A simplified machine learning model utilizing platelet-related genes for predicting poor prognosis in sepsis.

Background: Thrombocytopenia is a known prognostic factor in sepsis, yet the relationship between platelet-related genes and sepsis outcomes remains elusive. We developed a machine learning (ML) model based on platelet-related genes to predict poor prognosis in sepsis. The model underwent rigorous evaluation on six diverse platforms, ensuring reliable and versatile findings. Methods: A retrospective analysis of platelet data from 365 sepsis patients confirmed the predictive role of platelet count in prognosis. We employed COX analysis, Least Absolute Shrinkage and Selection Operator (LASSO) and Support Vector Machine (SVM) techniques to identify platelet-related genes from the GSE65682 dataset. Subsequently, these genes were trained and validated on six distinct platforms comprising 719 patients, and compared against the Acute Physiology and Chronic Health Evaluation II (APACHE II) and Sequential Organ-Failure Assessment (SOFA) score. Results: A PLT count <100×109/L independently increased the risk of death in sepsis patients (OR = 2.523; 95% CI: 1.084-5.872). The ML model, based on five platelet-related genes, demonstrated impressive area under the curve (AUC) values ranging from 0.5 to 0.795 across various validation platforms. On the GPL6947 platform, our ML model outperformed the APACHE II score with an AUC of 0.795 compared to 0.761. Additionally, by incorporating age, the model's performance was further improved to an AUC of 0.812. On the GPL4133 platform, the initial AUC of the machine learning model based on five platelet-related genes was 0.5. However, after including age, the AUC increased to 0.583. In comparison, the AUC of the APACHE II score was 0.604, and the AUC of the SOFA score was 0.542. Conclusion: Our findings highlight the broad applicability of this ML model, based on platelet-related genes, in facilitating early treatment decisions for sepsis patients with poor outcomes. Our study paves the way for advancements in personalized medicine and improved patient care.

A combined multilevel factor analysis and covariance regression model with mixed effects in the mean and variance structure.

Li et al developed a multilevel covariance regression (MCR) model as an extension of the covariance regression model of Hoff and Niu. This model assumes a hierarchical structure for the mean and the covariance matrix. Here, we propose the combined multilevel factor analysis and covariance regression model in a Bayesian framework, simultaneously modeling the MCR model and a multilevel factor analysis (MFA) model. The proposed model replaces the responses in the MCR part with the factor scores coming from an MFA model. Via a simulation study and the analysis of real data, we show that the proposed model is quite efficient when the responses of the MCR model are not measured directly but are latent variables such as the patient experience measurements in our motivating dataset.

A phase 1 dose-escalation study of the poly(ADP-ribose) polymerase inhibitor senaparib in Australian patients with advanced solid tumors.

BACKGROUND: Senaparib is a novel, selective poly(ADP-ribose) polymerase-1/2 inhibitor with strong antitumor activity in preclinical studies. This first-in-human, phase 1, dose-escalation study examined the safety and preliminary efficacy of senaparib in patients with advanced solid tumors. METHODS: Patients with advanced solid tumors were enrolled from three centers in Australia, using a conventional 3 + 3 design. Dose-escalation cohorts continued until the maximum tolerated dose or a recommended phase 2 dose was determined. Patients received one dose of oral senaparib and, if no dose-limiting toxicity occurred within 7 days, they received senaparib once daily in 3-week cycles. The primary end points were safety and tolerability. RESULTS: Thirty-nine patients were enrolled at 10 dose levels ranging from 2 to 150 mg. No dose-limiting toxicities were observed in any cohort. Most treatment-emergent adverse events were grade 1-2 (91%). Seven patients (17.9%) reported hematologic treatment-emergent adverse events. Treatment-related adverse events occurred in eight patients (20.5%), and the most frequent was nausea (7.7%). Two deaths were reported after the end of study treatment, one of which was considered a complication from senaparib-related bone marrow failure. Pharmacokinetic analysis indicated that senaparib the accumulation index was 1.06-1.67, and absorption saturation was 80-150 mg daily. In 22 patients with evaluable disease, the overall response rate was 13.6%, and the disease control rate was 81.8%. The overall response rate was 33.3% for the BRCA mutation-positive subgroup and 6.3% for the nonmutated subgroup. CONCLUSIONS: Senaparib was well tolerated in Australian patients with advanced solid tumors, with encouraging signals of antitumor activity. The recommended phase 2 dose for senaparib was determined to be 100 mg daily. GOV ID: NCT03507543.

