Multivariate Generalized Linear Mixed Models With Random Intercepts To Analyze Cardiovascular Risk Markers in Type-1 Diabetic Patients

Analysis of Survival Modification by Furosemide Use in a Cohort of Hospitalized COVID-19 Patients with Severe or Critical Disease in Mexico: Due to Its Chemical Structure, Furosemide Is More than Just a Diuretic

In the ongoing fight against Coronavirus Disease 2019 (COVID-19), researchers are exploring potential treatments to improve outcomes, especially in severe cases. This includes investigating the repurposing of existing medications, such as furosemide, which is widely available. This study aimed to evaluate the impact of furosemide on mortality rates among COVID-19 patients with severe or critical illness. We assessed a cohort of 515 hospitalized adults who experienced a high mortality rate of 43.9%. Using a multivariate analysis with adjusted risk ratios (AdRRs), factors like smoking (AdRR 2.48, 95% CI 1.53-4.01, p < 0.001), a high Pneumonia Severity Index (PSI) score (AdRR 7.89, 95% CI 5.82-10.70, p < 0.001), mechanical ventilation (AdRR 23.12, 95% CI 17.28-30.92, p < 0.001), neutrophilia (AdRR 2.12, 95% CI 1.52-2.95, p < 0.001), and an elevated neutrophil-to-lymphocyte ratio (NLR) (AdRR 2.39, 95% CI 1.72-3.32, p < 0.001) were found to increase mortality risk. In contrast, vaccination and furosemide use were associated with reduced mortality risk (AdRR 0.58, p = 0.001 and 0.60, p = 0.008; respectively). Furosemide showed a pronounced survival benefit in patients with less severe disease (PSI < 120) and those not on hemodialysis, with mortality rates significantly lower in furosemide users (3.7% vs. 25.7%). A Kaplan-Meier analysis confirmed longer survival and better oxygenation levels in patients treated with furosemide. Furthermore, a Structure-Activity Relationship analysis revealed that furosemide's sulfonamide groups may interact with cytokine sites such as tumor necrosis factor-alpha (TNF-α) and interleukin-6 (IL-6), potentially explaining its beneficial effects in COVID-19 management. These findings suggest that furosemide could be a beneficial treatment option in certain COVID-19 patient groups, enhancing survival and improving oxygenation.

Differences in the Evolution of Clinical, Biochemical, and Hematological Indicators in Hospitalized Patients with COVID-19 According to Their Vaccination Scheme: A Cohort Study in One of the World's Highest Hospital Mortality Populations

COVID-19 vaccines primarily prevent severe illnesses or hospitalization, but there is limited data on their impact during hospitalization for seriously ill patients. In a Mexican cohort with high COVID-19 mortality, a study assessed vaccination's effects. From 2021 to 2022, 462 patients with 4455 hospital days were analyzed. The generalized multivariate linear mixed model (GENLINMIXED) with binary logistic regression link, survival analysis and ROC curves were used to identify risk factors for death. The results showed that the vaccinated individuals were almost half as likely to die (adRR = 0.54, 95% CI = 0.30-0.97, p = 0.041). When stratifying by vaccine, the Pfizer group (BNT162b2) had a 2.4-times lower risk of death (adRR = 0.41, 95% CI = 0.2-0.8, p = 0.008), while the AstraZeneca group (ChAdOx1-S) group did not significantly differ from the non-vaccinated (adRR = 1.04, 95% CI = 0.5-2.3, p = 0.915). The Pfizer group exhibited a higher survival, the unvaccinated showed increasing mortality, and the AstraZeneca group remained intermediate (p = 0.003, multigroup log-rank test). Additionally, BNT162b2-vaccinated individuals had lower values for markers, such as ferritin and D-dimer. Biochemical and hematological indicators suggested a protective effect of both types of vaccines, possibly linked to higher lymphocyte counts and lower platelet-to-lymphocyte ratio (PLR). It is imperative to highlight that these results reinforce the efficacy of COVID-19 vaccines. However, further studies are warranted for a comprehensive understanding of these findings.

