissue infections: a 5-year cohort study

BACKGROUND

Necrotizing soft tissue infections (NSTI) are rapidly progressing and life-threatening conditions that require prompt diagnosis. However, differentiating NSTI from other non-necrotizing skin and soft tissue infections (SSTIs) remains challenging. We aimed to evaluate the diagnostic value of the biochemical analysis of soft tissue infectious fluid in distinguishing NSTIs from non-necrotizing SSTIs.

METHODS

This cohort study prospectively enrolled adult patients between May 2023 and April 2024, and retrospectively included patients from April 2019 to April 2023. Patients with a clinical suspicion of NSTI in the limbs who underwent successful ultrasound-guided aspiration to obtain soft tissue infectious fluid for biochemical analysis were evaluated and classified into the NSTI and non-necrotizing SSTI groups based on their final discharge diagnosis. Common extravascular body fluid (EBF) criteria were applied.

RESULTS

Of the 72 patients who met the inclusion criteria, 10 patients with abscesses identified via ultrasound-guided aspiration were excluded. Based on discharge diagnoses, 39 and 23 patients were classified into the NSTI and non-necrotizing SSTI groups, respectively. Biochemical analysis revealed significantly higher albumin, lactate, lactate dehydrogenase (LDH), and total protein levels in the NSTI group than in the non-necrotizing SSTI group, and the NSTI group had significantly lower glucose levels and pH in soft tissue fluids. In the biochemical analysis, LDH demonstrated outstanding discrimination (area under the curve (AUC) = 0.955; p < 0.001) among the biochemical markers. Albumin (AUC = 0.884; p < 0.001), lactate (AUC = 0.891; p < 0.001), and total protein (AUC = 0.883; p < 0.001) levels also showed excellent discrimination. Glucose level (AUC = 0.774; p < 0.001) and pH (AUC = 0.780; p < 0.001) showed acceptable discrimination. When the EBF criteria were evaluated, the total scores of Light's criteria (AUC = 0.925; p < 0.001), fluid-to-serum LDH ratio (AUC = 0.929; p < 0.001), and fluid-to-serum total protein ratio (AUC = 0.927; p < 0.001) demonstrated outstanding discrimination.

CONCLUSION

Biochemical analysis and EBF criteria demonstrated diagnostic performances ranging from acceptable to outstanding for NSTI when analyzing soft tissue infectious fluid. These findings provide valuable diagnostic insights into the recognition of NSTI. Further research is required to validate these findings.

Necrotizing Fasciitis: When skin confuses - An autopsy case report

Necrotizing Fasciitis (NF) is a severe life-threatening soft tissue infection characterized by the rapid destruction of muscle, fat and fascial layers. This report details an autopsy case report of a 40year old male, unclaimed body lacking the complete history except that given by the Police personnel accompanying in which there is no prior history of trauma. This person succumbed to septic shock secondary to NF, despite clinical interventions. This case emphasizes the importance of early diagnosis and the need for heightened clinical awareness to improve patient outcomes.

Identifying necrotizing soft tissue infection using infectious fluid analysis and clinical parameters based on machine learning algorithms

Background

Determining the presence of necrotizing soft tissue infection (NSTI) poses a significant hurdle. As of late, there has been a notable increase in the application of artificial intelligence (AI) machine learning techniques in identifying diseases, a shift that can be attributed to their exceptional efficiency, unbiased nature, and high precision.

Methods

Information was gathered from a cohort of 13 patients suffering from NSTI, alongside 12 patients with cellulitis. The construction of NSTI diagnostic machine learning models utilized four different algorithms, specifically random forest, k-nearest neighbors (KNN), support vector machine (SVM), and logistic regression. These models were constructed based on 28 distinctive attributes identified through statistical examination. Following this, the diagnostic efficiency of each algorithms was evaluated. A novel random forest model, streamlined for clinical use, was later developed by focusing on 6 attributes that had the most pronounced influence on the accuracy of our initial random forest model.

Results

The following data was noted regarding the sensitivity and specificity of the four NSTI diagnostic models:logistic regression displayed 78.2 % and 83.7 %, KNN presented 79.1 % and 87.1 %, SVM showed 83.5 % and 86.3 %, and random forest exhibited 89.6 % and 92.9 %, respectively. In comparison, lactate levels in fluid demonstrated 100 % sensitivity and 76.9 % specificity at an optimal cut-off point of 69.6 mg/dL. Among all four machine learning models, random forest outperformed the others and also showed better results than fluid lactate. A newly constructed random forest model, created using 6 of the 13 identified features, displayed promising results in diagnosing NSTI, having a sensitivity and specificity of 90.2 % and 92.2 %, respectively.

Conclusions

Developing a diagnostic model for NSTI employing the random forest algorithm has resulted in a diagnostic technique that is more efficient, cost-effective, and expedient. This approach could provide healthcare practitioners with the tools to identify and manage NSTI with greater efficacy.

Role of Procalcitonin for Early Discrimination Between Necrotizing Fasciitis and Cellulitis of the Extremities

Introduction Necrotizing fasciitis (NF) is a grave and life-threatening infection of the soft tissues. It is defined by the gradual necrosis of the fascia and subcutaneous tissue, which spreads along the fascial planes. Cellulitis, a prevalent skin infection, has led to suggestions that procalcitonin could serve as a diagnostic tool to distinguish it from other inflammatory skin conditions that resemble cellulitis. The study aims to assess the procalcitonin (PCT) levels in individuals with NF and cellulitis and determine its effectiveness in early differentiation between these two conditions. Methods After obtaining clearance from the institutional ethical committee, the study was conducted in the Department of General Surgery, Sri Devaraj Urs Medical College, over six months. Informed consent was obtained from all 30 patients included in this study. The study compared PCT levels in patients diagnosed with NF and cellulitis. Statistical analysis was performed using SPSS version 22 software (IBM Corp., Armonk, NY, USA). Results The mean age of subjects was 53.23 ± 8.78 years. Among patients, 21 (70%) were diagnosed with cellulitis and 9 (30%) were diagnosed with NF. The mean PCT levels were 0.34 ± 0.32 and 4.89 ± 1.98 among the cellulitis and NF groups, respectively. There was a significant difference (p<0.05). PCT had a sensitivity of 100% and a specificity of 100%, in differentiating cellulitis and necrotizing fasciitis. Conclusion PCT levels were notably elevated in cases of NF compared to cellulitis. Despite the study's limited sample size, it represents the first report highlighting the value of PCT as an early diagnostic tool for identifying necrotizing fasciitis.