Strain-tuning of edge magnetism in zigzag graphene nanoribbons.

Using the determinant quantum Monte-Carlo method, we elucidate the strain tuning of edge magnetism in zigzag graphene nanoribbons. Our intensive numerical results show that a relatively weak Coulomb interaction may induce a ferromagnetic-like behaviour with a proper strain, and the edge magnetism can be enhanced greatly as the strain along the zigzag edge increases, which provides another way to control graphene magnetism even at room temperature.

Phase I Dose-Escalation Study of Ramucirumab in Chinese Patients with Advanced Solid Tumors.

LESSONS LEARNED: Ramucirumab was well tolerated in Chinese patients with advanced solid tumors, and adverse events were manageable in this study.Pharmacokinetics characteristics in Chinese patients were similar to those in other populations. Immunogenicity was not detected.No efficacy conclusion could be drawn, and further randomized studies are warranted. BACKGROUND: This single-arm, nonrandomized, open-label, dose-escalation, phase I study was designed to evaluate the safety, tolerability, and pharmacokinetics (PK) of ramucirumab in Chinese patients with advanced solid tumors that were resistant to standard therapy or no standard therapy was available. METHODS: Dose escalation was a 3 + 3 design, with expansion in Cohorts 2 and 3 for PK. Ramucirumab was given intravenously at three different dosages: 6 mg/kg every 2 weeks, 10 mg/kg every 3 weeks, and 8 mg/kg every 2 weeks. Safety analyses included all patients. PK, immunogenicity, and antitumor activity were also assessed. RESULTS: Among 28 patients treated, 2 experienced dose-limiting toxicity, possibly related to ramucirumab. No maximum tolerated dose was determined. All patients experienced at least one treatment-emergent adverse event. Grade ≥3 adverse event was reported for 53.6% (n = 15) of patients. PK analyses indicated that ramucirumab had low clearance, small volume of distribution, and long half-life in Chinese patients, as in other populations. Immunogenicity was not detected. No patient had complete/partial response, and 64.3% (n = 18) had stable disease with a median duration of 5.55 months (95% confidence interval: 3.38-7.13 months). CONCLUSION: Ramucirumab appeared to be well tolerated in Chinese patients with advanced solid tumors. PK characteristics in Chinese patients were similar to those in other populations.

Bayesian Multilevel MIMIC Modeling for Studying Measurement Invariance in Cross-group Comparisons.

BACKGROUND: Recent methodological advancements should catalyze the evaluation of measurement invariance across groups, which is required for conducting meaningful cross-group comparisons. OBJECTIVE: The aim of this study was to apply a state-of-the-art statistical method for comparing latent mean scores and evaluating measurement invariance across managers' and frontline workers' ratings of the organization of hospital care. METHODS: On the 87 nursing units in a single institution, French-speaking and Dutch-speaking nursing unit managers' and staff nurses' ratings of their work environment were measured using the multidimensional 32-item practice environment scale of the nursing work index (PES-NWI). Measurement invariance and latent mean scores were evaluated in the form of a Bayesian 2-level multiple indicators multiple causes model with covariates at the individual nurse and nursing unit level. Role (manager, staff nurse) and language (French, Dutch) are of primary interest. RESULTS: Language group membership accounted for 7 of 11 PES-NWI items showing measurement noninvariance. Cross-group comparisons also showed that covariates at both within-level and between-level had significant effects on PES-NWI latent mean scores. Most notably, nursing unit managers, when compared with staff nurses, hold more positive views of several PES-NWI dimensions. CONCLUSIONS: Using a widely used instrument for measuring nurses' work environment, this study shows that precautions for the potential threat of measurement noninvariance are necessary in all stages of a study that relies on survey data to compare groups, particularly in multilingual settings. A Bayesian multilevel multiple indicators multiple causes approach can accommodate for detecting all possible instances of noninvariance for multiple covariates of interest at the within-level and between-level jointly.