Interactions between the principal risk factors for reduction of the eGFR in unvaccinated COVID‑19 survivors: Normal pre-COVID‑19 eGFR, not having diabetes and being hospitalized

There are contradictory results regarding changes in estimated glomerular filtration rate (eGFR) in coronavirus disease 2019 (COVID-19) survivors. An analysis of eGFR changes and clinical characteristics associated with those changes was conducted among COVID-19 survivors. eGFR values were compared at different time points (before and 4-, 8- and 12-months after COVID-19 infection). A multivariate generalized linear mixed model (GENLINMIXED procedure) with a binary logistic regression link was used to determine factors associated with eGFR reduction of ≥10 ml/min/1.73 m2. Being hospitalized (RR=2.90, 95% CI=1.10-7.68, P=0.032), treated with Ivermectin (RR=14.02, 95% CI=4.11-47.80, P<0.001) or anticoagulants (RR=6.51, 95% CI=2.69-15.73, P<0.001) are risk factors for a reduced eGFR. Having a low eGFR (<90 ml/min/1.73 m2) before COVID-19 infection, having B-positive blood type, diabetes, taking vitamin C during the acute phase of COVID-19 or suffering from chronic COVID-19 symptoms, were identified as protective factors. Analysis involving a two-way interaction (A x B, where A and B are factors) demonstrated that the combination of patients with a normal eGFR value before COVID-19 infection without diabetes (RR=58.60, 95% CI=11.62-295.38, P<0.001), or a normal eGFR value with being hospitalized for COVID-19 (RR=38.07, 95% CI=8.68-167.00, P<0.001), increased the probability of a reduced eGFR. The changes in eGFR in COVID-19 survivors varied depending on patient characteristics. Furthermore, the principal risk factors for post-COVID-19 eGFR reduction were analyzed in separate models.

Multivariate generalized linear mixed models for continuous bounded outcomes: Analyzing the body fat percentage data

We propose a multivariate regression model to handle multiple continuous bounded outcomes. We adopted the maximum likelihood approach for parameter estimation and inference. The model is specified by the product of univariate probability distributions and the correlation between the response variables is obtained through the correlation matrix of the random intercepts. For modeling continuous bounded variables on the interval (0,1) we considered the beta and unit gamma distributions. The main advantage of the proposed model is that we can easily combine different marginal distributions for the response variable vector. The computational implementation is performed using Template Model Builder, which combines the Laplace approximation with automatic differentiation. Therefore, the proposed approach allows us to estimate the model parameters quickly and efficiently. We conducted a simulation study to evaluate the computational implementation and the properties of the maximum likelihood estimators under different scenarios. Moreover, we investigate the impact of distribution misspecification in the proposed model. Our model was motivated by a data set with multiple continuous bounded outcomes, which refer to the body fat percentage measured at five regions of the body. Simulation studies and data analysis showed that the proposed model provides a general and rich framework to deal with multiple continuous bounded outcomes.

Advanced models for improved prediction of opioid-related overdose and suicide events among Veterans using administrative healthcare data

Veterans suffer disproportionate health impacts from the opioid epidemic, including overdose, suicide, and death. Prediction models based on electronic medical record data can be powerful tools for identifying patients at greatest risk of such outcomes. The Veterans Health Administration implemented the Stratification Tool for Opioid Risk Mitigation (STORM) in 2018. In this study we propose changes to the original STORM model and propose alternative models that improve risk prediction performance. The best of these proposed models uses a multivariate generalized linear mixed modeling (mGLMM) approach to produce separate predictions for overdose and suicide-related events (SRE) rather than a single prediction for combined outcomes. Further improvements include incorporation of additional data sources and new predictor variables in a longitudinal setting. Compared to a modified version of the STORM model with the same outcome, predictor and interaction terms, our proposed model has a significantly better prediction performance in terms of AUC (84% vs. 77%) and sensitivity (71% vs. 66%). The mGLMM performed particularly well in identifying patients at risk for SREs, where 72% of actual events were accurately predicted among patients with the 100,000 highest risk scores compared with 49.7% for the modified STORM model. The mGLMM’s strong performance in identifying true cases (sensitivity) among this highest risk group was the most important improvement given the model’s primary purpose for accurately identifying patients at most risk for adverse outcomes such that they are prioritized to receive risk mitigation interventions. Some predictors in the proposed model have markedly different associations with overdose and suicide risks, which will allow clinicians to better target interventions to the most relevant risks.

