The sudden increase in COVID-19 cases is putting high pressure on healthcare services worldwide. At this stage, fast, accurate and early clinical assessment of the disease severity is vital. To support decision making and logistical planning in healthcare systems, this study leverages a database of blood samples from 485 infected patients in the region of Wuhan, China, to identify crucial predictive biomarkers of disease mortality. For this purpose, machine learning tools selected three biomarkers that predict the mortality of individual patients more than 10 days in advance with more than 90% accuracy: lactic dehydrogenase (LDH), lymphocyte and high-sensitivity C-reactive protein (hs-CRP). In particular, relatively high levels of LDH alone seem to play a crucial role in distinguishing the vast majority of cases that require immediate medical attention. This finding is consistent with current medical knowledge that high LDH levels are associated with tissue breakdown occurring in various diseases, including pulmonary disorders such as pneumonia. Overall, this Article suggests a simple and operable decision rule to quickly predict patients at the highest risk, allowing them to be prioritized and potentially reducing the mortality rate.

Background: Most patients with coronavirus disease (COVID-19) who show mild symptoms are sent home by physicians to recover. However, the condition of some of these patients becomes severe or critical as the disease progresses. Objective: The aim of this study was to evaluate a telemedicine model that was developed to address the challenges of treating patients with progressive COVID-19 who are home-quarantined and shortages in the medical workforce. Methods: A telemedicine system was developed to continuously monitor the progression of home-quarantined patients with COVID-19. The system was built based on a popular social media smartphone app called WeChat; the app was used to establish two-way communication between a multidisciplinary team consisting of 7 medical workers and 188 home-quarantined individuals (including 74 confirmed patients with COVID-19). The system helped patients self-assess their conditions and update the multidisciplinary team through a telemedicine form stored on a cloud service, based on which the multidisciplinary team made treatment decisions. We evaluated this telemedicine system via a single-center retrospective study conducted at Tongji Hospital in Wuhan, China, in January 2020. Results: Among 188 individuals using the telemedicine system, 114 (60.6%) were not infected with COVID-19 and were dismissed. Of the 74 confirmed patients with COVID-19, 26 (35%) recovered during the study period and voluntarily stopped using the system. The remaining 48/76 confirmed patients with COVID-19 (63%) used the system until the end of the study, including 6 patients whose conditions progressed to severe or critical. These 6 patients were admitted to hospital and were stabilized (one received extracorporeal membrane oxygenation support for 17 days). All 74 patients with COVID-19 eventually recovered. Through a comparison of the monitored symptoms between hospitalized and nonhospitalized patients, we found prolonged persistence and deterioration of fever, dyspnea, lack of strength, and muscle soreness to be diagnostic of need for hospitalization. Conclusions: By continuously monitoring the changes in several key symptoms, the telemedicine system reduces the risks of delayed hospitalization due to disease progression for patients with COVID-19 quarantined at home. The system uses a set of scales for quarantine management assessment that enables patients to self-assess their conditions. The results are useful for medical staff to identify disease progression and, hence, make appropriate and timely treatment decisions. The system requires few staff to manage a large cohort of patients. In addition, the system can solicit help from recovered but self-quarantined medical workers to alleviate shortages in the medical workforce and free healthy medical workers to fight COVID-19 on the front line. Thus, it optimizes the usage of local medical resources and prevents cross-infections among medical workers and patients.

Hypertrophic scarring is a widespread skin disorder that affects a patient's confidence and quality of life. Intralesional injection of bleomycin is one of the most commonly employed treatments for hypertrophic scars, which, however, always brings pain and requires long-term treatment. To overcome these limitations, new methods for bleomycin delivery enabling painlessness, self-administration, the fast onset of action, and good bioavailability are urgently needed. In this study, we developed bleomycin-loaded dissolving microneedles (BMN) made up of hyaluronic acid (HA) with excellent aqueous solubility and enhanced efficacy for inhibiting hypertrophic scars. The as-obtained BMN possesses sufficient mechanical strength to pierce porcine skin ex vivo with an insertion depth of over 300 μm. Moreover, BMN can dissolve rapidly and release 20% of loaded bleomycin within 1 min and 52% within 10 min, and the BMN-treated skin could recover to its original status within 3 h, demonstrating good biocompatibility. Besides, the HA matrix also maintains the stability and activity of bleomycin. Furthermore, we show that BMN consisting of HA and bleomycin can inhibit the proliferation of human hypertrophic scar fibroblasts (hHSFs) and the secretion of transforming growth factor-β (TGF-β1) in vitro. Therefore, bleomycin-loaded dissolving HA microneedles provide a potential route to treat hypertrophic scars in a convenient, efficient, and minimally invasive manner.

With the development of artificial intelligence, the application of machine learning to develop predictive models for sepsis-associated acute kidney injury has made potential breakthroughs in early identification, grading, diagnosis, and prognosis determination. Here, we conducted a systematic search of the PubMed, Cochrane Library, Embase (Ovid), Web of Science, and Scopus databases on April 28, 2023, and screened relevant literature. Then, we comprehensively extracted relevant data related to machine learning algorithms, predictors, and predicted objectives. We subsequently performed a critical evaluation of research quality, data aggregation, and analyses. We screened 25 studies on predictive models for sepsis-associated acute kidney injury from a total of originally identified 2898 studies. The most commonly used machine learning algorithm is traditional logistic regression, followed by eXtreme gradient boosting. We categorized these predictive models into early identification models (60%), prognostic prediction models (32%), and subtype identification models (8%) according to their predictive purpose. The five most commonly used predictors were serum creatinine levels, lactate levels, age, blood urea nitrogen concentration, and diabetes mellitus. In addition, a single data source, insufficient assessment of clinical utility, lack of model bias assessment, and hyperparameter adjustment may be the main reasons for the low quality of the current research. However, studies on the nondeath prognostic outcomes, the long-term clinical outcomes, and the subtype identification models are insufficient. Additionally, the poor quality of the research and the insufficient practicality of the model are problems that need to be addressed urgently.