Organization of Hospital Nursing, Provision of Nursing Care, and Patient Experiences With Care in Europe.

This study integrates previously isolated findings of nursing outcomes research into an explanatory framework in which care left undone and nurse education levels are of key importance. A moderated mediation analysis of survey data from 11,549 patients and 10,733 nurses in 217 hospitals in eight European countries shows that patient care experience is better in hospitals with better nurse staffing and a more favorable work environment in which less clinical care is left undone. Clinical care left undone is a mediator in this relationship. Clinical care is left undone less frequently in hospitals with better nurse staffing and more favorable nurse work environments, and in which nurses work less overtime and are more experienced. Higher proportions of nurses with a bachelor's degree reduce the effect of worse nurse staffing on more clinical care left undone.

The Relationship between Visual Field Global Indices and Retinal Nerve Fiber Layer Thickness in Healthy Myopes.

The aim of the current study was to investigate the association between the thickness of the retinal nerve fiber layer (RNFL) and central visual field indices in otherwise healthy myopes. In total, 57 otherwise healthy subjects were cross-sectionally studied. General ophthalmic examinations, refractive measurements, RNFL thickness by spectral domain optical coherence tomography (OCT), and central visual fields were examined. Linear models were used to assess the associations. In this young and mid-aged population, the mean spherical equivalent was -4.79 (SD 1.66) and -4.59 (SD 1.88) diopters in the right and left eyes, respectively. Approximately 7% to 14% of the eyes showed the average RNFL thickness out of the normal range. The temporal RNFL was remarkably thicker, whereas the nasal RNFL was thinner. The higher the refractive error, the thinner the RNFL thickness. A thicker overall RNFL was significantly associated with decreased mean sensitivity and increased mean defect, and further adjustments for age, sex, refractive error, optic disk area, or ocular magnification did not change the association. Although nonpathologic myopia does not significantly affect central visual field global indices, its effects on the RNFL may be linked with performance on the central visual field test.

A multivariate multilevel Gaussian model with a mixed effects structure in the mean and covariance part.

A traditional Gaussian hierarchical model assumes a nested multilevel structure for the mean and a constant variance at each level. We propose a Bayesian multivariate multilevel factor model that assumes a multilevel structure for both the mean and the covariance matrix. That is, in addition to a multilevel structure for the mean we also assume that the covariance matrix depends on covariates and random effects. This allows to explore whether the covariance structure depends on the values of the higher levels and as such models heterogeneity in the variances and correlation structure of the multivariate outcome across the higher level values. The approach is applied to the three-dimensional vector of burnout measurements collected on nurses in a large European study to answer the research question whether the covariance matrix of the outcomes depends on recorded system-level features in the organization of nursing care, but also on not-recorded factors that vary with countries, hospitals, and nursing units. Simulations illustrate the performance of our modeling approach.

Levels and correlates of implicit rationing of nursing care in Swiss acute care hospitals--a cross sectional study.

