Search

Introduction: Accurately predicting patient outcomes is crucial for improving healthcare delivery, but large-scale risk prediction models are often developed and tested on specific datasets where clinical parameters and outcomes may not fully reflect local clinical settings. Where this is the case, whether to opt for de-novo training of prediction models on local datasets, direct porting of externally trained models, or a transfer learning approach is not well studied, and constitutes the focus of this study. Using the clinical challenge of predicting mortality and hospital length of stay on a Danish trauma dataset, we hypothesized that a transfer learning approach of models trained on large external datasets would provide optimal prediction results compared to de-novo training on sparse but local datasets or directly porting externally trained models., Methods: Using an external dataset of trauma patients from the US Trauma Quality Improvement Program (TQIP) and a local dataset aggregated from the Danish Trauma Database (DTD) enriched with Electronic Health Record data, we tested a range of model-level approaches focused on predicting trauma mortality and hospital length of stay on DTD data. Modeling approaches included de-novo training of models on DTD data, direct porting of models trained on TQIP data to the DTD, and a transfer learning approach by training a model on TQIP data with subsequent transfer and retraining on DTD data. Furthermore, data-level approaches, including mixed dataset training and methods countering imbalanced outcomes (e.g., low mortality rates), were also tested., Results: Using a neural network trained on a mixed dataset consisting of a subset of TQIP and DTD, with class weighting and transfer learning (retraining on DTD), we achieved excellent results in predicting mortality, with a ROC-AUC of 0.988 and an F2-score of 0.866. The best-performing models for predicting long-term hospitalization were trained only on local data, achieving an ROC-AUC of 0.890 and an F1-score of 0.897, although only marginally better than alternative approaches., Conclusion: Our results suggest that when assessing the optimal modeling approach, it is important to have domain knowledge of how incidence rates and workflows compare between hospital systems and datasets where models are trained. Including data from other health-care systems is particularly beneficial when outcomes are suffering from class imbalance and low incidence. Scenarios where outcomes are not directly comparable are best addressed through either de-novo local training or a transfer learning approach., Competing Interests: Authors MS and AB are employed by the company Aiomic. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest., (© 2023 Millarch, Bonde, Bonde, Klein, Folke, Rudolph and Sillesen.)

Virulence factors enable pathogenic bacteria to infect host cells, establish infection, and contribute to disease progressions. In Gram-positive pathogens such as Staphylococcus aureus (Sa) and Enterococcus faecalis (Ef), the pleiotropic transcription factor CodY plays a key role in integrating metabolism and virulence factor expression. However, to date, the structural mechanisms of CodY activation and DNA recognition are not understood. Here, we report the crystal structures of CodY from Sa and Ef in their ligand-free form and their ligand-bound form complexed with DNA. Binding of the ligands-branched chain amino acids and GTP-induces conformational changes in the form of helical shifts that propagate to the homodimer interface and reorient the linker helices and DNA binding domains. DNA binding is mediated by a non-canonical recognition mechanism dictated by DNA shape readout. Furthermore, two CodY dimers bind to two overlapping binding sites in a highly cooperative manner facilitated by cross-dimer interactions and minor groove deformation. Our structural and biochemical data explain how CodY can bind a wide range of substrates, a hallmark of many pleiotropic transcription factors. These data contribute to a better understanding of the mechanisms underlying virulence activation in important human pathogens., (© The Author(s) 2023. Published by Oxford University Press on behalf of Nucleic Acids Research.)

