Beta-Blocker Treatment in Ventilated COVID-19 patients – A Cox Regression with Time Dependent Covariate Analysis

Background: Patients with co-morbidities are particularly vulnerable to severe COVID-19 disease. Critically ill patients with COVID-19 frequently experience severe tachycardias and avoidance of these is important in some co-morbidities, for instance cardiovascular disease. There is growing interest in beta blockade in critical illness as their use been associated with improved outcomes in a variety of conditions. We report the real-world use of heart rate management in patients during the first wave of the COVID-19 pandemic. As retrospective data are prone to an Immortal Time Bias, we created a Cohort Trial such as might be used for a future prospective trial and used Time Dependent Covariate Analysis for its analysis. Methods: Data for all PCR-proven COVID-19 patients ventilated in the Intensive Care Unit (ICU) were extracted from the hospital databases. To compensate for the risk of immortal time bias, we restricted analysis to 144 patients who achieved a heart rate (HR) of 90 beats per minute for more than 12 hours and were treated with norepinephrine. We recorded time from these ‘entry criteria’ to first beta blocker dose. Those patients who did not receive a beta blocker were given a nominal time to beta blocker beyond the censor day. Outcome was mortality censored at 28 days.Results: In the study group, 83/144 patients (57.6%) received a beta blocker. The median interval from entry criteria to beta blocker was 7.91 days (IQR 3.89, 13.15) and median duration of treatment was 7.00 days (IQR 4.00, 14.00). Twenty-four beta blocker patients (28.9%) died within 28 days compared with 29 (47.5%) who did not (adjusted OR 0.43; 95% CI 0.20-0.95, P=0.036). Cox Regression with time-dependent covariate analysis revealed there was an increased, but not significant, risk of death with beta blocker delay (Hazard Ratio 1.42 p=0.264). Mortality was also reduced for each day treated with beta blockade (adjusted Odds Ratio 0.76, 95% CI 0.64-0.91; P=0.002).Conclusions: In a retrospective analysis of critically ill ventilated patients with COVID-19 who developed a tachycardia >90 beats per minute and were treated with norepinephrine, beta blockade was associated with reduced mortality.

Analysis of Youden Square Design with Two Missing Observations Belonging To the Same Treatment

Two missing observations can occur in a Youden Square Design in eight mutually exclusive ways. In the present study, the author has tried to discuss the case of two missing observations belonging to the same treatment. Estimates of the missing observations and variances of the various elementary treatment contrasts have been obtained by using Bartlett’s covariate analysis.

Efficient algorithms for covariate analysis with dynamic data using nonlinear mixed-effects model

Nonlinear mixed-effects modeling is one of the most popular tools for analyzing repeated measurement data, particularly for applications in the biomedical fields. Multiple integration and nonlinear optimization are the two major challenges for likelihood-based methods in nonlinear mixed-effects modeling. To solve these problems, approaches based on empirical Bayesian estimates have been proposed by breaking the problem into a nonlinear mixed-effects model with no covariates and a linear regression model without random effect. This approach is time-efficient as it involves no covariates in the nonlinear optimization. However, covariate effects based on empirical Bayesian estimates are underestimated and the bias depends on the extent of shrinkage. Marginal correction method has been proposed to correct the bias caused by shrinkage to some extent. However, the marginal approach appears to be suboptimal when testing covariate effects on multiple model parameters, a situation that is often encountered in real-world data analysis. In addition, the marginal approach cannot correct the inaccuracy in the associated p-values. In this paper, we proposed a simultaneous correction method (nSCEBE), which can handle the situation where covariate analysis is performed on multiple model parameters. Simulation studies and real data analysis showed that nSCEBE is accurate and efficient for both effect-size estimation and p-value calculation compared with the existing methods. Importantly, nSCEBE can be >2000 times faster than the standard mixed-effects models, potentially allowing utilization for high-dimension covariate analysis for longitudinal or repeated measured outcomes.

Population pharmacokinetic modeling of PF-06439535 (a bevacizumab biosimilar) and reference bevacizumab (Avastin®) in patients with advanced non-squamous non-small cell lung cancer

Purpose The objectives of this analysis were to characterize the population pharmacokinetics (PK) of PF-06439535 (a bevacizumab biosimilar) and reference bevacizumab (Avastin®) sourced from the European Union (bevacizumab-EU) in patients with advanced non-squamous non-small cell lung cancer (NSCLC), and to quantify the difference in PK parameters between the two drug products via covariate analysis. Methods Pooled PF-06439535 and bevacizumab-EU serum concentration data from a comparative clinical efficacy and safety study (NCT02364999) in patients with NSCLC (N = 719) were analyzed using a non-linear mixed-effects modeling approach. Patients received PF-06439535 plus chemotherapy or bevacizumab-EU plus chemotherapy every 21 days for 4–6 cycles, followed by monotherapy with PF-06439535 or bevacizumab-EU. PF-06439535 or bevacizumab-EU was administered intravenously at a dose of 15 mg/kg. Effects of patient and disease covariates, as well as the drug product (PF-06439535 versus bevacizumab-EU), on PK were investigated. Results Overall, 8632 serum bevacizumab concentrations from 351 patients in the PF-06439535 group and 354 patients in the bevacizumab-EU group were included in the analysis. A two-compartment model adequately described the combined data. Clearance (CL) and central volume of distribution (V1) estimates were 0.0113 L/h and 2.99 L for a typical 71-kg female patient with NSCLC administered bevacizumab-EU. CL and V1 increased with body weight and were higher in males than females even after accounting for differences in body weight. The 95% confidence intervals for the effect of drug product on CL and V1 encompassed unity. Conclusions The population PK of PF-06439535 and bevacizumab-EU were well characterized by a two-compartment model. Covariate analysis did not reveal any appreciable differences between PK parameters for PF-06439535 and bevacizumab-EU in patients with NSCLC. Clinical trial registration ClinicalTrials.gov, NCT02364999.

Finding latent groups in observed data: A primer on latent profile analysis in Mplus for applied researchers

The present guide provides a practical guide to conducting latent profile analysis (LPA) in the Mplus software system. This guide is intended for researchers familiar with some latent variable modeling but not LPA specifically. A general procedure for conducting LPA is provided in six steps: (a) data inspection, (b) iterative evaluation of models, (c) model fit and interpretability, (d) investigation of patterns of profiles in a retained model, (e) covariate analysis, and (f) presentation of results. A worked example is provided with syntax and results to exemplify the steps.