OBJECTIVES: (1) To describe the levels of implicit rationing of nursing care in Swiss acute care hospitals; (2) to explore the associations between nine selected potential rationing predictors and implicit rationing of nursing care. DESIGN: Cross sectional multi-center study. SETTINGS: A quota sample of 35 acute care hospitals from the German, French and Italian speaking regions of Switzerland participating in RN4CAST (Registered Nurse Forecasting) study. PARTICIPANTS: 1633 registered nurses working in randomly selected medical, surgical or medical-surgical units. METHODS: Implicit rationing of nursing care, i.e., the withholding of any of 32 necessary nursing measures due to a lack of nursing resources, was measured using the revised Basel Extent of Rationing of Nursing Care (BERNCA) instrument. Nine potential rationing predictors, e.g., staffing and resource adequacy, patient-to-nurse ratio, nurse education, and confounding variables, e.g., nurse gender and age, hospital typology, were assessed with validated instruments or single items of the RN4CAST study. Descriptive statistical procedures were used as appropriate. Three level regression models were used to investigate the effect of the selected nine predictors on rationing at the nurse, unit and hospital levels. RESULTS: Ninety-eight percent of the participating nurses reported that, in their last seven working days, they had to ration at least one of the 32 nursing tasks listed in the BERNCA. The mean rationing level of 1.69 (SD=0.571) indicates that on average the nurses reported 'rarely' being unable to perform the nursing tasks listed in the BERNCA. Multilevel regression analysis confirmed two of the nine tested predictors: better unit level staff resource adequacy and a more favorable hospital level safety climate were both consistently significantly associated with lower rationing levels. Counter to our assumptions, the other two nurse practice environment dimensions, the three workload measures, nurse experience and nurse education were not associated with rationing. CONCLUSIONS: Rationing frequency varied among the 32 BERNCA items, indicating differing prioritizations of necessary nursing tasks. The identified rationing predictors, staff resource adequacy and safety climate, can determine starting points for interventions, i.e., proactive changes to improve staff resource adequacy when rationing exceeds predefined thresholds, increasing the risk of negative effects on patient outcomes.

A multi-country perspective on nurses' tasks below their skill level: reports from domestically trained nurses and foreign trained nurses from developing countries.

BACKGROUND: Several studies have concluded that the use of nurses' time and energy is often not optimized. Given widespread migration of nurses from developing to developed countries, it is important for human resource planning to know whether nursing education in developing countries is associated with more exaggerated patterns of inefficiency. OBJECTIVES: First, to describe nurses' reports on tasks below their skill level. Second, to examine the association between nurses' migratory status (domestically trained nurse or foreign trained nurse from a developing country) and reports on these tasks. DESIGN: The Registered Nurse Forecasting Study used a cross-sectional quantitative research design to gather data from 33,731 nurses (62% response rate) in 486 hospitals in Belgium, England, Finland, Germany, Greece, Ireland, the Netherlands, Norway, Poland, Spain, Sweden and Switzerland. METHODS: For this analysis, nurse-reported information on migratory status and tasks below their skill level performed during their last shift was used. Random effects models estimated the effect of nurses' migratory status on reports of these tasks. RESULTS: 832 nurses were trained in a developing country (2.5% of total sample). Across countries, a high proportion of both domestically trained and foreign trained nurses from developing countries reported having performed tasks below their skill level during their last shift. After adjusting for nurses' type of last shift worked, years of experience, and level of education, there remained a pronounced overall effect of being a foreign trained nurse from a developing country and an increase in reports of tasks below skill level performed during the last shift. CONCLUSION: The findings suggest that there remains much room for improvement to optimize the use of nurses' time and energy. Special attention should be given to raising the professional level of practice of foreign trained nurses from developing countries. Further research is needed to understand the influence of professional practice standards, skill levels of foreign trained nurses from developing countries and values attached to these tasks resulting from previous work experiences in their home countries. This will allow us to better understand the conditions under which foreign trained nurses from developing countries can optimally contribute to professional nursing practice in developed country contexts.

Group-level impact of work environment dimensions on burnout experiences among nurses: a multivariate multilevel probit model.

