Conectar
This study was to create an interpretable machine learning model to predict the risk of mortality within 90 days for ICU patients suffering from pressure ulcers.
We retrospectively analysed 1774 ICU pressure ulcer patients from the Medical Information Mart for Intensive Care (MIMIC)-IV database.
We used the LASSO regression and the Boruta algorithm for feature selection. The dataset was split into training and test sets at a 7:3 ratio for constructing machine learning models. We employed logistic regression and nine other machine learning algorithms to build the prediction model. Restricted cubic spline (RCS) was used to analyse the linear relationship between the Braden score and the outcome, whereas the SHAP (Shapley additive explanations) method was applied to visualise the model's characteristics.
This study compared the predictive ability of the Braden Scale with other scoring systems (SOFA, APSIII, Charlson, SAPSII). The results showed that the Braden Scale model had the highest performance, and SHAP analysis indicated that the Braden Scale is an important influencing factor for the risk of 90-day mortality in the ICU. The restricted cubic spline curve demonstrated a significant negative correlation between the Braden Scale and mortality. Subgroup analysis showed no significant interaction effects among subgroups except for age.
The machine learning-enhanced Braden Scale has been developed to forecast the 90-day mortality risk for ICU patients suffering from pressure ulcers, and its efficacy as a clinically reliable tool has been substantiated.
Patients or public members were not directly involved in this study.
Chronic wounds pose an increasing burden on the healthcare system and data on wound outcomes and are needed to evaluate and address disparities and reform healthcare policies. In Singapore, data on chronic wounds are fragmented and to address this, we established a Chronic Wounds Registry (CWR) to collect harmonised data on chronic wounds and their outcomes over 6 months. This is a multi-centre prospective cohort study from Nov 2019 to Nov 2021. Patients with chronic wounds were enrolled at multi-speciality acute care hospitals and data were prospectively collected on baseline characteristics, including subject demographics, clinical data, wound images, interventions/treatment, cost burden and patient reported health-related quality of life (HRQOL). Patients were followed up for 6 months and wound outcomes recorded at Month 1, Month 3 and Month 6 time points. Despite the onset of COVID pandemic, a total of 812 patients were recruited in our study. Mean age was 63.5 ± 11.6 years with 66% men and 59% of Chinese ethnicity. Twenty percent of all the wounds were recurrent and patients with venous leg ulcers (VLU) (32%) had the highest recurrence. At 6 months, 225 (46%) of the neuro-ischemic ulcers (NIU), 152 (60%) of the VLUs and 29 (46%) of the pressure injuries (PI) had healed. Major (5%) amputations were highest in patients with NIUs. All-cause mortality was highest (30%) and HRQOL was lowest for patients with PIs (−0.18). Development of a national wound registry is both feasible and essential to consolidate key data elements on chronic wounds. The CWR in its current state captured the local epidemiology, patient journey in acute care hospitals, which will benefit in healthcare policymaking and harmonise care across different levels of healthcare system. The next phase of the CWR aims to track patients in all settings and collect data on the entire patient journey following an episode of wounding.
Emergency craniotomy in patients with traumatic brain injury poses a significant risk for surgical site infections (SSIs). Understanding the risk factors and pathogenic characteristics of SSIs in this context is crucial for improving outcomes. This comprehensive retrospective analysis spanned from February 2020 to February 2023 at our institution. We included 25 patients with SSIs post-emergency craniotomy and a control group of 50 patients without SSIs. Data on various potential risk factors were collected, including demographic information, preoperative conditions, and intraoperative details. The BACT/ALERT3D Automated Bacterial Culture and Detection System was utilized for rapid bacterial pathogen identification. Statistical analyses included univariate and multivariate logistic regression to identify significant risk factors for SSIs. The study identified Klebsiella pneumoniae, Escherichia coli, and Staphylococcus aureus as the most prevalent pathogens in SSIs. Significant risk factors for SSIs included the lack of preoperative antibiotic use, postoperative drainage tube placement, diabetes mellitus, and the incorporation of invasive procedures, all of which showed a significant association with SSIs in the univariate analysis. The multivariate analysis further highlighted the protective effect of preoperative antibiotics and the increased risks associated with anaemia, diabetes mellitus, postoperative drainage tube placement, and the incorporation of invasive procedures. Our research underscores the critical role of factors like insufficient preoperative antibiotics, postoperative drainage, invasive techniques, anaemia, and diabetes mellitus in elevating the risk of surgical site infections in traumatic brain injury patients undergoing emergency craniotomy. Enhanced focus on these areas is essential for improving surgical outcomes.
In recent years, breast conservation surgery (BCS) and modified radical mastectomy (MRM) have been widely used in the treatment of early-stage breast cancer. However, the effects of complications from these two surgical methods are still unclear. The purpose of this study was to evaluate the effects of BCS and MRM on postoperative wound complications in patients with early breast cancer. Eighty-eight patients with early breast cancer were randomly divided into BCS group (n = 46) and MRM group (n = 42). The occurrence of postoperative wound complications was compared between the two groups. The results showed that BCS patients had significantly lower intraoperative surgical times, blood losses, incision lengths, drainage volumes, drainage times and lengths of hospital stays than their MRM counterparts (p < 0.05). The overall incidence of postoperative wound complications in the BCS group was significantly lower than that in the MRM group at 1 month after surgery, but the difference was not statistically significant (p < 0.001). The rate of excellent breast cosmetic outcome was significantly higher in the BCS group than in the MRM group (p < 0.001). Compared to MRM, BCS has a lower incidence of intraoperative and postoperative wound complications, better cosmetic outcomes and higher clinical value in the treatment of early-stage breast cancer.
Diabetic foot ulcer (DFU), a common intractable chronic complication of diabetes mellitus (DM), has a prevalence of up to 25%, with more than 17% of the affected patients at risk of amputation or even death. Vascular risk factors, including vascular stenosis or occlusion, dyslipidemia, impaired neurosensory and motor function, and skin infection caused by trauma, all increase the risk of DFU in patients with diabetes. Therefore, diabetic foot is not a single pathogenesis. Preclinical studies have contributed greatly to the pathogenesis determination and efficacy evaluation of DFU. Many therapeutic tools are currently being investigated using DFU animal models for effective clinical translation. However, preclinical animal models that completely mimic the pathogenesis of DFU remain unexplored. Therefore, in this review, the preparation methods and evaluation criteria of DFU animal models with three major pathological mechanisms: neuropathy, angiopathy and DFU infection were discussed in detail. And the advantages and disadvantages of various DFU animal models for clinical sign simulation. Furthermore, the current status of vitro models of DFU and some preclinical studies have been transformed into clinical treatment programs, such as medical dressings, growth factor therapy, 3D bioprinting and pre-vascularization, Traditional Chinese Medicine treatment. However, because of the complexity of the pathological mechanism of DFU, the clinical transformation of DFU model still faces many challenges. We need to further optimize the existing preclinical studies of DFU to provide an effective animal platform for the future study of pathophysiology and clinical treatment of DFU.
FreshRSS 