Listeria monocytogenes is a facultative Gram-positive bacterium that causes listeriosis, a severe foodborne disease. We previously discovered that ring-fused 2-pyridone compounds can decrease virulence factor expression in Listeria by binding and inactivating the PrfA virulence activator. In this study, we tested PS900, a highly substituted 2-pyridone that was recently discovered to be bactericidal to other Gram-positive pathogenic bacteria, such as Staphylococcus aureus and Enterococcus faecalis. We show that PS900 can interact with PrfA and reduce the expression of virulence factors. Unlike previous ring-fused 2-pyridones shown to inactivate PrfA, PS900 had an additional antibacterial activity and was found to potentiate sensitivity toward cholic acid. Two PS900-tolerant mutants able to grow in the presence of PS900 carried mutations in the brtA gene, encoding the BrtA repressor. In wild-type (WT) bacteria, cholic acid binds and inactivates BrtA, thereby alleviating the expression of the multidrug transporter MdrT. Interestingly, we found that PS900 also binds to BrtA and that this interaction causes BrtA to dissociate from its binding site in front of the mdrT gene. In addition, we observed that PS900 potentiated the effect of different osmolytes. We suggest that the increased potency of cholic acid and osmolytes to kill bacteria in the presence of PS900 is due to the ability of the latter to inhibit general efflux, through a yet-unknown mechanism. Our data indicate that thiazolino 2-pyridones constitute an attractive scaffold when designing new types of antibacterial agents. IMPORTANCE Bacteria resistant to one or several antibiotics are a very large problem, threatening not only treatment of infections but also surgery and cancer treatments. Thus, new types of antibacterial drugs are desperately needed. In this work, we show that a new generation of substituted ring-fused 2-pyridones not only inhibit Listeria monocytogenes virulence gene expression, presumably by inactivating the PrfA virulence regulator, but also potentiate the bactericidal effects of cholic acid and different osmolytes. We identified a multidrug repressor as a second target of 2-pyridones. The repressor-2-pyridone interaction displaces the repressor from DNA, thus increasing the expression of a multidrug transporter. In addition, our data suggest that the new class of ring-fused 2-pyridones are efficient efflux inhibitors, possibly explaining why the simultaneous addition of 2-pyridones together with cholic acid or osmolytes is detrimental for the bacterium. This work proves conclusively that 2-pyridones constitute a promising scaffold to build on for future antibacterial drug design., Competing Interests: The authors declare a conflict of interest. F.A. and J.J. have ownership interests in Quretech Bio AB. Rest of the authors have no competing interest.

The risks of post trauma complications are regulated by the injury, comorbidities, and the clinical trajectories, yet prediction models are often limited to single time-point data. We hypothesize that deep learning prediction models can be used for risk prediction using additive data after trauma using a sliding windows approach. Using the American College of Surgeons Trauma Quality Improvement Program (ACS TQIP) database, we developed three deep neural network models, for sliding-windows risk prediction. Output variables included early- and late mortality and any of 17 complications. As patients moved through the treatment trajectories, performance metrics increased. Models predicted early- and late mortality with ROC AUCs ranging from 0.980 to 0.994 and 0.910 to 0.972, respectively. For the remaining 17 complications, the mean performance ranged from 0.829 to 0.912. In summary, the deep neural networks achieved excellent performance in the sliding windows risk stratification of trauma patients., (© 2023. The Author(s).)

The alarming rise of multidrug-resistant Gram-positive bacteria has precipitated a healthcare crisis, necessitating the development of new antimicrobial therapies. Here we describe a new class of antibiotics based on a ring-fused 2-pyridone backbone, which are active against vancomycin-resistant enterococci (VRE), a serious threat as classified by the Centers for Disease Control and Prevention, and other multidrug-resistant Gram-positive bacteria. Ring-fused 2-pyridone antibiotics have bacteriostatic activity against actively dividing exponential phase enterococcal cells and bactericidal activity against nondividing stationary phase enterococcal cells. The molecular mechanism of drug-induced killing of stationary phase cells mimics aspects of fratricide observed in enterococcal biofilms, where both are mediated by the Atn autolysin and the GelE protease. In addition, combinations of sublethal concentrations of ring-fused 2-pyridones and standard-of-care antibiotics, such as vancomycin, were found to synergize to kill clinical strains of VRE. Furthermore, a broad range of antibiotic resistant Gram-positive pathogens, including those responsible for the increasing incidence of antibiotic resistant healthcare-associated infections, are susceptible to this new class of 2-pyridone antibiotics. Given the broad antibacterial activities of ring-fused 2-pyridone compounds against Gram-positive (GmP) bacteria we term these compounds GmPcides, which hold promise in combating the rising tide of antibiotic resistant Gram-positive pathogens.

MRSA-1369 is a uropathogenic methicillin-resistant Staphylococcus aureus (MRSA) strain. Here, we present the complete genome sequence of MRSA-1369, which consists of one chromosome (2.87 Mb) and two plasmids (16.68 kb and 3.13 kb). This will serve as a reference genome for future Staphylococcus aureus pathogenesis and multiomic studies.