BACKGROUND: Nurses' work environments are associated with burnout experiences among nurses. The RN4CAST project provides data on these constructs within a four-level structure (nurse, nursing unit, hospital, and country), implying more complicated multilevel analysis strategies than have been used in previous efforts studying this relationship. OBJECTIVES: First, to explore and investigate the effect of the nursing unit, hospital, and country level variability on the relationship between dimensions of nurses' work environment and dimensions of burnout. Second, to explore the significance of the nursing unit, hospital, and country level variability among the burnout dimensions. DESIGN: Data from the RN4CAST project were available from a cross-sectional survey among 23,446 nurses in 2087 nursing units in 352 hospitals in 11 countries. METHODS: Nurse-reported information on their work environment (managerial support for nursing, doctor-nurse collegial relations, and promotion of care quality) and burnout experiences (emotional exhaustion, depersonalization and personal accomplishment) were available. We specified ecological measures of the nurse work environment dimensions at the three organizational levels and combined these with individual-level outcomes within a series of multilevel statistical models. The final model was a multivariate multilevel probit model in which we modeled the work environment and burnout dimensions jointly. RESULTS: Doctor-nurse collegial relations affected all burnout dimensions, but at the unit level only. For the dimension of promotion of care quality, the effect of the ecological exposure on burnout was pronounced at both the nursing unit and the hospital level for all three burnout dimensions. Findings for the dimensions of managerial support for nursing were ambiguous. CONCLUSION: Nurse work environment dynamics are related to nurses' burnout experiences at both the nursing unit and the hospital level. This implies that both hospital-wide and unit-specific interventions should be considered to achieve excellent work environments. The correlation structure among the three burnout outcomes varies across countries, but is stable between hospitals within countries and between nursing units within hospitals.

Logistic random effects regression models: a comparison of statistical packages for binary and ordinal outcomes.

BACKGROUND: Logistic random effects models are a popular tool to analyze multilevel also called hierarchical data with a binary or ordinal outcome. Here, we aim to compare different statistical software implementations of these models. METHODS: We used individual patient data from 8509 patients in 231 centers with moderate and severe Traumatic Brain Injury (TBI) enrolled in eight Randomized Controlled Trials (RCTs) and three observational studies. We fitted logistic random effects regression models with the 5-point Glasgow Outcome Scale (GOS) as outcome, both dichotomized as well as ordinal, with center and/or trial as random effects, and as covariates age, motor score, pupil reactivity or trial. We then compared the implementations of frequentist and Bayesian methods to estimate the fixed and random effects. Frequentist approaches included R (lme4), Stata (GLLAMM), SAS (GLIMMIX and NLMIXED), MLwiN ([R]IGLS) and MIXOR, Bayesian approaches included WinBUGS, MLwiN (MCMC), R package MCMCglmm and SAS experimental procedure MCMC.Three data sets (the full data set and two sub-datasets) were analysed using basically two logistic random effects models with either one random effect for the center or two random effects for center and trial. For the ordinal outcome in the full data set also a proportional odds model with a random center effect was fitted. RESULTS: The packages gave similar parameter estimates for both the fixed and random effects and for the binary (and ordinal) models for the main study and when based on a relatively large number of level-1 (patient level) data compared to the number of level-2 (hospital level) data. However, when based on relatively sparse data set, i.e. when the numbers of level-1 and level-2 data units were about the same, the frequentist and Bayesian approaches showed somewhat different results. The software implementations differ considerably in flexibility, computation time, and usability. There are also differences in the availability of additional tools for model evaluation, such as diagnostic plots. The experimental SAS (version 9.2) procedure MCMC appeared to be inefficient. CONCLUSIONS: On relatively large data sets, the different software implementations of logistic random effects regression models produced similar results. Thus, for a large data set there seems to be no explicit preference (of course if there is no preference from a philosophical point of view) for either a frequentist or Bayesian approach (if based on vague priors). The choice for a particular implementation may largely depend on the desired flexibility, and the usability of the package. For small data sets the random effects variances are difficult to estimate. In the frequentist approaches the MLE of this variance was often estimated zero with a standard error that is either zero or could not be determined, while for Bayesian methods the estimates could depend on the chosen "non-informative" prior of the variance parameter. The starting value for the variance parameter may be also critical for the convergence of the Markov chain.