Comparison of National Early Warning Score with shock index in patients with necrotizing fasciitis

Shock index (SI) and national early warning score (NEWS) are more frequently used as assessment tools in acute illnesses, patient disposition and early identification of critical condition. Both they are consisted of common vital signs and parameters including heart rate, systolic blood pressure, respiratory rate, oxygen saturation and level of conscious, which made it easy to evaluate in medical facilities. Its ability to predict mortality in patients with necrotizing fasciitis (NF) in the emergency department remains unclear. This study was conducted to compare the predictive capability of the risk scores among NF patients. A retrospective cohort study of hospitalized patients with NF was conducted in 2 tertiary teaching hospitals in Taiwan between January 2013 and March 2015. We investigated the association of NEWS and SI with mortality in NF patients. Of the 395 NF patients, 32 (8.1%) died in the hospital. For mortality, the area under the receiver curve value of NEWS (0.81, 95% confidence interval 0.76-0.86) was significantly higher than SI (0.76, 95% confidence interval 0.73-0.79, P = .016). The sensitivities of NEWS of 3, 4, and 5 for mortality were 98.1%, 95.6%, and 92.3%. On the contrast, the sensitivities of SI of 0.5, 0.6, and 0.7 for mortality were 87.8%, 84.7%, and 81.5%. NEWS had advantage in better discriminative performance of mortality in NF patients. The NEWS may be used to identify relative low risk patients among NF patients.

Prognostic factors for intraoperative detection of necrotizing fasciitis in severe soft tissue infections

BACKGROUND

Necrotizing fasciitis (NF) is a rare but lethal soft-tissue infection. There is still a paucity of information regarding the diagnostic tools and therapeutic strategies for the treatment of this devastating disease. This study aims to identify important perioperative parameters related to necrotizing fasciitis and to assess their relevance in terms of identifying NF.

METHODS AND MATERIAL

We retrospectively analyzed patients who underwent surgical exploration for suspected necrotizing fasciitis at a tertiary referral center, to explore the clinical features and factors related to the presence of necrotizing fasciitis and mortality.

RESULTS

Between 2010 and 2017, 88 patients underwent surgical exploration for suspected NF. The infection occurred in the lower extremities in 48 patients, in the thoracocervical region in 18 patients, and the perineum and abdomen in 22 patients. Histological evidence of NF was present in 59 of 88 patients. NF was associated with a longer hospital stay and ICU stay (p = 0.05 and 0.019 respectively) compared to patients without NF. ROC analysis showed that only macroscopic fascial appearance could discriminate patients with histological evidence of NF. Moreover, multivariate logistic regression revealed, that liver failure (p = 0.019), sepsis (p = 0.011), positive Gram stain (p = 0.032), and macroscopic fascial appearance (p <0.001) were independent prognostic parameters for histological evidence of NF.

CONCLUSION

Intraoperative tissue evaluation by an experienced surgeon is the most important diagnostic tool in identifying necrotizing fasciitis. An intraoperative Gram stain is an independent prognostic tool and therefore its use can be recommended especially in case of clinical uncertainty.

Differentiating necrotizing soft tissue infections from cellulitis by soft tissue infectious fluid analysis: a pilot study

Background

We conducted this study to evaluate the characteristics of the infectious fluid in soft tissue infection and investigate the utility of the biochemical tests and Gram stain smear of the infectious fluid in distinguishing necrotizing soft tissue infection (NSTI) from cellulitis.

Methods

This retrospective cohort study was conducted in a tertiary care hospital in Taiwan. From April 2019 to October 2020, patients who were clinically suspected of NSTI with infectious fluid accumulation along the deep fascia and received successful ultrasound-guided aspiration were enrolled. Based on the final discharge diagnosis, the patients were divided into NSTI group, which was supported by the surgical pathology report, or cellulitis group. The t test method and Fisher’s exact test were used to compare the difference between two groups. The receiver–operator characteristic (ROC) curves and area under the ROC curve (AUC) were used to evaluate the discriminating ability.

Results

Total twenty-five patients were enrolled, with 13 patients in NSTI group and 12 patients in cellulitis group. The statistical analysis showed lactate in fluid (AUC = 0.937) and LDH in fluid (AUC = 0.929) had outstanding discrimination. The optimal cut-off value of fluid in lactate was 69.6 mg/dL with corresponding sensitivity of 100% and specificity of 76.9%. The optimal cut-off value of fluid in LDH was 566 U/L with corresponding sensitivity of 83.3% and a specificity of 92.3%. In addition, albumin in fluid (AUC = 0.821), TP in fluid (AUC = 0.878) and pH in fluid (AUC = 0.858) also had excellent diagnostic accuracy for NSTI. The Gram stain smear revealed 50% bacteria present in NSTI group and all the following infectious fluid culture showed bacteria growth.

Conclusions

The analysis of infectious fluid along the deep fascia might provide high diagnostic accuracy to differentiate NSTI from cellulitis.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13017-022-00404-4.