P66, a bifunctional integral outer membrane protein, is necessary for Borrelia burgdorferi to establish initial infection and to disseminate in mice. The integrin binding function of P66 facilitates extravasation and dissemination, but the role of its porin function during murine infection has not been investigated. A limitation to studying P66 porin function during mammalian infection has been the lack of structural information for P66. In this study, we experimentally characterized specific domains of P66 with regard to structure and function. First, we aligned the amino acid sequences of P66 from Lyme disease-causing Borrelia and relapsing fever-causing Borrelia to identify conserved and unique domains between these disease-causing clades. Then, we examined whether specific domains of P66 are exposed on the surface of the bacteria by introducing c-Myc epitope tags into each domain of interest. The c-Myc epitope tag inserted C-terminally to E33 (highly conserved domain), to T187 (integrin binding region domain and a non-conserved domain), and to E334 (non-conserved domain) were all detected on the surface of Borrelia burgdorferi . The c-Myc epitope tag inserted C-terminally to E33 and D303 in conserved domains disrupted P66 oligomerization and porin function. In a murine model of infection, the E33 and D303 mutants exhibited decreased infectivity and dissemination. Taken together, these results suggest the importance of these conserved domains, and potentially P66 porin function, in vivo ., Competing Interests: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest., (Copyright © 2022 Curtis, Fierros, Hahn, Surdel, Kessler, Anderson, Vandewalle-Capo, Bonde, Zhu, Bergström and Coburn.)

Background: Patient-reported outcomes (PROs) are increasingly recognized as valuable sources of information to enhance our understanding of the quality of healthcare from the patient's perspective., Objective: This study aimed to describe the implementation process of the Danish nationwide PRO-Psychiatry project, including iterative tests of previously developed PRO measurement concept and an online data collection tool. Additional aims were to identify the 'best practice' for the routine use of PROs in hospital-based psychiatry and design information material about the project., Methods: We conducted an action-oriented observational study to explore the pilot implementation of the PRO-Psychiatry project, which was initiated in February 2018. The study was based on an iterative plan-do-learn approach. An inpatient unit and an outpatient unit from the same psychiatric department in the North Denmark Region were selected for the pilot implementation. The implementation was anchored in multidisciplinary implementation teams at unit level. These teams managed the implementation process according to four tasks defined by the department management., Results: The teams designed, tested, evaluated and adjusted the localized work practices relating to the use of PRO-Psychiatry. The comprehensibility of the predesigned PROs, the usability of the Information Technology(IT) system and the routine use of PROs during clinical consultations were repeatedly tested and adjusted until the functionality was satisfactory. Furthermore, the teams designed information material for patients (emails, posters, handouts and webpages) and clinicians (online clinical guidelines). The team members informed their colleagues about the progress of PRO-Psychiatry at staff meetings and rolled out the initiative through one-to-one teaching., Conclusions: The pilot implementation was deemed successful. PRO-Psychiatry was rolled out to other units in the region, and a national decision was made to pilot implement the initiative in the other four Danish regions., (© The Author(s) 2021. Published by Oxford University Press on behalf of International Society for Quality in Health Care. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.)

Background: Reporting of barriers and successes associated with the implementation and use of patient-reported outcomes (PROs) is limited as a means to ensure enhanced patient involvement, shared decision-making and improved treatment and care. We set out to evaluate the implementation and use of the PRO-Psychiatry initiative on patient-reported outcome measures in Danish mental health care. We aimed to described four specific areas: the quality of the clinical consultations before and after the implementation of PRO-Psychiatry as perceived by the patients (objective A), the motivation for participating in PRO-Psychiatry as perceived by patients and clinicians (objective B), the implementation process as perceived by patients, clinicians and managers (objective C) and suggestions for improvement (objective D)., Methods: The PRO-Psychiatry initiative was evaluated through a participatory approach, including patients, clinicians and managers. A repeated cross-sectional interview-based survey explored the quality of the clinical consultation before and after the implementation of PRO-Psychiatry. A three-step semi-structured group interview, inspired by the modified mini-Delphi method, was used to establish consensus on the evaluation of the implementation and use of the initiative., Results: The evaluation pointed at PRO-Psychiatry as a meaningful initiative, which motivated patients and supported clinicians. The patients emphasised the importance of PROs, but they also found that PROs were not used enough. Clinically relevant improvements were detected after the implementation of the initiative; more patients felt heard and experienced that clinicians took a greater interest in their problems. The clinicians valued the easily accessible real-time graphical display of the PRO responses in the electronic health record (EHR). Clinicians and managers agreed that clinical PRO practices, patient compliance and use of PROs in treatment and care should be supported during implementation., Conclusion: The evaluation was overall positive. Patients and clinicians were willing to participate, found the online reporting easy and valued the direct access to PRO responses in the EHR. An essential feature was the integration of well-defined and functional PRO practices into the existing clinical workflow. Using PROs in the clinical sessions in a way that was palpable to the patient was found to be a significant improvement need. At the individual level, PRO-Psychiatry can use patient outcome information to support dialogue, encourage shared decision-making and promote self-management during recovery. At the aggregated patient level, the PROs can be used for monitoring the patient-perceived quality of care and for research., (© The Author(s) 2022. Published by Oxford University Press on behalf of International Society for Quality in Health Care. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.)