Random effect models for multivariate mixed data: A Parafac-based finite mixture approach

We discuss a flexible regression model for multivariate mixed responses. Dependence between outcomes is introduced via the joint distribution of discrete outcome- and individual-specific random effects that represent potential unobserved heterogeneity in each outcome profile. A different number of locations can be used for each margin, and the association structure is described by a tensor that can be further simplified by using the Parafac model. A case study illustrates the proposal.

Academic detailing increases prescription drug monitoring program use among primary care practices

OBJECTIVE

Clinical review of a prescription drug monitoring program (PDMP) is considered a valuable tool for opioid prescribing risk mitigation; however, PDMP use is often low, even in states with mandatory registration and use policies. The objective was to evaluate the impact of an academic detailing (AD) outreach intervention on PDMP use among primary care prescribers.

METHODS

AD intervention was delivered to primary care based controlled substance prescribers (N = 87) and their associated PDMP delegates (n = 42) by a clinical pharmacist as 1 component of a large-scale, statewide initiative to improve opioid prescribing safety. Prescriber PDMP use behavior was assessed by prescriber self-report and analysis of objective 2016-2018 PDMP data regarding the number of monthly report requests. We compared means between pre- and postintervention using a paired t test and plotted the monthly average reports over time to assess the trend of mean reports over time. Generalized linear mixed model with a negative binomial distribution was used to assess the difference in the trend and magnitude of the combined count of reports for the entire sample and prescriber subsets that were segmented on the basis of the adoption status of PDMP.

RESULTS

The monthly mean of reports by combined prescribers and delegates significantly increased after the AD intervention (mean 28.1 pre vs. 53.0 post; P < 0.001), with the increase in delegate reports (mean 17.1 pre vs. 60.0 post; P < 0.001) driving the overall increase. Reports were requested 40.4 times more often than in the preintervention period (P < 0.001). Patterns of pre- to postchanges in mean monthly report requests differed by baseline PDMP adoption status.

CONCLUSION

The AD intervention was transformative in facilitating practice change to use delegates to run reports. Visits with both prescribers and delegates, including hands-on PDMP training and registration assistance, can be viewed as beneficial for practice facilitation.

Factors Influencing the Regional Dynamics of Butternut Canker

Butternut (Juglans cinerea) is an important component of native biodiversity in eastern North America. Of urgent concern is the survival of butternut, whose populations are declining rapidly, in large part due to an exotic pathogen, Ophiognomonia clavigignenti-juglandacearum, that causes butternut canker. The disease presently occurs throughout the range of butternut in North America, causing branch and stem cankers, dieback, and tree mortality. Despite the existential threat posed by O. clavigignenti-juglandacearum to butternut, a detailed understanding of the factors that drive cross-scale disease patterns is lacking. Therefore, we investigated the association of a range of factors, including tree attributes, topography, and weather, with butternut canker spatial dynamics at different scales using data collected in the province of Quebec, Canada. Trunk canker damage and dieback showed distinct geographic patterns. Bark phenotype was not significantly associated with trunk canker damage. Results suggest that open or dominant trees may show less dieback than intermediate or suppressed trees. Probability of the presence of trunk canker and percent dieback were proportional to the tree diameter at breast height. Temperature was positively associated with disease severity at a 1-km² scale. Our results provide strong evidence that multiple factors, notably weather, influence butternut canker epidemiology.