Search

Electronic health records (EHR) provide valuable resources for longitudinal studies and understanding risk factors associated with poor clinical outcomes. However, they may not contain complete follow-ups, and the missing data may not be at random since hospital discharge may depend in part on expected but unrecorded clinical outcomes that occur after patient discharge. These non-ignorable missing data requires appropriate analysis methods. Here, we are interested in measuring and analyzing individual treatment benefits of medical treatments in patients recorded in EHR databases. We present a method for predicting individual benefits that handles non-ignorable missingness due to hospital discharge. The longitudinal clinical outcome of interest is modeled simultaneously with the hospital length of stay using a joint mixed-effects model, and individual benefits are predicted through a frequentist approach: the empirical Bayesian approach. We illustrate our approach by assessing individual pain management benefits to patients who underwent spinal fusion surgery. By calculating sample percentiles of empirical Bayes predictors of individual benefits, we examine the evolution of individual benefits over time. We additionally compare these percentiles with percentiles calculated with a Monte Carlo approach. We showed that empirical Bayes predictors of individual benefits do not only allow examining benefits in specific patients but also reflect overall population trends reliably. [ABSTRACT FROM AUTHOR]

An increasing number of empirical studies aim to quantify individual variation in demographic parameters because these patterns are key for evolutionary and ecological processes. Advanced approaches to estimate individual heterogeneity are now using a multivariate normal distribution with correlated individual random effects to account for the latent correlations among different demographic parameters occurring within individuals. Despite the frequent use of multivariate mixed models, we lack an assessment of their reliability when applied to Bernoulli variables.Using simulations, we estimated the reliability of multivariate mixed effect models for estimating correlated fixed individual heterogeneity in demographic parameters modelled with a Bernoulli distribution. We evaluated both bias and precision of the estimates across a range of scenarios that investigate the effects of life‐history strategy, levels of individual heterogeneity and presence of temporal variation and state dependence. We also compared estimates across different sampling designs to assess the importance of study duration, number of individuals monitored and detection probability.In many simulated scenarios, the estimates for the correlated random effects were biased and imprecise, which highlight the challenge in estimating correlated random effects for Bernoulli variables. The amount of fixed among‐individual heterogeneity was frequently overestimated, and the absolute value of the correlation between random effects was almost always underestimated. Simulations also showed contrasting performances of mixed models depending on the scenario considered. Generally, estimation bias decreases and precision increases with slower pace of life, large fixed individual heterogeneity and large sample size.We provide guidelines for the empirical investigation of individual heterogeneity using correlated random effects according to the life‐history strategy of the species, as well as, the volume and structure of the data available to the researcher. Caution is warranted when interpreting results regarding correlated individual random effects in demographic parameters modelled with a Bernoulli distribution. Because bias varies with sampling design and life history, comparisons of individual heterogeneity among species is challenging. The issue addressed here is not specific to demography, making this warning relevant for all research areas, including behavioural and evolutionary studies. [ABSTRACT FROM AUTHOR]

Bivariate observations of binary and ordinal data arise frequently and require a bivariate modeling approach in cases where one is interested in aspects of the marginal distributions as separate outcomes along with the association between the two. We consider methods for constructing such bivariate models based on latent variables with logistic marginals and propose a model based on the Ali-Mikhail-Haq bivariate logistic distribution. We motivate the model as an extension of that based on the Gumbel type 2 distribution as considered by other authors and as a bivariate extension of the logistic distribution, which preserves certain natural characteristics. Basic properties of the obtained model are studied and the proposed methods are illustrated through analysis of two data sets: a basic science cognitive experiment of visual recognition and awareness and a clinical data set describing assessments of walking disability among multiple sclerosis patients. [ABSTRACT FROM AUTHOR]